research and results zeitschrift

Fachzeitungen.de - Das unabhängige Portal für Fachmagazine Fachpublikationen & eBooks

• ‹‹ vorige Zeitschrift
• nächste Zeitschrift ››

Research & Results

Research & Results, die aktuelle Zeitschrift für Media-, Markt- und Werbeforschung, hat sich in der Branche erfolgreich etabliert. Nicht umsonst steht Research & Results als praxisnahes, anwendungsorientiertes und doch fachlich fundiertes Magazin für Informationskompetenz in der Forschungsbranche. Denn das Inhaber- und Herausgeber-Team von Research & Results kann auf langjährige Erfahrung in der Forschungs-, Medien- und Marketingbranche verweisen.

Publikationen zum Thema:

• Research - Results - Media - Markt

Zeitschriften zum Thema: Marktforschung - Planung - Analyse - Motivationsforschung

Sales excellence.

Sales Excellence ist die führende Plattform für Vertriebsexperten. Unsere Redaktionen aus Print und Online bieten Ihnen fundierte, praxisorientierte Fachinformation, Hintergrundberichte sowie ...

fitness MANAGEMENT international

fitness MANAGEMENT international I Offizielles Organ des Arbeitgeberverbands deutscher Fitness- und Gesundheits-Anlagen e.V. (DSSV e.V.) I Erscheint alle 2 Monate in einer Auflage von 20.000 ...

iBusiness Executive Summary

iBusiness, 1991 entstanden, ist der Thinktank für Zukunftsforschung rund um digitale, interaktive und disruptive Märkte. Strategische Entscheider aus Unternehmen und Agenturen, Dienstleister und ...

HORIZONT (Österreich)

Jeden Freitag neu: Die Wochenzeitung HORIZONT bietet aktuelle, ausführliche Berichterstattung über die Welt der Medien, des Marketings, der Werbung und der Kommunikation. Wir ...

planung & analyse

planung & analyse veröffentlicht acht Mal im Jahr praxisorientierte, aktuelle und fundierte Originalbeiträge zu Projekten, Konzepten und Methoden der Markt- und Marketingforschung. Im ...

HORIZONT ist die Fachzeitung für Marketing, Werbung und Medien und berichtet wöchentlich über das gesamte Spektrum von Marketing und Kommunikation in Unternehmen, Agenturen und Medien. Mit ...

E-Books zum Thema: Marketing - Werbung - PR

Dieses Lehrbuch vermittelt die gesamten Grundlagen des Marketing und nimmt dabei konsequent Bezug auf die Praxis. Die anwendungsorientierte Darstellung versetzt den Leser in die Lage,…

Der Profi-Verkäufer: Wegweiser zum Erfolg mit 100 Motivationsimpulsen für den Alltag

Das Handbuch für den Top-Verkäufer Jedes Kapitel steckt voller Anregungen und Tips für gekonntes Verkaufen. Die Analyse eigener Verhaltensmuster liefert Impulse zur kontinuierlichen…

Die Zukunft des Kunstmarktes

Mit Beiträgen von Jean-Christophe Ammann, Bazon Brock, Peter Conradi, Wilfried Dörstel, Andreas Dornbracht, Walter Grasskamp, Bernd Günter/Andrea Hausmann, Arnold Hermanns, Bernd Kauffmann, Tobias…

Management von Kundenbeziehungen. 7 Bausteine für ein effizientes Kundenmanagement

Bei uns ist der Kunde König!" - Das behaupten viele Unternehmen. Doch damit sich der Kunde tatsächlich als König fühlt und langfristig Wachstum und Geschäftserfolg sichert, müssen die Beziehungen…

Texten für das Web

Dieses Buch bietet das nötige Handwerkszeug, um die Qualität der eigenen Web-Texte zu verbessern bzw. eingekaufte Texte sicherer beurteilen zu können. Es liefert klare Kriterien für die Textanalyse,…

Total Loyalty Marketing

Die immer gleiche Marke als Lebensabschnittsbegleiter hat ausgedient - könnte man meinen. Doch zu einigen Produkten und Dienstleistungen haben deren Verwender eine besondere Beziehung: eine…

Professionelles Direkt- und Dialogmarketing per E-Mail

Inhalte richtig formulieren und gestalten E-Mails sicher zustellen und auswerten Praxistipps und Fallbeispiele2., vollständig überarbeitete und erweiterte Auflage!Die Bedeutung von Direkt- und…

Die 11 Geheimnisse des IKEA-Erfolgs

Schaust du noch oder liest du schon? IKEA ist Kult. IKEA ist unglaublich erfolgreich. Sein umstrittener Gründer Ingvar Kamprad gilt manchen (vor allem Schweden) heute als der reichste Mann der Welt.…

Wachstumsmarkt Alter

Produkte und Dienstleistungen, welche sich an den Bedürfnissen älterer Menschen orientieren, sind derzeit noch Mangelware, bilden aber einen enormen Wachstumsmarkt. Wie Sie von diesem Markt…

Die Marke in der Marke

Marketing lebt von der Innovation und dem Kundennutzen. Unternehmen, die beides kombinieren können, sind auf dem Weg zum Erfolg. Mit Ingredient Branding bekommen Lieferanten und Endprodukt-Hersteller…

• Architecture and Design
• Asian and Pacific Studies
• Business and Economics
• Classical and Ancient Near Eastern Studies
• Computer Sciences
• Cultural Studies
• Engineering
• General Interest
• Geosciences
• Industrial Chemistry
• Islamic and Middle Eastern Studies
• Jewish Studies
• Library and Information Science, Book Studies
• Life Sciences
• Linguistics and Semiotics
• Literary Studies
• Materials Sciences
• Mathematics
• Social Sciences
• Sports and Recreation
• Theology and Religion
• Publish your article
• The role of authors
• Promoting your article
• Abstracting & indexing
• Publishing Ethics
• Why publish with De Gruyter
• How to publish with De Gruyter
• Our book series
• Our subject areas
• Your digital product at De Gruyter
• Contribute to our reference works
• Product information
• Tools & resources
• Product Information
• Promotional Materials
• Orders and Inquiries
• FAQ for Library Suppliers and Book Sellers
• Repository Policy
• Free access policy
• Open Access agreements
• Database portals
• For Authors
• Customer service
• People + Culture
• Journal Management
• How to join us
• Working at De Gruyter
• Mission & Vision
• De Gruyter Foundation
• De Gruyter Ebound
• Our Responsibility
• Partner publishers

Your purchase has been completed. Your documents are now available to view.

The Current State of Big Data Research in Tourism : Results of a Systematic Literature Analysis

Kim Hartmann is PhD Candidate at the Department of Marketing, Strathclyde Business School, University of Strathclyde (UK) and lecturer at the International School of Management (ISM) in Munich. She holds tourism and business degrees from ISM Dortmund, Hochschule Worms University of Applied Sciences, and University of the Sunshine Coast. Her professional background includes roles in project management and marketing for the FMCG, cruise, and accommodation industries. Her teaching and research activities focus on tourism management and marketing.

Matthias Lederer is Professor for Information Systems at the OTH Technical University of Applied Sciences Amberg-Weiden. Prior to this, he was a professor at the International School of Management (ISM) in Munich and at the same time Chief Process Officer at the IT Service Center of the Bavarian justice system. His previous positions include research assistant at the University of Erlangen-Nuremberg and strategy consultant at the German industrial company REHAU. His research and studies focus on business process management and IT management. Prof. Lederer holds a doctorate & master’s degree in international information systems and is the author of over 60 scientific publications in this field.

The use of large and diverse data in real time (called Big Data) affects many business processes and models. The tourism industry, characterized by manifold sub-sectors and players, provides a variety of starting points for Big Data usage. Examples are the optimization of transport offers using transaction data or a comprehensive analysis of destination trends based on social media posts. Big Data is a trending topic, however, the general discourse centres around potential ideas but fewer practical solutions. Based on a systematic literature analysis of initially 148 peer-reviewed journal articles, this article evaluates the current state of Big Data research within tourism. For this purpose, research articles centering around tourism-related Big Data were investigated according to the actual state of implementation of an IT solution, whether they truly grasp or represent Big Data in technological terms, and which added value they create for the tourism industry and research community. One key finding is that traditional data analysis is often wrongfully subsumed under the Big Data label. Further, the scientific literature predominantly discusses ideas or theoretical considerations, fewer tangible Big Data implementations, and fails to address and/or meet all requirements to be classified as Big Data. Only a minority of the presented solutions processes data in real time, whereas many rely on only one data source or structured data. Furthermore, most articles revolve around post-trip data analyses and are set to a destination context. In contrast, other tourism sectors as well as data interpretation and usage in pre-trip and on-trip phases are less represented. Lastly, this literature analysis provides an overview of true Big Data solutions already in operation and enables researchers to validly classify their own research activities in order to plan initiatives more specifically.

About the authors

Athanasopoulos, G., Hyndman, R.J., Song, H., & Wu, D.C. (2011). The tourism forecasting competition. International Journal of Forecasting , 27(3), 822–844. 10.1016/j.ijforecast.2010.04.009 Search in Google Scholar

Alaei, A.R., Becken, S., & Stantic, B. (2019). Sentiment analysis in tourism: Capitalizing on big data. Journal of Travel Research , 58(2), 175–191. 10.1177/0047287517747753 Search in Google Scholar

Basaran, M.A., Dogan, S., & Kantarci, K. (2020). On modeling of responses generated by travel 2.0 implementation: Fuzzy rule-based systems. International Journal of Contemporary Hospitality Management , 32(4), 1503–1522. 10.1108/IJCHM-03-2019-0279 Search in Google Scholar

Batistič, S., & der Laken, P. (2019). History, evolution and future of big data and analytics: A bibliometric analysis of its relationship to performance in organizations. British Journal of Management , 30(2), 229–251. 10.1111/1467-8551.12340 Search in Google Scholar

Best, W., & Weth, M. (2010). Process Excellence . Cham: Springer. 10.1007/978-3-8349-8950-5 Search in Google Scholar

Bhat, W., & Quadri, A. (2015). Big Data promises value: is hardware technology taken onboard?. Industrial Management & Data Systems , 115(9), 1577–1595. 10.1108/IMDS-04-2015-0160 Search in Google Scholar

Booth, A., Sutton, A., & Papaioannou, D. (2016). Systematic approaches to a successful literature review . Los Angeles: SAGE. Search in Google Scholar

Centobelli, P., & Ndou, V. (2019). Managing customer knowledge through the use of big data analytics in tourism research. Current Issues in Tourism , 22(15), 1862–1882. 10.1080/13683500.2018.1564739 Search in Google Scholar

Chen, M., Mao, S., Zhang, Y., & Leung, V.C.M. (2014). Big data: Related technologies, challenges and future prospects . Cham: Springer. Search in Google Scholar

Cheng, X., & Zhao, C. (2019). Prediction of tourist consumption based on bayesian network and big data. Ingénierie Des Systèmes D’Information , 24(5), 491–497. 10.18280/isi.240505 Search in Google Scholar

Cooper, C., Fletcher, J., Fyall, A., Gilbert, D., & Wanhill, S. (2008). Tourism: Principles and practices (4th ed.). Essex: Pearson Education. Search in Google Scholar

Del Vecchio, P., Mele, G., Ndou, V., & Secundo, G. (2018). Creating value from social big data: Implications for smart tourism destinations. Information Processing & Management , 54(5), 847–860. 10.1016/j.ipm.2017.10.006 Search in Google Scholar

Donovan, C., Flaherty, E.T., & Quinn Healy, E. (2017). Using big data from Wikipedia page views for official tourism statistics. Statistical Journal of the IAOS , 33(4), 997–1003. 10.3233/SJI-160320 Search in Google Scholar

Edwards, D., Cheng, M., Wong, I.A., Zhang, J., & Wu, Q. (2017). Ambassadors of knowledge sharing: Co-produced travel information through tourist-local social media exchange. International Journal of Contemporary Hospitality Management , 29(2), 690–708. 10.1108/IJCHM-10-2015-0607 Search in Google Scholar

Fronzetti Colladon, A., Guardabascio, B., & Innarella, R. (2019). Using social network and semantic analysis to analyze online travel forums and forecast tourism demand. Decision Support Systems , 123, 113075. 10.1016/j.dss.2019.113075 Search in Google Scholar

Fuchs, M., Höpken, W., & Lexhagena, M. (2014). Big data analytics for knowledge generation in tourism destinations – A case from Sweden. Journal of Destination Marketing & Management , 3(4), 198–209. 10.1016/j.jdmm.2014.08.002 Search in Google Scholar

Gandomi, A., & Haider, M. (2015). Beyond the hype: Big data concepts, methods, and analytics. International Journal of Information Management , 35(2), 137–144. 10.1016/j.ijinfomgt.2014.10.007 Search in Google Scholar

Gibbs, C., Guttentag, D., Gretzel, U., Yao, L., & Morton, J. (2018). Use of dynamic pricing strategies by Airbnb hosts. International Journal of Contemporary Hospitality Management , 30(1), 2–20. 10.1108/IJCHM-09-2016-0540 Search in Google Scholar

Gupta, U.G., & Gupta, A. (2016). Vision: A missing key dimension in the 5V big data framework. Journal of International Business Research and Marketing , 1(3), 50–56. 10.18775/jibrm.1849-8558.2015.13.3005 Search in Google Scholar

Happ, E., & Ivancsó-Horváth, Z. (2018). Digital tourism is the challenge of future – a new approach to tourism. Knowledge Horizons – Economics , 10(2), 9–16. Search in Google Scholar

Hasuike, T., & Ichimura, T. (2013). Web intelligence for tourism using railway data by a simplified fuzzy reasoning method. Journal of Intelligent & Fuzzy Systems , 24(2), 251–259. 10.3233/IFS-2012-0551 Search in Google Scholar

Holloway, J.C., Humphreys, C., & Davidson, R. (2009). The business of tourism (8th ed.). Essex: Pearson Education. Search in Google Scholar

Holmlund, M., Van Vaerenbergh, Y., Ciuchita, R.v, Ravald, A., Sarantopoulos, P., Villarroel Ordenes, F., & Zaki, M. (2020). Customer experience management in the age of big data analytics: A strategic framework. Journal of Business Research , 116, 356–365. 10.1016/j.jbusres.2020.01.022 Search in Google Scholar

Huang, Z., Cao, F., Jin, C., Yu, Z., & Huang, R. (2017). Carbon emission flow from self-driving tours and its spatial relationship with scenic spots – A traffic-related big data method. Journal of Cleaner Production , 142, 946–955. 10.1016/j.jclepro.2016.09.129 Search in Google Scholar

Iafrate, F. (2015). From big data to smart data . London: Wiley. 10.1002/9781119116189 Search in Google Scholar

Jun, S.H., Vogt, C.A., & MacKay, K.J. (2007). Relationships between travel information search and travel product purchase in pretrip contexts. Journal of Travel Research , 45(February), 266–274. 10.1177/0047287506295945 Search in Google Scholar

Kandampully, J., Zhang, T., & Jaakkola, E. (2018). Customer experience management in hospitality: A literature synthesis, new understanding, and research agenda. International Journal of Contemporary Hospitality Management , 30(1), 21–56. 10.1108/IJCHM-10-2015-0549 Search in Google Scholar

Kirilenko, A.P., Stepchenkova, S.O., Kim, H., & Li, X. (2018). Automated sentiment analysis in tourism: Comparison of approaches. Journal of Travel Research , 57(8), 1012–1025. 10.1177/0047287517729757 Search in Google Scholar

Kornmeier, M. (2007). Wissenschaftstheorie und wissenschaftliches Arbeiten . Physica, Mannheim. 10.1007/978-3-7908-1919-9 Search in Google Scholar

Laudon, C., & Laudon, J.P. (2017). Management Information Systems: Managing the Digital Firm . London: Pearson. Search in Google Scholar

Lederer, M. (2019). What’s going to happen to Business Process Management? Current Status and Future of a Discipline. In Proceedings of the S-BPM ONE 2019. Sevilla: CEUR-WS. Search in Google Scholar

Lederer, M., & Lederer J. (2021): Are they Ready for the Big Thing? Big Data Applications Requirements for Process Management & Evaluation of Current Software Solutions. In: J. P. Chigwada & G. Tsvuura (Hrsg.): Handbook of Research on Information and Records Management in the Fourth Industrial Revolution . Hershey: Information Science Reference. 10.4018/978-1-7998-7740-0.ch013 Search in Google Scholar

Lee, H., Chung, N., & Nam, Y. (2019). Do online information sources really make tourists visit more diverse places? Based on the social networking analysis. Information Processing & Management , 56(4), 1376–1390. 10.1016/j.ipm.2018.01.005 Search in Google Scholar

Li, J., Xu, L., Tang, L., Wang, S., & Li, L. (2018). Big data in tourism research: A literature review. Tourism Management , 68, 301–323. 10.1016/j.tourman.2018.03.009 Search in Google Scholar

Lin, N. (2014). Applied Business Analytics: Integrating Business Process, Big Data, and Advanced Analytics . Jew Jersey: FT Press. Search in Google Scholar

Liu, Q., Wang, Z., & Ye, X. (2018). Comparing mobility patterns between residents and visitors using geo‐tagged social media data. Transactions in GIS , 22(6), 1372–1389. 10.1111/tgis.12478 Search in Google Scholar

Mandal, S. (2018). Exploring the influence of big data analytics management capabilities on sustainable tourism supply chain performance: The moderating role of technology orientation. Journal of Travel & Tourism Marketing , 35(8), 1104–1118. 10.1080/10548408.2018.1476302 Search in Google Scholar

Mariani, M., Baggio, R., Fuchs, M.v & Höpken, W. (2018). Business intelligence and big data in hospitality and tourism: a systematic literature review. International Journal of Contemporary Hospitality Management , 30(12), 3514–3554. 10.1108/IJCHM-07-2017-0461 Search in Google Scholar

Mayring, P. (2000). Qualitative content analysis. Forum Qualitative Sozialforschung , 1(2). Search in Google Scholar

Medlik, S. (2002). Dictionary of travel, tourism, and hospitality (3rd ed.). Oxford: Elsevier. Search in Google Scholar

Miah, S.J., Vu, H.Q., Gammack, J., & McGrath, M. (2017). A Big Data Analytics Method for Tourist Behaviour Analysis. Information & Management , 54(6), 771–785. 10.1016/j.im.2016.11.011 Search in Google Scholar

Min, W., Yu, L., Yu, L., & He, S. (2018). People logistics in smart cities. Communications of the ACM , 61(11), 54–59. 10.1145/3239546 Search in Google Scholar

Mindur, M. (2019). The determination of potential locations for hotel and service facilities in relation to the transport system – the logistic approach. Logistics & Transport , 43(3), 27–35. 10.26411/83-1734-2015-3-43-1-19 Search in Google Scholar

Moro, S., Esmerado, J., Ramos, P., & Alturas, B. (2019). Evaluating a guest satisfaction model through data mining. International Journal of Contemporary Hospitality Management , 32(4), 1523–1538. 10.1108/IJCHM-03-2019-0280 Search in Google Scholar

Nilashi, M., Ahani, A., Esfahani, M.D., Yadegaridehkordi, E., Samad, S., Ibrahim, O., Sharef, N.M., & Akbari, E. (2019). Preference learning for eco-friendly hotels recommendation: A multi-criteria collaborative filtering approach. Journal of Cleaner Production , 215, 767–783. 10.1016/j.jclepro.2019.01.012 Search in Google Scholar

Nilashi, M., Bin Ibrahim, O., Ithnin, N., & Sarmin, N.H. (2015). A multi-criteria collaborative filtering recommender system for the tourism domain using Expectation Maximization (EM) and PCA–ANFIS. Electronic Commerce Research and Applications , 14(6), 542–562. 10.1016/j.elerap.2015.08.004 Search in Google Scholar

Önder, I., Koerbitz, W., & Hubmann-Haidvogel, A. (2016). Tracing tourists by their digital footprints. Journal of Travel Research , 55(5), 566–573. 10.1177/0047287514563985 Search in Google Scholar

Önder, I. (2017). Forecasting tourism demand with Google trends: Accuracy comparison of countries versus cities. The International Journal of Tourism Research , 19(6), 648–660. 10.1002/jtr.2137 Search in Google Scholar

Pan, B., & Yang, Y. (2017). Forecasting destination weekly hotel occupancy with big data. Journal of Travel Research , 56(7), 957–970. 10.1177/0047287516669050 Search in Google Scholar

Pantano, E., & Dennis, C. (2019). Store buildings as tourist attractions: Mining retail meaning of store building pictures through a machine learning approach. Journal of Retailing and Consumer Services , 51, 304–310. 10.1016/j.jretconser.2019.06.018 Search in Google Scholar

Park, S.B., Ok, C.M., & Chae, B.K. (2016). Using Twitter data for cruise tourism marketing and research. Journal of Travel & Tourism Marketing , 33(6), 885–898. 10.1080/10548408.2015.1071688 Search in Google Scholar

Petcu, N. (2011). Statistical tests applied in tourism. Economic Sciences , 4(1), 135–140. Search in Google Scholar

Pizam, A. (2009). What is the hospitality industry and how does it differ from the tourism and travel industries? International Journal of Hospitality Management, 28, 183–184. 10.1016/j.ijhm.2009.01.001 Search in Google Scholar

Pohl, K. (2010). Requirements Engineering: Fundamentals, Principles, and Techniques . Cham: Springer. 10.1007/978-3-642-12578-2 Search in Google Scholar

Ren, Q., Xu, F., & Ji, X. (2019). Use of the pathfinder network scaling to measure online customer reviews: A theme park study. Strategic Change , 28(5), 333–344. 10.1002/jsc.2288 Search in Google Scholar

Scuderi, R., & Dalle Nogare, C. (2018). Mapping tourist consumption behaviour from destination card data: What do sequences of activities reveal?. The International Journal of Tourism Research , 20(5), 554–565. 10.1002/jtr.2205 Search in Google Scholar

Shapoval, V., Wang, M.C., Hara, T., & Shioya, H. (2018). Data mining in tourism data analysis: Inbound visitors to Japan. Journal of Travel Research , 57(3), 310–323. 10.1177/0047287517696960 Search in Google Scholar

Shereni, N.C., & Chambwe, M. (2020). Hospitality big data analytics in developing countries. Journal of Quality Assurance in Hospitality & Tourism , 21(3), 361–369. 10.1080/1528008X.2019.1672233 Search in Google Scholar

Sinha, S., Bhatnagar, V., & Bansal, A (2018). Multi-label naïve bayes classifier for identification of top destination and issues to accost by tourism sector. Journal of Global Information Management , 26(3), 37–54. 10.4018/JGIM.2018070104 Search in Google Scholar

Song, H., Gao, B.Z., & Lin, V.S. (2013). Combining statistical and judgmental forecasts via a web-based tourism demand forecasting system. International Journal of Forecasting , 29(2), 295–310. 10.1016/j.ijforecast.2011.12.003 Search in Google Scholar

Stryk, B. (2015). How do organizations prepare and clean big data to achieve better data governance? A Delphi Study . Dissertation, Capella University. Search in Google Scholar

Stylos, N., & Zwiegelaar, J. (2019). Big data as a game changer: How does it shape business intelligence within a tourism and hospitality industry context? In: Marianna Sigala, Roya Rahimi, Mike Thelwall (Eds.): Big Data and Innovation in Tourism, Travel, and Hospitality (p.163–181). Cham: Springer. 10.1007/978-981-13-6339-9_11 Search in Google Scholar

Tingting, S. (2019). Spatial data mining and big data analysis of tourist travel behavior. Ingénierie Des Systèmes D’Information , 24(2), 167–173. 10.18280/isi.240206 Search in Google Scholar

Tsai, T. (2014). A self-learning advanced booking model for railway arrival forecasting. Transportation Research. Part C, Emerging Technologies , 39, 80–93. 10.1016/j.trc.2013.11.021 Search in Google Scholar

Vaid, J. (2018). Role of big data analytics in social media marketing of MICE tourism. Global Journal of Enterprise Information System , 10(1), 55–62. Search in Google Scholar

Vinod, B. (2016). Big data in the travel marketplace. Journal of Revenue and Pricing Management , 15(5), 352–359. 10.1057/rpm.2016.30 Search in Google Scholar

Webster, J., Watson, R.T. (2002). Analyzing the past to prepare for the future: Writing a literature review. MIS Quarterly , 26, 13–23. Search in Google Scholar

Wu, H.-Q., Liu, Y.-X., Zhuang, D.-L., & Zhu, W.-D. (2014). A DEA-based measurement of effectiveness of provincial image advertisement for local tourism destination: Evidence from China. International Journal of Management & Decision Making , 13(2), 192–207. 10.1504/IJMDM.2014.060860 Search in Google Scholar

Xiang, Z. (2018). From digitization to the age of acceleration: On information technology and tourism. Tourism Management Perspectives , 25, 147–150. 10.1016/j.tmp.2017.11.023 Search in Google Scholar

Xu, F., Nash, N., & Whitmarsh, L. (2020). Big data or small data? A methodological review of sustainable tourism. Journal of Sustainable Tourism , 28(2), 147–166. 10.1080/09669582.2019.1631318 Search in Google Scholar

Yadegaridehkordi, E., Nilashi, M., Shuib, L., Hairul Nizam Bin Md Nasir, M., Asadi, S., Samad, S., & Fatimah Awang, N. (2020). The impact of big data on firm performance in hotel industry. Electronic Commerce Research and Applications , 40, 100921. 10.1016/j.elerap.2019.100921 Search in Google Scholar

Yang, Y., Pan, B., & Song, H. (2014). Predicting hotel demand using destination marketing organization’s web traffic data. Journal of Travel Research , 53(4), 433–447. 10.1177/0047287513500391 Search in Google Scholar

Yao, B., Qiu, R.T.R., Fan, D.X.F., Liu, A., & Buhalis, D. (2019). Standing out from the crowd – an exploration of signal attributes of Airbnb listings. International Journal of Contemporary Hospitality Management , 31(12), 4520–4542. 10.1108/IJCHM-02-2019-0106 Search in Google Scholar

Zhao, P., & Hu, H. (2019). Geographical patterns of traffic congestion in growing megacities: Big data analytics from Beijing. Cities , 92, 164–174. 10.1016/j.cities.2019.03.022 Search in Google Scholar

© 2021 Walter de Gruyter GmbH, Berlin/Boston

• X / Twitter

Supplementary Materials

Please login or register with De Gruyter to order this product.

Journal and Issue

Articles in the same issue.

Cross-National Comparative Research—Analytical Strategies, Results, and Explanations

International vergleichende Forschung – Analysestrategien, Ergebnisse und Erklärungen

• Abhandlungen
• Published: 22 May 2019
• Volume 71 , pages 1–28, ( 2019 )

Cite this article

• Hans-Jürgen Andreß 1 ,
• Detlef Fetchenhauer 1 &
• Heiner Meulemann 1  

1945 Accesses

13 Citations

Explore all metrics

This introductory article reviews the history of cross-national comparative research, discusses its typical research designs and research questions, and ultimately summarizes the contributions to this special issue with respect to two questions: (i) What are the methodological challenges of cross-national comparative research today? (ii) What typical effects of the national context have been identified up to now?

Zusammenfassung

In diesem einleitenden Artikel wird die Geschichte der ländervergleichenden Forschung dargestellt, es werden die typischen Forschungsdesigns und Forschungsfragen erörtert und schließlich die Beiträge dieses Sonderhefts in Bezug auf zwei Fragen zusammengefasst: (i) Was sind die methodologischen Herausforderungen der ländervergleichenden Forschung heute? (ii) Welche typischen Auswirkungen des nationalen Kontexts wurden bisher festgestellt?

This is a preview of subscription content, log in via an institution to check access.

Access this article

Price excludes VAT (USA) Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Rent this article via DeepDyve

Institutional subscriptions

Similar content being viewed by others

Internationally Comparative Research Designs in the Social Sciences: Fundamental Issues, Case Selection Logics, and Research Limitations

Achim Goerres, Markus B. Siewert & Claudius Wagemann

Multilevel Structural Equation Modeling for Cross-National Comparative Research

Bart Meuleman

Comparative Economic Studies and Comparative Economics: Six Decades and Counting

Josef C. Brada & Paul Wachtel

The analysis of context effects is not only prominent in CNCR, but also in regional science and urban sociology. The 2014 special issue of the Kölner Zeitschrift für Soziologie und Soziologie discusses predominantly local contexts such as urban districts or other lower-level regional units (Friedrichs and Nonnenmacher 2014 ).

The Survey of Health, Ageing and Retirement in Europe (SHARE 2019 ) and the Generations and Gender Survey (GGS 2019 ) are two such examples.

Nonnenmacher and Friedrichs ( 2013 ) review 22 articles using at least one of these different forms of multi-country studies to explain life satisfaction.

To be precise: it takes account of all unobserved heterogeneity that is uncorrelated with the explanatory variables.

Data are available on request from the first author. The following journals were analyzed: American Sociological Review, European Sociological Review, International Journal of Sociology, American Journal of Political Science, European Journal of Political Research, Political Research Quarterly, and Social Science Research .

Of course, countries which have the same welfare regime may install councils in order to learn from each other—as the Scandinavian welfare states did. If such councils attain power over their constituent countries, they can become a collective actor in their own right, and the borderline from aggregation to social reality will be transgressed—just as in the case of the European Union. Furthermore, such councils are examples of the interaction between collective actors, which is beyond the purview of CNCR. International relations may be a complementary research arena to cross-national comparison.

Alderson, Arthur S., and Francois Nielsen. 2002. Globalization and the great U‑turn: Income inequality trends in 16 OECD countries. American Journal of Sociology 107:1244–1299.

Article   Google Scholar  

Andreß, Hans-Jürgen, Katrin Golsch and Alexander W. Schmidt. 2013. Applied panel analysis for economic and social surveys . Berlin: Springer.

Book   Google Scholar  

Blau, Peter M., and Joseph E. Schwartz. 1984. Crosscutting social circles: Testing a macrostructural theory of intergroup relations . Orlando: Academic Press.

Google Scholar  

Blossfeld, Hans-Peter, Nevena Kulic, Jan Skopek, Moris Triventi, Elina Kilpi-Jakonen, Daniela Vono de Vilhena, and Sandra Buchholz. 2019. Conditions and consequences of unequal educational opportunities in the life course: Results from the Cross-National Comparative eduLIFE Project. In Cross-national comparative research – analytical strategies, results and explanations . Sonderheft Kölner Zeitschrift für Soziologie und Sozialpsychologie . Eds. Hans-Jürgen Andreß, Detlef Fetchenhauer and Heiner Meulemann. Wiesbaden: Springer VS. https://doi.org/10.1007/s11577-019-00595-w .

Bollen, Kenneth A., Barbara Entwisle and Arthur S. Alderson. 1993. Macrocomparative research methods. Annual Review of Sociology 19:321–351.

Boomgaarden, Hajo G., and Hyunjin Song. 2019. Media use and its effects in a cross-national perspective. In Cross-national comparative research – analytical strategies, results and explanations . Sonderheft Kölner Zeitschrift für Soziologie und Sozialpsychologie . Eds. Hans-Jürgen Andreß, Detlef Fetchenhauer and Heiner Meulemann. Wiesbaden: Springer VS. https://doi.org/10.1007/s11577-019-00596-9 .

Bornschier, Volker, and Christopher K. Chase-Dunn. 1985. Transnational corporations and underdevelopment . New York: Praeger.

Boyd, Lawrence H., and Gudmund R. Iversen. 1979. Contextual analysis: Concepts and statistical techniques . Belmont, CA: Wadsworth.

Bryk, Anthony S., and Stephen W. Raudenbush. 2002. Hierarchical linear models: Applications and data analysis methods. Second edition. Newbury Park: Sage.

Careja, Romana. 2019. Immigrants’ labor market outcomes: Contributions from multilevel studies. In Cross-national comparative research – analytical strategies, results and explanations . Sonderheft Kölner Zeitschrift für Soziologie und Sozialpsychologie . Eds. Hans-Jürgen Andreß, Detlef Fetchenhauer and Heiner Meulemann. Wiesbaden: Springer VS. https://doi.org/10.1007/s11577-019-00597-8 .

Cieciuch, Jan, Eldad Davidov, Peter Schmidt and René Algesheimer. 2019. How to obtain comparable measures for cross-national comparisons. In Cross-national comparative research – analytical strategies, results and explanations . Sonderheft Kölner Zeitschrift für Soziologie und Sozialpsychologie . Eds. Hans-Jürgen Andreß, Detlef Fetchenhauer and Heiner Meulemann. Wiesbaden: Springer VS. https://doi.org/10.1007/s11577-019-00598-7 .

CNCR. 2019. Cross-national Comparative Research. KZfSS Special Issue—(website). http://eswf.uni-koeln.de/cncr/ (Date of access: 8 Feb. 2019).

Cohen, Jeffrey E. 2004. Economic perceptions and executive approval in comparative perspective. Political Behavior 26:27–43.

Coleman, James S. 1990. Foundations of social theory . Cambridge, MA: Belknap Press of Harvard University Press.

Context. 2019. ESS Multilevel Data. https://www.europeansocialsurvey.org/data/multilevel/guide/about.html (Date of access: 29 Jan. 2019).

Davidov, Eldad, Bart Meuleman, Jan Cieciuch, Peter Schmidt and Jaak Billiet. 2014. Measurement equivalence in cross-national research. Annual Review of Sociology 40:55–75.

Davidov, Eldad, Peter Schmidt, Jaak Billiet and Bart Meuleman. 2018. Cross-cultural analysis: Methods and applications, second edition . New York: Taylor & Francis.

Diener, Ed, and Shigehiro Oishi. 2000. Money and happiness: Income and subjective well-being across nations. In Culture and subjective well-being , eds. Ed Diener and Eunkook Mark Suh, 185–218. Cambridge: The MIT Press.

Diener, Ed, Louis Tay and David G. Myers. 2011. The religion paradox: If religion makes people happy, why are so many dropping out? Journal of Personality and Social Psychology 101:1278–1290.

Easterlin, Richard A. 1974. Does economic growth improve the human lot? Some empirical evidence. In Nations and households in economic growth: Essays in honor of Moses Abramovitz , eds. Paul A. David and Melvin W. Reder, 89–125. New York: Academic Press.

Erlinghagen, Marcel. 2019. Employment and its institutional contexts. In Cross-national comparative research – analytical strategies, results and explanations . Sonderheft Kölner Zeitschrift für Soziologie und Sozialpsychologie . Eds. Hans-Jürgen Andreß, Detlef Fetchenhauer and Heiner Meulemann. Wiesbaden: Springer VS. https://doi.org/10.1007/s11577-019-00599-6 .

ESS. 2019. European Social Survey. https://www.europeansocialsurvey.org (Date of access: 29 Jan. 2019).

EVS. 2019. European Values Study. https://europeanvaluesstudy.eu (Date of access: 29 Jan. 2019).

Flora, Peter. 1983 and 1986. State, economy, and society in Western Europe . Frankfurt a. M.: Campus.

Freedman, David A. 2004. The ecological fallacy. In The SAGE Encyclopedia of Social Science Research Methods, Volume 1 , eds. Michael S. Lewis-Beck, Alan Bryman and Tim Futing Liao, 293. Thousand Oaks: Sage.

Frick, Joachim R., Stephen P. Jenkins, Dean R. Lillard, Oliver Lipps and Mark Wooden. 2007. The Cross-National Equivalent File (CNEF) and its member country household panel studies. Schmollers Jahrbuch 127:627–654.

Friedrichs, Jürgen, and Alexandra Nonnenmacher. 2014. Die Analyse sozialer Kontexte (An analysis of social contexts). In Soziale Kontexte und soziale Mechanismen (Kölner Zeitschrift für Soziologie Sonderheft 54/2014) , eds. Jürgen Friedrichs and Alexandra Nonnenmacher, 1–16. Wiesbaden: Springer.

Gallie, Duncan, and Serge Paugam. Eds. 2000. Welfare regimes and the experience of unemployment in Europe . Oxford: Oxford University Press.

Gauthier, Anne Hélène. 1996. The state and the family: a comparative analysis of family policies in industrialized countries . Oxford: Clarendon Press.

GGS. 2019. Generations and Gender Survey. https://www.ggp-i.org (Date of access: 29 Jan. 2019).

Giesselmann, Marco, and Alexander W. Schmidt-Catran. 2018. Getting the within estimator of cross-level interactions in multilevel models with pooled cross-sections: Why country dummies (sometimes) do not do the job. Sociological Methodology Online First, https://doi.org/10.1177/0081175018809150 .

Goerres, Achim, Markus B. Siewert and Claudius Wagemann. 2019. Internationally comparative research designs in the social sciences: Fundamental issues, case selection logics, and research limitations. In Cross-national comparative research – analytical strategies, results and explanations . Sonderheft Kölner Zeitschrift für Soziologie und Sozialpsychologie . Eds. Hans-Jürgen Andreß, Detlef Fetchenhauer and Heiner Meulemann. Wiesbaden: Springer VS. https://doi.org/10.1007/s11577-019-00600-2 .

Goldstein, Harvey. 1986. Multilevel mixed linear model analysis using iterative generalized least squares. Biometrika 73:43–56.

Grunow, Daniela. 2019. Comparative analyses of housework and its relation to paid work: Institutional contexts and individual agency. In Cross-national comparative research – analytical strategies, results and explanations . Sonderheft Kölner Zeitschrift für Soziologie und Sozialpsychologie . Eds. Hans-Jürgen Andreß, Detlef Fetchenhauer and Heiner Meulemann. Wiesbaden: Springer VS. https://doi.org/10.1007/s11577-019-00601-1 .

Guerin, Daniel, Jean Crete and Jean Mercier. 2001. A multilevel analysis of the determinants of recycling behavior in the European countries. Social Science Research 30:195–218.

Hall, Peter A., and Rosemary C. R. Taylor. 1996. Political science and the three new institutionalisms. Political Studies 44:936–957.

Halman, Loek, and John Gelissen. 2019. Values in life domains in a cross-national perspective. In Cross-national comparative research – analytical strategies, results and explanations . Sonderheft Kölner Zeitschrift für Soziologie und Sozialpsychologie . Eds. Hans-Jürgen Andreß, Detlef Fetchenhauer and Heiner Meulemann. Wiesbaden: Springer VS.

Hank, Karsten, and Anja Steinbach. 2019. Families and their institutional contexts: The role of family policies and legal regulations. In Cross-national comparative research – analytical strategies, results and explanations . Sonderheft Kölner Zeitschrift für Soziologie und Sozialpsychologie . Eds. Hans-Jürgen Andreß, Detlef Fetchenhauer and Heiner Meulemann. Wiesbaden: Springer VS. https://doi.org/10.1007/s11577-019-00603-z .

Hantrais, Linda. 2008. International comparative research: Theory, methods and practice . Basingstoke: Palgrave Macmillan.

Harkness, Janet A., Michael Braun, Brad Edwards, Timothy P. Johnson, Lars Lyberg, Peter Ph. Mohler, Beth-Ellen Pennell and Tom W. Smith. 2010. Survey methods in multinational, multiregional, and multicultural contexts . New York: Wiley.

Harkness, Janet A., Fons J. R. Van De Vijver and Peter Ph. Mohler. Eds. 2003. Cross-cultural survey methods . Hoboken: Wiley.

Henrich, Joseph, Steven J. Heine and Ara Norenzayan. 2010. The weirdest people in the world? Behavioral and Brain Sciences 33:61–135.

Hoffmeyer-Zlotnik, Jürgen H. P., and Christof Wolf. 2011. Advances in cross-national comparison: A European working book for demographic and socio-economic variables . New York: Springer.

Hosoya, Georg, Tobias Koch and Michael Eid. 2014. Längsschnittdaten und Mehrebenenanalyse (Longitudinal data and multi-level analyses). In Soziale Kontexte und soziale Mechanismen (Kölner Zeitschrift für Soziologie Sonderheft 54/2014) , eds. Jürgen Friedrichs and Alexandra Nonnenmacher, 189–219. Wiesbaden: Springer.

Inkeles, Alex. 1960. Industrial man: The relation of status to experience, perception, and value. American Journal of Sociology 66:1–31.

ISSP. 2019. International Social Survey Programme. https://www.issp.org (Date of access: 29 Jan. 2019).

Johnson, Timothy P., Beth-Ellen Pennell, Ineke A. L. Stoop and Brita Dorer. 2018. Advances in comparative survey methods: Multinational, multiregional, and multicultural contexts (3MC) . New York: Wiley.

Just, Aida, and Christopher J. Anderson. 2012. Immigrants, Citizenship and Political Action in Europe. British Journal of Political Science 42:481–509. https://doi.org/10.1017/S0007123411000378 .

Kaltenthaler, Karl C., and Christopher J. Anderson. 2001. Europeans and their money: Explaining public support for the common European currency. European Journal of Political Research 40:139–170.

Kohn, Melvin L. 1987. Cross-national research as an analytic strategy: American Sociological Association, 1987 Presidential Address. American Sociological Review 52:713–731.

Kohn, Melvin L. 2015. Change and stability: A cross-national analysis of social structure and personality . New York: Taylor & Francis.

Kroneberg, Clemens. 2019. Theory development in comparative social research. In Cross-national comparative research – analytical strategies, results and explanations . Sonderheft Kölner Zeitschrift für Soziologie und Sozialpsychologie . Eds. Hans-Jürgen Andreß, Detlef Fetchenhauer and Heiner Meulemann. Wiesbaden: Springer VS. https://doi.org/10.1007/s11577-019-00604-y .

Landman, Todd. 2017. Issues and methods in comparative politics. An introduction. 4th edition . London and New York: Routledge.

Lazarsfeld, Paul, and Herbert Menzel. 1969. On the relation between individual and collective properties. In A sociological reader on complex organization. 2nd edition, ed. Amitai Etzioni, 422–440. New York: Holt, Rinehart and Winston.

Lewis, Jeffrey B., and Drew A. Linzer. 2005. Estimating regression models in which the dependent variable is based on estimates. Political Analysis 13:345–364.

Lipset, Seymour Martin. 1959. Democracy and working-class authoritarianism. American Sociological Review 24:482–501.

March, James G., and Johan P. Olsen. 1989. Rediscovering institutions . New York: Free Press.

Meuleman, Bart. 2019. Multilevel Structural Equation Modeling for Cross-National Comparative Research. In Cross-national comparative research – analytical strategies, results and explanations . Sonderheft Kölner Zeitschrift für Soziologie und Sozialpsychologie . Eds. Hans-Jürgen Andreß, Detlef Fetchenhauer and Heiner Meulemann. Wiesbaden: Springer VS. https://doi.org/10.1007/s11577-019-00605-x .

Meuleman, Bart, Eldad Davidov and Jaak Billiet. 2018. Modeling multiple-country repeated cross-sections. A societal growth curve model for studying the effect of the economic crisis on perceived ethnic threat. Methoden, Daten, Analysen 12:185–210.

Meulemann, Heiner. 2002. Ed. Social capital in Europe: Similarity of countries and diversity of people? Multi-level analyses of the European Social Survey 2002 . Leiden: Brill.

Meulemann, Heiner. 2013. Soziologie von Anfang an. Eine Einführung in Themen, Ergebnisse und Literatur. 3rd ed. Wiesbaden: Springer.

Meulemann, Heiner. 2017. The secularization theory—not disconfirmed, yet rarely tested. Analyse und Kritik 39:325–355.

Meyer, John W., John Boli, George M. Thomas and Francisco O. Ramirez. 1997. World society and the nation-state. American Journal of Sociology 103:144–181.

Minkov, Michael. 2013. Cross-cultural analysis: The science and art of comparing the world’s modern societies and their cultures . Los Angeles: Sage.

Nonnenmacher, Alexandra, and Jürgen Friedrichs. 2013. The missing link: Deficits of country level studies. A review of 22 articles explaining life satisfaction. Social Indicators Research 110(3):1221–1244.

Norris, Pippa, and Ronald Inglehart. 2004. Sacred and secular. Religion and politics worldwide . Cambridge: Cambridge University Press.

Opp, Karl-Dieter. 2009. Theories of political protest and social movements. A multidisciplinary introduction, critique and synthesis . London: Routledge

Parsons, Talcott. 1964. Evolutionary universals in society. American Sociological Review 29:339–357.

Pescosolido, Bernice. 2007. Sociology of social networks. In 21st century sociology: A reference handbook, Volume 1 , eds. Clifton D. Bryant and Dennis L. Peck, 208–217. Thousand Oaks: Sage.

Chapter   Google Scholar  

Pförtner, Timo-Kolja, Holger Pfaff and Frank J. Elgar. 2019. The role of welfare state characteristics for health and inequalities in health from a cross-national perspective: A critical research synthesis. In Cross-national comparative research – analytical strategies, results and explanations . Sonderheft Kölner Zeitschrift für Soziologie und Sozialpsychologie . Eds. Hans-Jürgen Andreß, Detlef Fetchenhauer and Heiner Meulemann. Wiesbaden: Springer VS. https://doi.org/10.1007/s11577-019-00606-w .

Przeworski, Adam, and Henri Teune. 1970. The logic of comparative social inquiry . New York: Wiley-Interscience.

Robinson, William S. 1950. Ecological correlations and the behavior of individuals. American Sociological Review 15:351–357.

Saxonhouse, Gary R. 1976. Estimated parameters as dependent variables. The American Economic Review 66:178–183.

Saxonhouse, Gary R. 1977. Regressions from samples having different characteristics. The Review of Economics and Statistics 59:234–237.

Schmidt-Catran, Alexander W. and Malcolm Fairbrother. 2016. The random effects in multilevel models: Getting them wrong and getting them right. European Sociological Review 32:23–38.

Schmidt-Catran, Alexander W., Malcolm Fairbrother and Hans-Jürgen Andreß. 2019. Multilevel models for the analysis of comparative survey data: Common problems and some solutions. In Cross-national comparative research – analytical strategies, results and explanations . Sonderheft Kölner Zeitschrift für Soziologie und Sozialspychologie . Eds. Hans-Jürgen Andreß, Detlef Fetchenhauer and Heiner Meulemann. Wiesbaden: Springer VS. https://doi.org/10.1007/s11577-019-00607-9 .

Schmitt-Beck, Rüdiger. 2019. Political systems and electoral behavior: A review of internationally comparative multilevel research. In Cross-national comparative research – analytical strategies, results and explanations . Sonderheft Kölner Zeitschrift für Soziologie und Sozialpsychologie . Eds. Hans-Jürgen Andreß, Detlef Fetchenhauer and Heiner Meulemann. Wiesbaden: Springer VS. https://doi.org/10.1007/s11577-019-00608-8 .

Schröder, Martin. 2019. Varieties of capitalism and welfare regime theories: Assumptions, accomplishments, and the need for different Methods. In Cross-national comparative research – analytical strategies, results and explanations . Sonderheft Kölner Zeitschrift für Soziologie und Sozialpsychologie . Eds. Hans-Jürgen Andreß, Detlef Fetchenhauer and Heiner Meulemann. Wiesbaden: Springer VS. https://doi.org/10.1007/s11577-019-00609-7 .

SHARE. 2019. Survey of Health, Ageing and Retirement in Europe. https://www.share-project.org (Date of access: 29 Jan. 2019).

Siegers, Pascal. 2019. Is the influence of religiosity on attitudes and behaviors stronger in less religious or more religious societies? A review of theories and contradictory evidence. In Cross-national comparative research – analytical strategies, results and explanations . Sonderheft Kölner Zeitschrift für Soziologie und Sozialpsychologie . Eds. Hans-Jürgen Andreß, Detlef Fetchenhauer and Heiner Meulemann. Wiesbaden: Springer VS. https://doi.org/10.1007/s11577-019-00610-0 .

Spies, Dennis Christopher, and Simon T. Franzmann. 2019. Party competition and vote choice. In Cross-national comparative research – analytical strategies, results and explanations . Sonderheft Kölner Zeitschrift für Soziologie und Sozialpsychologie . Eds. Hans-Jürgen Andreß, Detlef Fetchenhauer and Heiner Meulemann. Wiesbaden: Springer VS. https://doi.org/10.1007/s11577-019-00611-z .

Stavrova, Olga. 2015. Religion, self-rated health, and mortality: Whether religiosity delays death depends on the cultural context. Social Psychological and Personality Science 6:911–922.

Stavrova, Olga. 2019. How much do sources of happiness vary across countries? A review of the empirical literature. In Cross-national comparative research – analytical strategies, results and explanations . Sonderheft Kölner Zeitschrift für Soziologie und Sozialpsychologie . Eds. Hans-Jürgen Andreß, Detlef Fetchenhauer and Heiner Meulemann. Wiesbaden: Springer VS. https://doi.org/10.1007/s11577-019-00612-y .

Swamy, Paravastu A. V. B. 1970. Efficient inference in a random coefficient regression model. Econometrica 38:311–323.

The Local. 2015. Danish police: Refugees can travel on to Sweden. TT/AFP/The Local, September 10, 2015. https://www.thelocal.se/20150910/danish-police-allow-refugees-to-continue-on-to-sweden (Date of access: 15 Jan. 2019).

Treiman, Donald J. 1977. Occupational prestige in comparative perspective . New York: Academic Press.

Wallerstein, Immanuel Maurice. 2011a. The modern world-system III: The second era of great expansion of the capitalist world-economy, 1730s–1840s . Berkeley and Los Angeles: University of California Press.

Wallerstein, Immanuel Maurice. 2011b. The modern world-system IV: Centrist liberalism triumphant, 1789–1914 . Berkeley and Los Angeles: University of California Press.

Wallerstein, Immanuel Maurice. 2011c. The modern world-system II: Mercantilism and the consolidation of the European world-economy, 1600–1750 . Berkeley: University of California Press.

Wallerstein, Immanuel Maurice. 2011d. The modern world-system I: Capitalist agriculture and the origins of the European world-economy in the sixteenth century. Berkeley and Los Angeles: University of California Press.

Weber, Max. 1980. Soziologische Grundbegriffe (Basic sociological terms). In Max Weber Schriften 1894–1922 , ed. Dirk Käsler, 653–716. Stuttgart: Kröner (originally published 1920).

Welch, Michael R., Charles R. Tittle and Thomas Petee. 1991. Religion and deviance among adult Catholics: A test of the “moral communities” hypothesis. Journal for the Scientific Study of Religion 30:159–172.

Wilkinson, Richard G. 2006. The impact of inequality: How to make sick societies healthier . New York: New Press.

Zapf, Wolfgang. 1994. Modernisierung, Wohlfahrtsentwicklung und Transformation. Soziologische Aufsätze 1987–1994 . Berlin: Edition sigma.

Ziller, Conrad. 2019. Policy Effects on Political Engagement. In Cross-national comparative research – analytical strategies, results and explanations . Sonderheft Kölner Zeitschrift für Soziologie und Sozialpsychologie . Eds. Hans-Jürgen Andreß, Detlef Fetchenhauer and Heiner Meulemann. Wiesbaden: Springer VS. https://doi.org/10.1007/s11577-019-00613-x .

Download references

Acknowledgements

We would like to thank Romana Careja, Clemens Kroneberg, and Conrad Ziller for their helpful comments on earlier versions of this article. Moreover, as editors of this special issue, we acknowledge that this project would not have been possible without the help of colleagues and funding organizations. First of all, we would like to thank the editors and the editorial team of the Kölner Zeitschrift für Soziologie und Soziologie for discussing and finally accepting our proposal for a special issue on cross-national comparative research, and for their input and practical support in also finalizing it. Second, we thank our authors for their willingness to follow our guidelines for the publication project and for their patience with our numerous revision requests. Third, the contributions to this special issue greatly benefitted from the reviews and discussions during an authors’ conference held in Cologne 2017. Our thanks go to Rolf Becker, Gerhard Bosch, Miriam Bröckel, Hilke Brockmann, Marius Busemeyer, Christian Czymara, Claudia Diehl, Nico Dragano, Malcolm Fairbrother, Jürgen Friedrichs, Catherine Hakim, Loek Halman, Johannes Huinink, Staffan Kumlin, Steffen Lehndorff, Bart Meuleman, Karl-Dieter Opp, Gert Pickel, Ingo Rohlfing, Stefano Ronchi, Sigrid Roßteutscher, Markus Wagner, and Michael Wagner for their valuable input. Fourth and finally, several people helped in organizing and putting the whole project into practice. Ravena Penning together with Lukas Hofheinz organized the conference. She also made sure that all the contributions complied with the KZfSS guidelines, while Neil Mussett did the final English editing.

The authors’ conference on which this special issue is based was partly financed by a grant from the Thyssen Foundation, for which we are highly grateful. All other costs were covered by a grant to the University of Cologne from the German Research Foundation, which supported the Research Training Group “Social Order and Life Chances in Cross-National Comparison (SOCLIFE)” between 2008 and 2017, for which we are also highly grateful. Finally, the idea for this special issue would not have been born without the enthusiasm and academic success of our SOCLIFE students, who have been inspiring us with their PhD projects for almost a decade, this having been—for the three of us—a form of coda to our common ten-year endeavors of teaching—and researching—in SOCLIFE.

Author information

Authors and affiliations.

Institut für Soziologie und Sozialpsychologie, Universität zu Köln, Universitätsstr. 24, 50931, Cologne, Germany

Hans-Jürgen Andreß, Detlef Fetchenhauer & Heiner Meulemann

You can also search for this author in PubMed   Google Scholar

Corresponding author

Correspondence to Hans-Jürgen Andreß .

Rights and permissions

Reprints and permissions

About this article

Andreß, HJ., Fetchenhauer, D. & Meulemann, H. Cross-National Comparative Research—Analytical Strategies, Results, and Explanations. Köln Z Soziol 71 (Suppl 1), 1–28 (2019). https://doi.org/10.1007/s11577-019-00594-x

Download citation

Published : 22 May 2019

Issue Date : 03 June 2019

DOI : https://doi.org/10.1007/s11577-019-00594-x

Share this article

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

• Multilevel analysis
• Mixed effects models
• Cross-sectional and longitudinal designs
• Context effects

Schlüsselwörter

• Mehrebenenanalyse
• Mixed-Effects-Modelle
• Querschnitte und Längsschnitte
• Kontexteffekte
• Find a journal
• Publish with us
• Track your research

• Search Menu
• Advance articles
• Author Guidelines
• Submission Site
• Open Access
• Why Publish?
• About Research Evaluation
• Editorial Board
• Advertising and Corporate Services
• Journals Career Network
• Self-Archiving Policy
• Dispatch Dates
• Journals on Oxford Academic
• Books on Oxford Academic

Julia Melkers

Emanuela Reale

Thed van Leeuwen

About the journal

Research Evaluation is an interdisciplinary peer-reviewed, international journal. Its subject matter is the evaluation of activities concerned with scientific research, technological development and innovation …

Why submit to Research Evaluation ?

Learn more about the submission process, our author benefits, and how you can join our prestigious author community.

Find out more

High Impact Articles

Keep up to date with the latest research your peers are reading and citing, with a collection of the most cited articles published in  Research Evaluation  in recent years.

The JHPS editorial board

Learn more about JHPS’ s international editorial board.

Measuring the Impact of Arts and Humanities Research in Europe

Read the latest virtual issue exploring the societal value of arts and humanities research.

Latest articles

Email alerts

Register to receive table of contents email alerts as soon as new issues of Research Evaluation are published online.

Recommend to your library

Fill out our simple online form to recommend Research Evaluation to your library.

Recommend now

Committee on Publication Ethics (COPE)

This journal is a member of and subscribes to the principles of the Committee on Publication Ethics (COPE)

publicationethics.org

Related Titles

• Recommend to your Library

Affiliations

• Online ISSN 1471-5449
• Print ISSN 0958-2029
• Copyright © 2024 Oxford University Press
• About Oxford Academic
• Publish journals with us
• University press partners
• What we publish
• New features  
• Open access
• Institutional account management
• Rights and permissions
• Get help with access
• Accessibility
• Advertising
• Media enquiries
• Oxford University Press
• Oxford Languages
• University of Oxford

Oxford University Press is a department of the University of Oxford. It furthers the University's objective of excellence in research, scholarship, and education by publishing worldwide

• Copyright © 2024 Oxford University Press
• Cookie settings
• Cookie policy
• Privacy policy
• Legal notice

This Feature Is Available To Subscribers Only

Sign In or Create an Account

This PDF is available to Subscribers Only

For full access to this pdf, sign in to an existing account, or purchase an annual subscription.

Sen. Mark Warner convenes biotech leaders in Roanoke and New River valleys

Key university and industry players gathered at the Fralin Biomedical Research Institute at VTC to discuss research, economic development, and a vision for Virginia’s proposed research triangle.

• Matt Chittum

28 Mar 2024

• Share on Facebook
• Share on Twitter
• Copy address link to clipboard

A rare glimpse of the biotechnology ecosystem in the region occurred this week when U.S. Sen. Mark Warner convened many of its tech sector leaders at the Fralin Biomedical Research Institute at VTC .

The Research, Innovation, and Economic Development in Biotech event highlighted progress and potential  as well as Virginia’s proposed research triangle, which primarily includes Virginia Tech and the research institute, the University of Virginia, and Virginia Commonwealth University.

“To have Sen. Warner here, and this group of stakeholders together, is a special day for us to be able to share our common goals and interests,” said Michael Friedlander , executive director of the research institute and Virginia Tech’s vice president for health sciences and technology. “With such a well-informed, influential, and supportive person as a Sen. Warner sharing a dialogue with us, I think great things are going to come out of this meeting.”

Virginia Tech President Tim Sands welcomed Warner and other officials at the event.

“It was a wonderful opportunity to get all of our partners together across these sectors that all support the biotechnology enterprise in the region. And to have Sen. Warner convene us was a very nice twist in the sense that he has a lot of insights at the federal level and certainly in Virginia,” Sands said. “Having all these people who represent different parts of the ecosystem on hand to hear the same messages really knits everyone together. It was fantastic.”

As the chairman of the U.S. Senate’s intelligence committee, Warner sees the kind of research going on at Virginia Tech with partners at Carilion Clinic and the Children’s National Research & Innovation Campus in Washington, D.C., as potentially fitting into the national security envelope.

“National security is no longer simply who has the most tanks and guns and ships and plans, it’s who’s going to dominate in a variety of technology domains [such as] synthetic biology, bio-manufacturing, and the next generation of biotech,” Warner said. “In so many ways, what is happening at Fralin Biomedical Research Institute and at the medical school is really cutting edge. The next generation of bio is something I’d love to be an advocate for.”

“What we’re building together is attracting the brightest minds in health care, technology, and research who are putting their ideas to work here,” said Carilion Clinic CEO Nancy Agee. “It’s exciting to think about the cures, therapies, and medications of tomorrow that will be developed here and benefit patients everywhere.”

The roundtable included more than two dozen biotech, health care, economic, and academic leaders from the region and the state, including Joe Benevento, Virginia’s deputy secretary of commerce and trade and interim CEO of the Virginia Innovation Partnership Corp., and Heywood Fralin, Roanoke businessman, economic development proponent and the namesake benefactor of the Fralin Biomedical Research Institute. Fralin noted that the broad region of the Roanoke and New River valleys and the area around Martinsville and Danville to the south was the fastest growing economically in the last year, according to Virginia data.

Friedlander cited the research institute’s growth from its opening 14 years ago to more than 40 faculty-led research teams this year. The research institute, with over 500 employees and students, holds more than $220 million in current active outside funding. Its funding per faculty member rivals some of the country’s larger and more established research centers. He also noted the globally rich talent pool that the research institute draws upon is a major contributor to its success and that is at least party attributable to the welcoming environment provided by the Roanoke community, consistent with Warner’s assessment of the impact on our nation’s global competitiveness.

A panel discussion featured Erin Burcham, president of Verge Alliance and executive director of the Roanoke-Blacksburg Technology Council; Tony Seupaul, executive vice president and chief physician executive at Carilion Clinic; Sarah Snider, CEO and co-founder of BEAM Diagnostics, a company founded from research at Fralin Biomedical Research Institute that develops technology to support behavioral health care; and Rob Gourdie, a research institute professor and founder of the Tiny Cargo Co., which is developing a therapeutic drug delivery system.

Moderated by Virginia Tech Associate Vice President for Innovation and Partnerships  Brandy Salmon , panelists extolled the support from Virginia Tech, the research institute, state and local governments, and RAMP, Roanoke’s start-up business accelerator program. Multiple attendees also noted that Roanoke’s natural beauty, outdoor amenities, and vibrant arts scene are valuable assets for drawing the people needed to grow the biotech sector.

FROM LEFT: A panel featuring Sarah Snider, CEO and co-founder of BEAM Diagnostics, a company founded from research at Fralin Biomedical Research Institute that develops technology to support behavioral health care; Rob Gourdie, a research institute professor and founder of the Tiny Cargo Co., which is developing a therapeutic drug delivery system; Erin Burcham, president of Verge Alliance and executive director of the Roanoke-Blacksburg Technology Council; and Tony Seapaul, executive vice president and chief physician executive at Carilion Clinic; discussed ways to foster innovation in the region. Photos by Ryan Anderson for Virginia Tech.

Michael Friedlander, executive director of the Fralin Biomedical Research Institute at VTC and Virginia Tech vice president for health sciences and technology, provided an overview of growth that began with a single scientist at the research institute in 2010 to nearly 40 Virginia Tech faculty-led research teams today.

From left: Fralin Biomedical Research Institute Executive Director Michael Friedlander, Virginia Tech President Tim Sands, and Carilion Clinic CEO Nancy Howell Agee were on hand to support research initiatives.

Leigh Anne Kelley

540-526-2002

John Pastor

(540) 526-2222

• Center for Vascular and Heart Research
• Children's National Hospital
• Children's National Research & Innovation Campus
• Faculty of Health Sciences
• Fralin Biomedical Research Institute at VTC
• Fralin Biomedical Research Institute at VTC - top news
• Friedlander News
• Gourdie Lab
• Industry Partners
• Industry Partnerships
• LINK+LICENSE+LAUNCH
• Office of the President
• Roanoke, Va.

Related Content

An official website of the United States government

The .gov means it’s official. Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

The site is secure. The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

• Publications
• Account settings

Preview improvements coming to the PMC website in October 2024. Learn More or Try it out now .

• Advanced Search
• Journal List

Benefits and harms of medical cannabis: a scoping review of systematic reviews

Misty pratt.

1 Knowledge Synthesis Group, Ottawa Methods Centre, Ottawa Hospital Research Institute, The Ottawa Hospital, General Campus, 501 Smyth Road, Ottawa, Ontario K1H 8 L6 Canada

Adrienne Stevens

2 TRIBE Graduate Program, University of Split School of Medicine, Split, Croatia

Micere Thuku

Claire butler.

3 Department of Pharmacology and Therapeutics, McGill University, Montreal, Quebec H3A 2B4 Canada

Becky Skidmore

4 Ottawa, Canada

L. Susan Wieland

5 Center for Integrative Medicine, University of Maryland School of Medicine, Baltimore, MD USA

Mark Clemons

6 School of Epidemiology and Public Health, University of Ottawa, 451 Smyth Road, Ottawa, Ontario K1H 8 M5 Canada

7 Division of Medical Oncology and Department of Medicine, University of Ottawa, Ottawa, Canada

Salmaan Kanji

8 Department of Pharmacy, The Ottawa Hospital, Ottawa, Canada

9 Clinical Epidemiology Program, The Ottawa Hospital Research Institute, Ottawa, Canada

Brian Hutton

Associated data.

All data generated or analyzed during this study are included in this published article (and its supplementary information files).

There has been increased interest in the role of cannabis for treating medical conditions. The availability of different cannabis-based products can make the side effects of exposure unpredictable. We sought to conduct a scoping review of systematic reviews assessing benefits and harms of cannabis-based medicines for any condition.

A protocol was followed throughout the conduct of this scoping review. A protocol-guided scoping review conduct. Searches of bibliographic databases (e.g., MEDLINE®, Embase, PsycINFO, the Cochrane Library) and gray literature were performed. Two people selected and charted data from systematic reviews. Categorizations emerged during data synthesis. The reporting of results from systematic reviews was performed at a high level appropriate for a scoping review.

After screening 1975 citations, 72 systematic reviews were included. The reviews covered many conditions, the most common being pain management. Several reviews focused on management of pain as a symptom of conditions such as multiple sclerosis (MS), injury, and cancer. After pain, the most common symptoms treated were spasticity in MS, movement disturbances, nausea/vomiting, and mental health symptoms. An assessment of review findings lends to the understanding that, although in a small number of reviews results showed a benefit for reducing pain, the analysis approach and reporting in other reviews was sub-optimal, making it difficult to know how consistent findings are when considering pain in general. Adverse effects were reported in most reviews comparing cannabis with placebo (49/59, 83%) and in 20/24 (83%) of the reviews comparing cannabis to active drugs. Minor adverse effects (e.g., drowsiness, dizziness) were common and reported in over half of the reviews. Serious harms were not as common, but were reported in 21/59 (36%) reviews that reported on adverse effects. Overall, safety data was generally reported study-by-study, with few reviews synthesizing data. Only one review was rated as high quality, while the remaining were rated of moderate ( n = 36) or low/critically low ( n = 35) quality.

Conclusions

Results from the included reviews were mixed, with most reporting an inability to draw conclusions due to inconsistent findings and a lack of rigorous evidence. Mild harms were frequently reported, and it is possible the harms of cannabis-based medicines may outweigh benefits.

Systematic review registration

The protocol for this scoping review was posted in the Open Access ( https://ruor.uottawa.ca/handle/10393/37247 ).

Interest in medical applications of marijuana ( Cannabis sativa ) has increased dramatically during the past 20 years. A 1999 report from the National Academies of Sciences, Engineering, and Medicine supported the use of marijuana in medicine, leading to a number of regulatory medical colleges providing recommendations for its prescription to patients [ 1 ]. An updated report in 2017 called for a national research agenda, improvement of research quality, improvement in data collection and surveillance efforts, and strategies for addressing barriers in advancing the cannabis agenda [ 2 ].

Proponents of medical cannabis support its use for a highly varied range of medical conditions, most notably in the fields of pain management [ 3 ] and multiple sclerosis [ 4 ]. Marijuana can be consumed by patients in a variety of ways including smoking, vaporizing, ingesting, or administering sublingually or rectally. The plant consists of more than 100 known cannabinoids, the main ones of relevance to medical applications being tetrahydrocannabinol (THC) and cannabidiol (CBD) [ 5 ]. Synthetic forms of marijuana such as dronabinol and nabilone are also available as prescriptions in the USA and Canada [ 6 ].

Over the last decade, there has been an increased interest in the use of medical cannabis products in North America. It is estimated that over 3.5 million people in the USA are legally using medical marijuana, and a total of USD$6.7 billion was spent in North America on legal marijuana in 2016 [ 7 ]. The number of Canadian residents with prescriptions to purchase medical marijuana from Health Canada–approved growers tripled from 30,537 in 2015 to near 100,000 in 2016 [ 8 ]. With the legalization of recreational-use marijuana in parts of the USA and in Canada in October 2018, the number of patients using marijuana for therapeutic purposes may become more difficult to track. The likely increase in the numbers of individuals consuming cannabis also necessitates a greater awareness of its potential benefits and harms.

Plant-based and plant-derived cannabis products are not monitored as more traditional medicines are, thereby increasing the uncertainty regarding its potential health risks to patients [ 3 ]. While synthetic forms of cannabis are available by prescription, different cannabis plants and products contain varied concentrations of THC and CBD, making the effects of exposure unpredictable [ 9 ]. While short-lasting side effects including drowsiness, loss of short-term memory, and dizziness are relatively well known and may be considered minor, other possible effects (e.g., psychosis, paranoia, anxiety, infection, withdrawal) may be more harmful to patients.

There remains a considerable degree of clinical equipoise as to the benefits and harms of marijuana use for medical purposes [ 10 – 13 ]. To understand the extent of synthesized evidence underlying this issue, we conducted a scoping review [ 14 ] of systematic reviews evaluating the benefits and/or harms of cannabis (plant-based, plant-derived, and synthetic forms) for any medical condition. We located and mapped systematic reviews to summarize research that is available for consideration for practice or policy questions in relation to medical marijuana.

A scoping review protocol was prepared and posted to the University of Ottawa Health Sciences Library’s online repository ( https://ruor.uottawa.ca/handle/10393/37247 ). We used the PRISMA for Scoping Reviews checklist to guide the reporting of this report (see Additional file 1 ) [ 15 ].

Literature search and process of study selection

An experienced medical information specialist developed and tested the search strategy using an iterative process in consultation with the review team. Another senior information specialist peer-reviewed the strategy prior to execution using the PRESS Checklist [ 16 ]. We searched seven Ovid databases: MEDLINE®, including Epub Ahead of Print and In-Process & Other Non-Indexed Citations, Embase, Allied and Complementary Medicine Database, PsycINFO, the Cochrane Database of Systematic Reviews, the Database of Abstracts of Reviews of Effects, and the Health Technology Assessment Database. The final peer-reviewed search strategy for MEDLINE was translated to the other databases (see Additional file 2 ). We performed the searches on November 3, 2017.

The search strategy incorporated controlled vocabulary (e.g., “Cannabis,” “Cannabinoids,” “Medical Marijuana”) and keywords (e.g., “marijuana,” “hashish,” “tetrahydrocannabinol”) and applied a broad systematic review filter where applicable. Vocabulary and syntax were adjusted across the databases and where possible animal-only and opinion pieces were removed, from the search results.

Gray literature searching was limited to relevant drug and mental health databases, as well as HTA (Health Technology Assessment) and systematic review databases. Searching was guided by the Canadian Agency for Drugs and Technologies in Health’s (CADTH) checklist for health-related gray literature (see Additional file 3 ). We performed searches between January and February 2018. Reference lists of overviews were searched for relevant systematic reviews, and we searched for full-text publications of abstracts or protocols.

Management of all screening was performed using Distiller SR Software ® (Evidence Partners Inc., Ottawa, Canada). Citations from the literature search were collated and de-duplicated in Reference Manager (Thomson Reuters: Reference Manager 12 [Computer Program]. New York: Thomson Reuters 2011), and then uploaded to Distiller. The review team used Distiller for Levels 1 (titles and abstracts) and 2 (full-text) screening. Pilot testing of screening questions for both levels were completed prior to implementation. All titles and abstracts were screened in duplicate by two independent reviewers (MT and MP) using the liberal accelerated method [ 17 ]. This method requires only one reviewer to assess an abstract as eligible for full-text screening, and requires two reviewers to deem the abstract irrelevant. Two independent reviewers (MT and MP) assessed full-text reports for eligibility. Disagreements during full-text screening were resolved through consensus, or by a third team member (AS). The process of review selection was summarized using a PRISMA flow diagram (Fig. ​ (Fig.1) 1 ) [ 18 ].

PRISMA-style flow diagram of the review selection process

Review selection criteria

English-language systematic reviews were included if they reported that they investigated harms and/or benefits of medical or therapeutic use of cannabis for adults and children for any indication. Definitions related to medical cannabis/marijuana are provided in Table ​ Table1. 1 . We also included synthetic cannabis products, which are prescribed medicines with specified doses of THC and CBD. Reviews of solely observational designs were included only in relation to adverse effects data, in order to focus on the most robust evidence available. We considered studies to be systematic reviews if at least one database was searched with search dates reported, at least one eligibility criterion was reported, the authors had assessed the quality of included studies, and there was a narrative or quantitative synthesis of the evidence. Reviews assessing multiple interventions (both pharmacological and complementary and alternative medicine (CAM) interventions) were included if the data for marijuana studies was reported separately. Published and unpublished guidelines were included if they conducted a systematic review encompassing the criteria listed above.

Context for the use of cannabis-related terms during the review selection process

We excluded overviews of systematic reviews, reviews in abstract form only, and review protocols. We further excluded systematic reviews focusing on recreational, accidental, acute, or general cannabis use/abuse and interventions such as synthetic cannabinoids not approved for therapeutic use (e.g., K2 or Spice).

Data collection and quality assessment

All data were collected electronically in a pre-developed form using Microsoft Excel software (Microsoft Corporation, Seattle, USA). The form was pilot tested on three included reviews by three people. One reviewer (MP or CB) independently extracted all data, and a second reviewer (MT) verified all of the items collected and checked for any omitted data. Disagreements were resolved by consensus and consultation with a third reviewer if necessary. A data extraction form with the list of included variables is provided in Additional file 4 . All collected data has also been made available in the online supplemental materials associated with this report.

Quality assessment of systematic reviews was performed using the AMSTAR-2 [ 20 ] tool. One reviewer (MP or CB) independently assessed quality, while a second reviewer (MT) verified the assessments. Disagreements were resolved by consensus and consultation with a third reviewer if necessary. The tool consists of 16 items in total, with four critical domains and 12 non-critical domains. The AMSTAR-2 tool is not intended to generate an overall score, and instead allows for an overall rating based on weaknesses in critical domains. Reviews were rated as high (no critical flaws with zero or one non-critical flaw), moderate (no critical flaws with ≥ 1 non-critical flaw), low (one critical flaw with/without non-critical weakness), or critically low (> 1 critical flaw with/without non-critical weakness) quality.

Evidence synthesis

We used a directed content analytic approach [ 21 ] with an initial deductive framework [ 22 ] that allowed flexibility for inductive analysis if refinement or development of new categorization was needed. The framework used to categorize outcome data results is outlined in Table ​ Table2. 2 . Where reviews had a mix of narrative and quantitative data, results from meta-analyses were prioritized over count data or study-by-study data. The extraction and reporting of data results was performed at a high level and did not involve an in-depth evaluation, which is appropriate for a scoping review [ 14 ]. Review authors’ conclusions and/or recommendations were extracted and reported narratively.

Outcome result categorization

Changes from the study protocol

For feasibility, we decided to limit the inclusion of systematic reviews of only observational study designs to those that addressed adverse events data. All other steps of the review were performed as planned.

Search findings

The PRISMA flow diagram describing the process of review selection is presented in Fig. ​ Fig.1. 1 . After duplicates were removed, the search identified a total of 1925 titles and abstracts, of which 47 references were located through the gray literature search. Of the total 1925 citations assessed during Level 1 screening, 1285 were deemed irrelevant. We reviewed full-text reports for the 640 reviews of potential relevance, and of these, 567 were subsequently excluded, leaving a total of 72 systematic reviews that were included; the associated data collected are provided in Additional file 5 . A listing of the reports excluded during full-text review is provided in Additional file 6 .

Characteristics of included reviews

There were 63 systematic reviews [ 4 , 19 , 23 – 83 ] and nine guidelines with systematic reviews [ 84 – 92 ]. Overall, 27 reviews were performed by researchers in Europe, 16 in the USA, 15 in Canada, eight in Australia, two in Brazil, and one each in Israel, Singapore, South Africa, and China. Funding was not reported in 29 (40%) of the reviews, and the remaining reviews received funding from non-profit or academic ( n = 20; 28%), government ( n = 14; 19%), industry ( n = 3; 4%), and mixed ( n = 1; 1%) sources. Five reviews reported that they did not receive any funding for the systematic review. Tables ​ Tables3, 3 , ​ ,4, 4 , ​ ,5, 5 , ​ ,6, 6 , ​ ,7, 7 , ​ ,8, 8 , ​ ,9, 9 , ​ ,10, 10 , ​ ,11, 11 , ​ ,12, 12 , and ​ and13 13 provide an overview of the characteristics of the 72 included systematic reviews.

Multiple sclerosis

MS multiple sclerosis, NICE National Institute for Health and Care Excellence, No . number, NR not reported, NRS numerical rating scale, QoL quality of life, RMI Rivermead Mobility Index, SBS study-by-study, VAS visual analog scale

*A colon indicates that there were separate analyses for each comparator

Movement disorders

HD Huntington’s disease, MS multiple sclerosis, NR not reported, PD Parkinson’s disease, SBS study-by-study, SCL-90R Symptoms Checklist-90 Revised, QoL quality of life, STSSS Shapiro Tourette Syndrome Severity Scale, THC tetrahydrocannabinol, TS-CGI Tourette Syndrome Clinical Global Impressions, TSSL Tourette’s Syndrome Symptom List (patient rated), VAS visual analog scale, YGTSS Yale Global Tic Severity Scale

AE : adverse effect, NICE National Institute for Health and Care Excellence, NNT numbers needed to treat, NP neuropathic pain, NR not reported, QoL quality of life, QST quantitative sensory testing, SBS study-by-study, VAS visual analog scale

*A colon indicates that there were separate analyses for each comparator; a “+” sign indicates placebo was combined with another comparator

AE adverse effect, NP neuropathic pain, NR not reported, NRS numerical rating scale, QoL quality of life, THC tetrahydrocannabinol, SIGN Scottish Intercollegiate Guidelines Network, SBS study-by-study

Rheumatic disease

AE adverse event, FM fibromyalgia, NR not reported, NRS numerical rating scale, OA osteoarthritis, RA rheumatoid arthritis, SBS study-by-study

NP neuropathic pain, NR not reported, QoL quality of life, SBS study-by-study

Mental health

PTSD posttraumatic stress disorder, SBS study-by-study

NP neuropathic pain, NR not reported, SBS study-by-study

Neurological conditions

AE adverse effect, ALS amyotrophic lateral sclerosis, CADTH Canadian Agency for Drugs and Technologies in Health, NR not reported

Various conditions

AE adverse effect, AD Alzheimer’s disease, ALS amyotrophic lateral sclerosis, CADTH Canadian Agency for Drugs and Technologies in Health, CGI-C Clinical Global Impression of Change scale, COPD Chronic Obstructive Pulmonary Disease, FIQ fibromyalgia impact questionnaire, FM fibromyalgia, HD Huntington’s disease, IBD inflammatory bowel disease, MS multiple sclerosis, NP neuropathic pain, NR not reported, PD Parkinson’s disease, PTSD posttraumatic stress disorder, RA rheumatoid arthritis, SBS study-by-study, SCI spinal cord injury

Other conditions

CADTH Canadian Agency for Drugs and Technologies in Health, IBS irritable bowel syndrome, NR not reported, QoL quality of life, SBS study-by-study, VAS visual analog scale

The reviews were published between 2000 and 2018 (median year 2014), and almost half (47%) were focused solely on medical cannabis. Four (6%) reviews covered both medical and other cannabis use (recreational and substance abuse), 19 (26%) reported multiple pharmaceutical interventions (cannabis being one), six (8%) reported various CAM interventions (cannabis being one), and nine (13%) were mixed pharmaceutical and CAM interventions (cannabis being one). Multiple databases were searched by almost all of the reviews (97%), with Medline/PubMed or Embase common to all.

Cannabis use

Figure ​ Figure2 2 illustrates the different cannabis-based interventions covered by the included reviews. Plant-based cannabis consists of whole plant products such as marijuana or hashish. Plant-derived cannabinoids are active constituents of the cannabis plant, such as tetrahydrocannabinol (THC), cannabidiol (CBD), or a combination of THC:CBD (also called nabiximols, under the brand name Sativex) [ 3 ]. Synthetic cannabinoids are manufactured rather than extracted from the plant and include drugs such as nabilone and dronabinol.

Review coverage of the various cannabis-based interventions

Twenty-seven reviews included solely interventions from plant-derived cannabinoids, 10 studied solely synthetic cannabinoids, and eight included solely studies on plant-based cannabis products. Twenty-four reviews covered a combination of different types of cannabis, and the remaining three systematic reviews did not report which type of cannabinoid was administered in the included studies.

The systematic reviews covered a wide range of conditions and illnesses, the most notable being pain management. Seventeen reviews looked at specific types of pain including neuropathic [ 31 , 42 , 62 , 69 , 85 , 90 ], chronic [ 26 , 32 , 52 , 58 , 80 ], cancer [ 84 , 87 ], non-cancer [ 41 , 68 ], and acute [ 38 ] types of pain (one review covered all types of pain) [ 65 ]. Twenty-seven reviews (38%) also focused on management of pain as a symptom of conditions such as multiple sclerosis (MS) [ 6 , 23 , 27 , 43 , 46 , 52 , 63 , 85 , 92 ], injury [ 29 , 35 , 36 , 69 ], cancer [ 37 , 43 , 65 , 88 ], inflammatory bowel disease (IBD) [ 28 ], rheumatic disease (RD) [ 49 , 51 , 73 ], diabetes [ 68 – 70 ], and HIV [ 48 , 53 , 67 ]. In Fig. ​ Fig.3, 3 , the types of illnesses addressed by the set of included reviews are graphically represented, with overlap between various conditions and pain. Some systematic reviews covered multiple diseases, and therefore the total number of conditions represented in Fig. ​ Fig.3 3 is greater than the total number of included reviews.

Conditions or symptoms across reviews that were treated with cannabis. IBD inflammatory bowel disease, MS multiple sclerosis, RD rheumatic disease

One review included a pediatric-only population, in the evaluation of marijuana for nausea and vomiting following chemotherapy [ 54 ]. Although trials in both adult and child populations were eligible for thirteen (18%) reviews, only two additional reviews included studies in children; these reviews evaluated cannabis in cancer [ 60 ] and a variety of conditions [ 25 ]. Many of the reviews ( n = 25, 35%) included only adults ≥ 18 years of age. Almost half of the reviews ( n = 33, 46%) did not report a specific population for inclusion.

Cannabis was prescribed for a wide range of medical issues. The indication for cannabis use is illustrated in Fig. ​ Fig.4. 4 . Pain management ( n = 27) was the most common indication for cannabis use. A number of reviews sought to address multiple disease symptoms ( n = 12) or explored a more holistic treatment for the disease itself ( n = 11). After pain, the most common symptoms being treated with cannabis were spasticity in MS, movement disturbances (such as dyskinesia, tics, and spasms), weight or nausea/vomiting, and mental health symptoms.

Indications for cannabis use across included reviews

Figure ​ Figure5 5 summarizes the breadth of outcomes analyzed in the included reviews. The most commonly addressed outcomes were withdrawal due to adverse effects, “other pain,” neuropathic pain, spasticity, and the global impression of the change in clinical status. Many outcomes were reported using a variety of measures across reviews. For example, spasticity was measured both objectively (using the Ashworth scale) and subjectively (using a visual analog scale [VAS] or numerical rating scale [NRS]). Similarily, outcomes for pain included VAS or NRS scales, reduction in pain, pain relief, analgesia, pain intensity, and patient assessment of change in pain.

Quality of the systematic reviews

Quality assessments of the included reviews based upon AMSTAR-2 are detailed in Additional file 7 and Additional file 8 . Only one review was rated as high quality [ 45 ]. All other reviews were deemed to be of moderate ( n = 36) or low/critically low ( n = 35) methodological quality. Assessments for the domains deemed of critical importance for determining quality ratings are described below.

Only 20% of reviews used a comprehensive search strategy; another 47% were given a partial score because they had not searched the reference lists of the included reviews, trial registries, gray literature, and/or the search date was older than 2 years. The remaining reviews did not report a comprehensive search strategy.

Over half of the reviews (51%) used a satisfactory technique for assessing risk of bias (ROB) of the individual included studies, while 35% were partially satisfactory because they had not reported whether allocation sequence was truly random and/or they had not assessed selective reporting. The remaining reviews did not report a satisfactory technique for assessing ROB.

Most reviews (71%) could not be assessed for an appropriate statistical method for combining results in a meta-analysis, as they synthesized study data narratively. Approximately 19% of reviews used an appropriate meta-analytical approach, leaving 10% that used inappropriate methods.

The final critical domain for the AMSTAR-2 determines whether review authors accounted for ROB in individual studies when discussing or interpreting the results of the review. The majority of reviews (83%) did so in some capacity.

Mapping results of included systematic reviews

We mapped reviews according to authors’ comparisons, the conditions or symptoms they were evaluating, and the categorization of the results (see Table ​ Table2). 2 ). In some cases, reviews contributed to more than one comparison (e.g., cannabis versus placebo or active drug). As pain was the most commonly addressed outcome, we mapped this outcome separately from all other endpoints. This information is shown for all reviews and then restricted to reviews of moderate-to-high quality (as determined using the AMSTAR-2 criteria): cannabis versus placebo (Figs. ​ (Figs.6 6 and ​ and7), 7 ), cannabis versus active drugs (Figs. ​ (Figs.8 8 and ​ and9), 9 ), cannabis versus a combination of placebo and active drug (Figs. ​ (Figs.10 10 and ​ and11), 11 ), one cannabis formulation versus other (Figs. ​ (Figs.12 12 and ​ and13), 13 ), and cannabis analyzed against all other comparators (Fig. ​ (Fig.14). 14 ). Details on how to read the figures are provided in the corresponding figure legends. The median number of included studies across reviews was four, and ranged from one to seventy-nine (not shown in figures).

Cannabis vs. placebo. Authors’ presentations of the findings were mapped using the categorization shown in Table ​ Table2. 2 . According to the reviews’ intended scope for the condition being treated, outcomes were mapped into “pain,” “non-pain outcomes,” and “adverse events.” For each condition and outcome pair (i.e., each row in the grid), the number of reviews reporting findings is shown according to the results categorization. For pain, reviews numbered in different categories signal discordant findings across those reviews. For non-pain outcomes, reviews presenting findings in the different categories would signal different results for different outcomes, as well as discordant findings within and across reviews. Adverse events are grouped as a whole and “favors intervention” would be interpreted as a decrease in events with cannabis when compared with the control group. Favors int = favors intervention; Favors Ctrl = favors control; Not stat sig = not statistically significant

Cannabis vs. placebo, high and moderate quality reviews. Authors’ presentations of the findings were mapped using the categorizations shown in Table ​ Table2. 2 . According to the reviews’ intended scope for the condition being treated, outcomes were mapped into “pain,” “non-pain outcomes,” and “adverse events.” For each condition and outcome pair (i.e., each row in the grid), the number of reviews reporting findings is shown according to the results categorization. For pain, reviews numbered in different categories signal discordant findings across those reviews. For non-pain outcomes, reviews presenting findings in the different categories would signal different results for different outcomes, as well as discordant findings within and across reviews. Adverse events are grouped as a whole and “favors intervention” would be interpreted as a decrease in events with cannabis when compared with the control group. Favors int = favors intervention; Favors Ctrl = favors control; Not stat sig = not statistically significant

Cannabis vs. active drugs. Authors’ presentations of the findings were mapped using the categorizations shown in Table ​ Table2. 2 . According to the reviews’ intended scope for the condition being treated, outcomes were mapped into “pain,” “non-pain outcomes,” and “adverse events.” For each condition and outcome pair (i.e., each row in the grid), the number of reviews reporting findings is shown according to the results categorization. For pain, reviews numbered in different categories signal discordant findings across those reviews. For non-pain outcomes, reviews presenting findings in the different categories would signal different results for different outcomes, as well as discordant findings within and across reviews. Adverse events are grouped as a whole and “favors intervention” would be interpreted as a decrease in events with cannabis when compared with the control group. Favors int = favors intervention; Favors Ctrl = favors control; Not stat sig = not statistically significant

Cannabis vs. active drugs, high and moderate quality reviews. Authors’ presentations of the findings were mapped using the categorizations shown in Table ​ Table2. 2 . According to the reviews’ intended scope for the condition being treated, outcomes were mapped into “pain,” “non-pain outcomes,” and “adverse events.” For each condition and outcome pair (i.e., each row in the grid), the number of reviews reporting findings is shown according to the results categorization. For pain, reviews numbered in different categories signal discordant findings across those reviews. For non-pain outcomes, reviews presenting findings in the different categories would signal different results for different outcomes, as well as discordant findings within and across reviews. Adverse events are grouped as a whole and “favors intervention” would be interpreted as a decrease in events with cannabis when compared with the control group. Favors int = favors intervention; Favors Ctrl = favors control; Not stat sig = not statistically significant

Cannabis vs. placebo + active drug. Authors’ presentations of the findings were mapped using the categorizations shown in Table ​ Table2. 2 . According to the reviews’ intended scope for the condition being treated, outcomes were mapped into “pain,” “non-pain outcomes,” and “adverse events.” For each condition and outcome pair (i.e., each row in the grid), the number of reviews reporting findings is shown according to the results categorization. For pain, reviews numbered in different categories signal discordant findings across those reviews. For non-pain outcomes, reviews presenting findings in the different categories would signal different results for different outcomes, as well as discordant findings within and across reviews. Adverse events are grouped as a whole and “favors intervention” would be interpreted as a decrease in events with cannabis when compared with the control group. Favors int = favors intervention; Favors Ctrl = favors control; Not stat sig = not statistically significant

Cannabis vs. placebo + active drug, high and moderate quality reviews. Authors’ presentations of the findings were mapped using the categorizations shown in Table ​ Table2. 2 . According to the reviews’ intended scope for the condition being treated, outcomes were mapped into “pain,” “non-pain outcomes,” and “adverse events.” For each condition and outcome pair (i.e., each row in the grid), the number of reviews reporting findings is shown according to the results categorization. For pain, reviews numbered in different categories signal discordant findings across those reviews. For non-pain outcomes, reviews presenting findings in the different categories would signal different results for different outcomes, as well as discordant findings within and across reviews. Adverse events are grouped as a whole and “favors intervention” would be interpreted as a decrease in events with cannabis when compared with the control group. Favors int = favors intervention; Favors Ctrl = favors control; Not stat sig = not statistically significant

One cannabis formulation vs. other. Authors’ presentations of the findings were mapped using the categorizations shown in Table ​ Table2. 2 . According to the reviews’ intended scope for the condition being treated, outcomes were mapped into “pain,” “non-pain outcomes,” and “adverse events.” For each condition and outcome pair (i.e., each row in the grid), the number of reviews reporting findings is shown according to the results categorization. For pain, reviews numbered in different categories signal discordant findings across those reviews. For non-pain outcomes, reviews presenting findings in the different categories would signal different results for different outcomes, as well as discordant findings within and across reviews. Adverse events are grouped as a whole and “favors intervention” would be interpreted as a decrease in events with cannabis when compared with the control group. Favors int = favors intervention; Favors Ctrl = favors control; Not stat sig = not statistically significant

One cannabis formulation vs. other, high and moderate quality reviews. Authors’ presentations of the findings were mapped using the categorizations shown in Table ​ Table2. 2 . According to the reviews’ intended scope for the condition being treated, outcomes were mapped into “pain,” “non-pain outcomes,” and “adverse events.” For each condition and outcome pair (i.e., each row in the grid), the number of reviews reporting findings is shown according to the results categorization. For pain, reviews numbered in different categories signal discordant findings across those reviews. For non-pain outcomes, reviews presenting findings in the different categories would signal different results for different outcomes, as well as discordant findings within and across reviews. Adverse events are grouped as a whole and “favors intervention” would be interpreted as a decrease in events with cannabis when compared with the control group. Favors int = favors intervention; Favors Ctrl = favors control; Not stat sig = not statistically significant

Cannabis vs. all comparators combined. Authors’ presentations of the findings were mapped using the categorizations shown in Table ​ Table2. 2 . According to the reviews’ intended scope for the condition being treated, outcomes were mapped into “pain,” “non-pain outcomes,” and “adverse events.” For each condition and outcome pair (i.e., each row in the grid), the number of reviews reporting findings is shown according to the results categorization. For pain, reviews numbered in different categories signal discordant findings across those reviews. For non-pain outcomes, reviews presenting findings in the different categories would signal different results for different outcomes, as well as discordant findings within and across reviews. Adverse events are grouped as a whole and “favors intervention” would be interpreted as a decrease in events with cannabis when compared with the control group. Favors int = favors intervention; Favors Ctrl = favors control; Not stat sig = not statistically significant

Cannabis versus placebo

Most reviews (59/72, 82%) compared cannabis with placebo. Of these reviews, 34 (58%) addressed pain outcomes and 47 (80%) addressed non-pain outcomes, with most outcomes addressed by three reviews or fewer (Fig. ​ (Fig.6). 6 ). Some reviews had a mix of quantitative syntheses and study-by-study data reported (13/59, 22%), while another group of reviews (14/59, 24%) only reported results study-by-study. Overall, 24% (14/59) of the cannabis versus placebo reviews had only one included study.

• i. Reviews focused on addressing pain across conditions. In most cases, findings were discordant across reviews for the pain outcomes measured. For chronic non-cancer pain, however, two reviews favored cannabis over placebo for decreasing pain. One review assessing acute pain for postoperative pain relief found no difference between various cannabinoid medications and placebo. The distribution of findings was similar when restricting to moderate-to-high-quality reviews.
• ii. Reviews focused on treating a condition or family of related conditions . Various results were observed for pain. For MS and HIV/AIDS, one review each reported quantitative results favoring cannabis for decreased pain but with other reviews reporting results study-by-study, it is difficult to know, broadly, how consistent those findings are. For cancer, two reviews reported results favoring cannabis for decreased pain. For rheumatic disease, findings are discordant between two reviews, and another two reviews reported results study-by-study. One review that included studies of MS or paraplegia found no difference in pain between groups. For treating injury, one review showed that the placebo group had less pain and one review reported data study-by-study. No reviews addressed pain in movement disorders, neurological conditions, and IBD.

For those reviews assessing pain as part of a focus on treating a range of conditions, two showed cannabis reduced pain [ 43 , 52 ], but one showed mixed results depending on how pain was measured [ 43 ]. These reviews covered several different conditions, including injury, chronic pain, rheumatoid arthritis, osteoarthritis, fibromyalgia, HIV/AIDS, cancer, and MS or paraplegia.

When restricting to moderate-to-high-quality reviews, only one review each in multiple sclerosis and HIV/AIDS with a study-by-study analysis on pain remained. One review on cancer favored cannabis for pain reduction. Findings remained the same for MS or paraplegia and rheumatic disease. No review for injury and paint outcomes was of higher quality.

• 2. Non-pain outcomes

The types of non-pain outcomes included in the reviews varied by condition/illness. The most commonly reported outcomes (see Fig. ​ Fig.5 5 for overall outcomes) when comparing cannabis to placebo included muscle- or movement-related outcomes ( n = 20), quality of life ( n = 14), and sleep outcomes ( n = 10).

There was no consistent pattern for non-pain outcomes either within or across medical conditions. Many ( n = 24, 33%) reviews assessing non-pain outcomes reported the results of those analyses study-by-study. Conflicting results are observed in some cases due to the use of different measures, such as different ways of quantifying spasticity in patients with multiple sclerosis [ 56 , 91 ]. One review each addressing neurological conditions [ 50 ] (outcome: muscle cramps) and MS/paraplegia [ 27 ] (outcomes: spasticity, spasm, cognitive function, daily activities, motricity, and bladder function) showed no difference between groups.

• 3. Adverse effects

Adverse effects were reported in most reviews comparing cannabis with placebo (49/59, 83%). Most adverse events were reported study-by-study, with few reviews ( n = 16/59, 27%) conducting a narrative or quantitative synthesis. Serious adverse effects were reported in 21/59 (36%) reviews, and minor adverse effects were reported in 30/59 (51%) reviews. The remaining reviews did not define the difference between serious and minor adverse events. The most commonly reported serious adverse events included psychotic symptoms ( n = 6), severe dysphoric reactions ( n = 3), seizure ( n = 3), and urinary tract infection ( n = 2). The most commonly reported minor adverse events included somnolence/drowsiness ( n = 28), dizziness ( n = 27), dry mouth ( n = 20), and nausea ( n = 18). Many reviews ( n = 37/59, 63%) comparing cannabis to placebo reported both neurocognitive and non-cognitive adverse effects. Withdrawals due to adverse events were reported in 22 (37%) reviews.

Of the moderate-/high-quality reviews, adverse effect analyses were reported in reviews on pain, multiple sclerosis, cancer, HIV/AIDS, movement disorders, rheumatic disease, and several other conditions. Two reviews on pain showed fewer adverse events with cannabis for euphoria, events linked to alternations in perception, motor function, and cognitive function, withdrawal due to adverse events, sleep, and dizziness or vertigo [ 58 , 90 ]. One review on MS showed that there was no statistically significant difference between cannabis and placebo for adverse effects such as nausea, weakness, somnolence, and fatigue [ 91 ], while another review on MS/paraplegia reported fewer events in the placebo group for dizziness, somnolence, nausea, and dry mouth [ 27 ]. Within cancer reviews, one review found no statistically significant difference between cannabis and placebo for dysphoria or sedation but reported fewer events with placebo for “feeling high,” and fewer events with cannabis for withdrawal due to adverse effects [ 40 ]. In rheumatic disease, one review reported fewer total adverse events with cannabis and found no statistically significant difference between cannabis and placebo for withdrawal due to adverse events [ 51 ].

Cannabis versus other drugs

Relatively fewer reviews compared cannabis with active drugs ( n = 23/72, 32%) (Fig. ​ (Fig.8). 8 ). Many of the reviews did not synthesize studies quantitatively, and results were reported study-by-study. The most common conditions in reviews comparing cannabis to active drugs were pain, cancer, and rheumatic disease. Comparators included ibuprofen, codeine, diphenhydramine, amitriptyline, secobarbital, prochlorperazine, domperidone, metoclopramide, amisulpride, neuroleptics, isoproterenol, megestrol acetate, pregabalin, gabapentin, and opioids.

• i. Reviews focused on addressing pain across conditions. When comparing across reviews, a mix of results are observed (see Fig. ​ Fig.8), 8 ), and some were reported study-by-study. One review found no statistically significant difference between cannabinoids and codeine for nociceptive pain, postoperative pain, and cancer pain [ 65 ]. Another review favored “other drugs” (amitriptyline and pregabalin) over cannabinoids for neuropathic pain [ 90 ]. The distribution of findings was similar when restricting to moderate-to-high-quality reviews.
• ii. Reviews focused on treating a condition or family of related conditions. One review on cancer compared cannabinoids and codeine or secobarbital and reported pain results study-by-study. Another review on fibromyalgia comparing synthetic cannabinoids with amitriptyline also reported pain data study-by-study [ 39 ].
• Non-pain outcomes

Two reviews on cancer favored cannabinoids over active drugs (prochlorperazine, domperidone, metoclopramide, and neuroleptics) for patient preference and anti-emetic efficacy [ 40 , 60 ]. Non-pain outcomes were reported study-by-study for the outcome of sleep in neuropathic pain [ 90 ] and rheumatic disease [ 39 , 49 ]. In a review covering various conditions (pain, MS, anorexia, cancer, and immune deficiency), results were unclear or indeterminate for subjective measures of sleep [ 46 ].

Adverse effects were reported in 20/24 (83%) of the reviews comparing cannabis to active drugs, and only 6/20 (30%) reported a narrative or quantitative synthesis. Many reviews that reported narrative data did not specify whether adverse effects could be attributed to a placebo or active drug comparator.

Of the moderate-to-high-quality reviews, two pain reviews found no statistically significant difference for cannabis compared to codeine or amitriptyline for withdrawals due to adverse events [ 65 , 90 ]. Results from one cancer review were mixed, with fewer adverse events for cannabis (compared to prochlorperazine, domperidone, or metoclopramide) or no difference between groups, depending on the type of subgroup analysis that was conducted [ 40 ].

Cannabis + active drugs versus placebo + active drugs

Two reviews compared cannabis with placebo cannabis in combination with an active drug (opioids and gabapentin) (Figs. ​ (Figs.10 10 and ​ and11). 11 ). Both were scored to be of moderate quality. Although one review showed that cannabis plus opioids decreased chronic pain [ 80 ], another review on pain in MS included only a single study [ 81 ], precluding the ability to determine concordance of results. Cannabis displayed varied effects on non-pain outcomes, including superiority of placebo over cannabis for some outcomes. One review reported withdrawal due to adverse events study-by-study and also reported that side effects such as nausea, drowsiness, and dizziness were more frequent with higher doses of cannabinoids (data from two included studies) [ 80 ].

Cannabis versus other cannabis comparisons

Six (8%) reviews compared different cannabis formulations or doses (Figs. ​ (Figs.12 12 and ​ and13). 13 ). Almost all were reported as study-by-study results, with two reviews including only one RCT. One review for PTSD found only observational data [ 33 ] and another review on anxiety and depression combined data from one RCT with cross-sectional study data [ 19 ]. A single review on MS reported a narrative synthesis that found a benefit for spasticity. However, it was unclear if the comparator was placebo or THC alone [ 56 ]. Four reviews reported adverse effects study-by-study, with a single review comparing side effects from different dosages; in this review, combined extracts of THC and CBD were better tolerated than extracts of THC alone [ 56 ].

Cannabis versus all comparators

One review combined all comparators for the evaluation (Fig. ​ (Fig.14). 14 ). The review (combining non-users, placebo and ibuprofen) covered a range of medical conditions and was rated as low quality [ 30 ]. No adverse effects were evaluated for this comparison.

Mapping the use of quality assessment and frameworks to interpret the strength of evidence

Although 83% of reviews incorporated risk of bias assessments in their interpretation of the evidence, only 11 (15%) reviews used a framework such as GRADE to evaluate important domains other than risk of bias that would inform the strength of the evidence.

Mapping authors’ conclusions or recommendations

Most reviews (43/72 60%) indicated an inability to draw conclusions, whether due to uncertainty, inconsistent findings, lack of (high quality) evidence, or focusing their conclusion statement on the need for more research. Almost 15% of reviews (10/72) reported recommendations or conclusions that included some uncertainty. One review (1%) provided a statement of the extent of the strength of the evidence, which differed according to outcome.

Eleven reviews provided clearer conclusions (14%). Four indicated that cannabis was not effective or not cost-effective compared to placebo in relation to multiple sclerosis, acute pain, cancer, and injury. Three reviews addressing various conditions provided varying conclusions: one stated cannabis was not effective, one indicated it was modestly safe and effective, and one concluded that cannabis was safe and efficacious as short-term treatment; all reviews were of low quality. The three remaining reviews stated moderate or modest effects for improving chronic pain, compared with placebo or other analgesia; two of those reviews were of medium AMSTAR-2 quality, and one used the GRADE framework for interpreting the strength of the evidence.

The eight remaining included reviews (11%) did not provide a clear conclusion statement or reported only limitations.

Mapping authors’ limitations of the research

Several of the reviews indicated that few studies, small sample sizes, short duration of treatment, and issues related to outcomes (e.g., definition, timing, and types) were drawbacks to the literature. Some reviews noted methodological issues with and heterogeneity among studies as limitations. A few authors stated that restricting eligibility to randomized trials, English-language studies, or full publications may have affected their review results.

With the increasing use of medical cannabis, an understanding of the landscape of available evidence syntheses is needed to support evidence-informed decision-making, policy development, and to inform a research agenda. In this scoping review, we identified 72 systematic reviews evaluating medical cannabis for a range of conditions and illnesses. Half of the reviews were evaluated as being of moderate quality, with only one review scoring high on the AMSTAR-2 assessment tool.

There was disparity in the reported results across reviews, including non-synthesized (study-by-study) data, and many were unable to provide a definitive statement regarding the effectiveness of cannabis (as measured by pain reduction or other relevant outcomes), nor the extent of increased side effects and harms. This is consistent with the limitations declared in general across reviews, such as the small numbers of relevant studies, small sample sizes of individual studies, and methodological weaknesses of available studies. This common theme in review conclusions suggests that while systematic reviews may have been conducted with moderate or high methodological quality, the strength of their conclusions are driven by the availability and quality of the relevant underlying evidence, which was often found to be limited.

Relatively fewer reviews addressed adverse effects associated with cannabis, except to narratively summarize study level data. Although information was provided for placebo-controlled comparisons, none of the comparative effectiveness reviews quantitatively assessed adverse effects data. For the placebo-controlled data, although the majority of adverse effects were mild, the number of reviews reporting serious adverse effects such as psychotic symptoms [ 25 , 42 ] and suicidal ideation [ 68 , 85 ] warrants caution.

A mix of reviews supporting and not supporting the use of cannabis, according to authors’ conclusions, was identified. Readers may wish to consider the quality of the reviews, the use of differing quality assessment tools, additional considerations covered by the GRADE framework, and the potential for spin as possible reasons for these inconsistencies. It is also possible that cannabis has differing effects depending on its type (e.g., synthetic), dose, indication, the type of pain being evaluated (e.g., neuropathic), and the tools used for outcome assessment, which can be dependent on variations in condition. Of potential interest to readers may be a closer examination of the reviews evaluating chronic pain, in order to locate the source(s) of discordance. For example, one review was deemed of moderate quality, used the GRADE framework, and rated the quality of evidence for the effectiveness of cannabis for reducing neuropathic pain as moderate, suggesting that further investigation of cannabis for neuropathic pain may be warranted [ 80 ]. The exploration aspects outlined in this paragraph are beyond the purview of scoping review methodology; a detailed assessment of the reviews, including determining the overlap of included studies among similar reviews, potential reasons for the observed discordance of findings, what re-analysis of study-by-study analyses would yield, and an undertaking of missing GRADE assessments would fall outside the bounds of a scoping review and require the use of overview methodology [ 14 ].

Our findings are consistent with a recently published summary of cannabis-based medicines for chronic pain management [ 3 ]. This report found inconsistent results in systematic reviews of cannabis-based medicines compared to placebo for chronic neuropathic pain, pain management in rheumatic diseases and painful spasms in MS. The authors also concluded that cannabis was not superior to placebo in reducing cancer pain. Four out of eight included reviews scored high on the original AMSTAR tool. The variations between the two tools can be attributed to the differences in our overall assessments. Lastly, the summary report included two reviews that were not located in our original search due to language [ 93 ] and the full-text [ 94 ] of an abstract [ 95 ] that was not located in our search.

This scoping review has identified a plethora of synthesized evidence in relation to medical cannabis. For some conditions, the extent of review replication may be wasteful. Many reviews have stated that additional trials of methodologically robust design and, where possible, of sufficient sample size for precision, are needed to add to the evidence base. This undertaking may require the coordination of multi-center studies to ensure adequate power. Future trials may also help to elucidate the effect of cannabis on different outcomes.

Given authors’ reporting of issues in relation to outcomes, future prospective trials should be guided by a standardized, “core” set of outcomes to strive for consistency across studies and ensure relevance to patient-centered care. Development of those core outcomes should be developed using the Core Outcome Measures in Effectiveness Trials (COMET) methodology [ 96 ], and further consideration will need to be made in relation to what outcomes may be common across all cannabis research and which outcomes are condition-specific. With maturity of the evidence base, future systematic reviews should seek and include non-journal-published (gray literature) reports and ideally evaluate any non-English-language papers; authors should also adequately assess risk of bias and undertake appropriate syntheses of the literature.

The strengths of this scoping review include the use of an a priori protocol, peer-reviewed search strategies, a comprehensive search for reviews, and consideration of observational designs for adverse effects data. For feasibility, we restricted to English-language reviews, and it is unknown how many of the 39 reviews in other languages that we screened would have met our eligibility criteria. The decision to limit the inclusion of reviews of observational data to adverse effects data was made during the process of full-text screening and for pragmatic reasons. We also did not consider a search of the PROSPERO database for ongoing systematic reviews; however, in preparing this report, we performed a search and found that any completed reviews were already considered for eligibility or were not available at the time of our literature search. When charting results, we took a broad perspective, which may be different than if these reviews were more formally assessed during an overview of systematic reviews.

Cannabis-based medicine is a rapidly emerging field of study, with implications for both healthcare practitioners and patients. This scoping review is intended to map and collate evidence on the harms and benefits of medical cannabis. Many reviews were unable to provide firm conclusions on the effectiveness of medical cannabis, and results of reviews were mixed. Mild adverse effects were frequently but inconsistently reported, and it is possible that harms may outweigh benefits. Evidence from longer-term, adequately powered, and methodologically sound RCTs exploring different types of cannabis-based medicines is required for conclusive recommendations.

Supplementary information

Acknowledgements.

Not applicable.

Abbreviations

Authors’ contributions.

MP, AS, and BH drafted the initial version of the report. BS designed and implemented the literature search. MP, MT, and CB contributed to review of abstracts and full texts as well as data collection. MP, AS, and BH were responsible for analyses. All authors (MP, AS, MT, CB, BS, SW, MC, SK, BH) contributed to interpretation of findings and revision of drafts and approved the final version of the manuscript.

Research reported in this publication was supported by the National Center for Complementary and Integrative Health of the National Institutes of Health under award number R24AT001293. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.

Availability of data and materials

Ethics approval and consent to participate, consent for publication, competing interests.

BH has previously received honoraria from Cornerstone Research Group for provision of methodologic advice related to the conduct of systematic reviews and meta-analysis. All other authors declare that they have no conflicts of interest.

Publisher’s Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Contributor Information

Misty Pratt, Email: ac.irho@ttarpim .

Adrienne Stevens, Email: ac.irho@snevetsda .

Micere Thuku, Email: ac.irho@ukuhtm .

Claire Butler, Email: ac.irho@reltublc .

Becky Skidmore, Email: moc.sregor@eromdiksb .

L. Susan Wieland, Email: moc.liamg@dnaleiwsl .

Mark Clemons, Email: ac.hot@snomelcm .

Salmaan Kanji, Email: ac.hot@ijnaks .

Brian Hutton, Email: ac.irho@nottuhb .

Supplementary information accompanies this paper at 10.1186/s13643-019-1243-x.

Editor's Choice: Self-Managed Abortion Before and After the Dobbs Decision

• Original Investigation Effect of Early vs Late Inguinal Hernia Repair on Serious Adverse Event Rates in Preterm Infants: A Randomized Clinical Trial Martin L. Blakely, MD, MS; Andrea Krzyzaniak, MA; Melvin S. Dassinger, MD; et al deptHead Go-With 01 Title Here Authors et al -->

Just Published

• Industry Payments to US Physicians by Specialty and Product Type Ahmed Sayed, MBBS; et al. Research Letter online first Ahmed Sayed, MBBS; et al.
• Health Equity Adjustment and Hospital Performance in the Medicare Value-Based Purchasing Program Michael Liu, MPhil; et al. Original Investigation online first has active quiz Michael Liu, MPhil; et al.
• Early vs Late Inguinal Hernia Repair in Preterm Infants HIP Trial Investigators Original Investigation HIP Trial Investigators
• Provision of Medications for Self-Managed Abortion Before and After the Dobbs Decision Abigail R. A. Aiken, PhD; et al. Original Investigation online first free access has multimedia Abigail R. A. Aiken, PhD; et al.
• Donor Heart Acceptance and Race and Gender of Patients on the Transplant Waiting List Khadijah Breathett, MD, MS; et al. Original Investigation online first Khadijah Breathett, MD, MS; et al. Editorial
• Reporting on Health and War in Medical Journals Gregory Curfman, MD; et al. Editor's Note online first free access Gregory Curfman, MD; et al.
• Reporting Health Consequences of War in Medical Journals Philip Greenland, MD; et al. Viewpoint online first free access Philip Greenland, MD; et al. Editor's Note
• How’s Your Soul David Vermette, MD, MBA, MHS A Piece of My Mind online first free access David Vermette, MD, MBA, MHS
• In Love: the Sonnet and Medicine Rafael Campo, MD, MA Editor's Note online first free access Rafael Campo, MD, MA
• The Waiting Room, Revisited Rafael Campo, MD, MA Editor's Note free access Rafael Campo, MD, MA
• Risk Assessment and Prevention of Falls in Older Community-Dwelling Adults Cathleen S. Colón-Emeric, MD, MHS; et al. Review online first has active quiz has multimedia Cathleen S. Colón-Emeric, MD, MHS; et al.
• Guidelines on Falls Prevention in Older Adults Peggy B. Leung, MD; et al. JAMA Clinical Guidelines Synopsis online first has active quiz has multimedia Peggy B. Leung, MD; et al.
• Review of Common Oral Conditions Eric T. Stoopler, DMD; et al. Review has active quiz has multimedia Eric T. Stoopler, DMD; et al.
• Effect Scores to Characterize Heterogeneity of Treatment Effects Guanbo Wang, PhD; et al. JAMA Guide to Statistics and Methods online first Guanbo Wang, PhD; et al.
• Arteriovenous Access for Hemodialysis Charmaine E. Lok, MD, MSc; et al. Review online first has active quiz has multimedia Charmaine E. Lok, MD, MSc; et al.

Latest from the USPSTF

• USPSTF Recommendation: Primary Care Interventions to Prevent Child Maltreatment
• USPSTF Recommendation: Screening for Speech and Language Delay and Disorders
• USPSTF Recommendation: Screening and Preventive Interventions for Oral Health in Adults
• 47,187 Views GLP-1 Agonists for Obesity—A New Recipe for Success?
• 45,507 Views Adverse Events After XBB.1.5-Containing COVID-19 mRNA Vaccines
• 37,704 Views Brain Waves Appear to Wash Out Waste During Sleep
• 34,836 Views International Consensus Criteria for Pediatric Sepsis and Septic Shock
• 29,830 Views Effect of Tirzepatide on Maintenance of Weight Reduction
• 27,270 Views Provision of Medications for Self-Managed Abortion Before and After the Dobbs Decision
• 24,744 Views Study Provides Insight Into ME/CFS
• 23,429 Views Pharmacotherapy and Mortality in Individuals With ADHD
• 20,897 Views Paclitaxel-Coated Balloon vs Uncoated Balloon for Coronary In-Stent Restenosis
• 20,656 Views Long-Term Outcomes of Medical Management vs Bariatric Surgery in Type 2 Diabetes
• 713 Citations Antibody Response to 2-Dose SARS-CoV-2 mRNA Vaccine Series in Solid Organ Transplant Recipients
• 597 Citations Pancreatic Cancer
• 597 Citations Strengthening the Reporting of Observational Studies in Epidemiology Using Mendelian Randomization
• 582 Citations Updated Guidance on the Reporting of Race and Ethnicity in Medical and Science Journals
• 574 Citations USPSTF Recommendation: Screening for Colorectal Cancer
• 507 Citations Effect of 2 Inactivated SARS-CoV-2 Vaccines on Symptomatic COVID-19 Infection in Adults
• 446 Citations The Leading Causes of Death in the US for 2020
• 439 Citations Effect of Intermediate- vs Standard-Dose Anticoagulation on Outcomes of Patients With COVID-19
• 438 Citations Association Between IL-6 Antagonists and Mortality Among Patients Hospitalized for COVID-19
• 402 Citations Association Between 3 Doses of mRNA COVID-19 Vaccine and Symptomatic Infection Caused by Omicron and Delta Variants
• Register for email alerts with links to free full-text articles
• Access PDFs of free articles
• Manage your interests
• Save searches and receive search alerts

2024 Undergraduate Research Symposium: March 30

Please join the 2024 Undergraduate Research Symposium on Saturday, March 30 from 8am to 1pm! Posters and artwork will be available to view in the Learning Commons @ Perry Library and its Gallery section. Also, presentations will be shared in Perry Library's conference rooms. In addition, the Libraries encourage the Monarch community to explore student abstracts submitted to ODU Digital Commons , as well as some full text presentations, posters, and artwork.

The Old Dominion University Undergraduate Research Symposium, through the Perry Honors College, provides undergraduates the opportunity to present the results of their research and creative projects in a public forum. Notably, the symposium has been an important and successful annual event since 2010.

For more information, please contact Dr. David Metzger at [email protected] .  

Enhance your college career by gaining relevant experience with the skills and knowledge needed for your future career. Discover our experiential learning opportunities.

Picture yourself in the classroom, speak with professors in your major, and meet current students.

From sports games to concerts and lectures, join the ODU community at a variety of campus events. 

• Skip to Content
• Skip to Main Navigation
• Skip to Catalog Search

Indiana University Bloomington Indiana University Bloomington Bloomington

• Online Only

Search Results

• Export to Refworks
• Export to EndNote
• Librarian View

Zeitschrift für Lebensmittel-Untersuchung und -Forschung

Blmgtn - auxiliary library facility, select campus.

Access to this resource may be restricted to users from specific IU campuses.

Please choose your campus in order to proceed:

Browse by call number

• Zeitschrift für Untersuchung der Nahrungs- und Genussmittel TX501 .Z4 Index v.1-16 1898-1908
• Zeitschrift für Untersuchung der Lebensmittel TX501 .Z4 Index v.51-80
• Zeitschrift für Lebensmittel-Untersuchung und -Forschung TX501 .Z4 v.100 1955
• Nutrients in processed foods TX511 .A4 v.1
• Analysis of volatiles : methods and applications : proceedings, international workshop, Würzburg, Federal Republic of Germany, September 28-30, 1983 TX511 .A53 1984

Thank you for visiting nature.com. You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

• View all journals
• My Account Login
• Explore content
• About the journal
• Publish with us
• Sign up for alerts
• Open access
• Published: 26 March 2024

Predicting and improving complex beer flavor through machine learning

• Michiel Schreurs   ORCID: orcid.org/0000-0002-9449-5619 1 , 2 , 3   na1 ,
• Supinya Piampongsant 1 , 2 , 3   na1 ,
• Miguel Roncoroni   ORCID: orcid.org/0000-0001-7461-1427 1 , 2 , 3   na1 ,
• Lloyd Cool   ORCID: orcid.org/0000-0001-9936-3124 1 , 2 , 3 , 4 ,
• Beatriz Herrera-Malaver   ORCID: orcid.org/0000-0002-5096-9974 1 , 2 , 3 ,
• Christophe Vanderaa   ORCID: orcid.org/0000-0001-7443-5427 4 ,
• Florian A. Theßeling 1 , 2 , 3 ,
• Łukasz Kreft   ORCID: orcid.org/0000-0001-7620-4657 5 ,
• Alexander Botzki   ORCID: orcid.org/0000-0001-6691-4233 5 ,
• Philippe Malcorps 6 ,
• Luk Daenen 6 ,
• Tom Wenseleers   ORCID: orcid.org/0000-0002-1434-861X 4 &
• Kevin J. Verstrepen   ORCID: orcid.org/0000-0002-3077-6219 1 , 2 , 3  

Nature Communications volume  15 , Article number:  2368 ( 2024 ) Cite this article

7171 Accesses

681 Altmetric

Metrics details

• Chemical engineering
• Gas chromatography
• Machine learning
• Metabolomics
• Taste receptors

The perception and appreciation of food flavor depends on many interacting chemical compounds and external factors, and therefore proves challenging to understand and predict. Here, we combine extensive chemical and sensory analyses of 250 different beers to train machine learning models that allow predicting flavor and consumer appreciation. For each beer, we measure over 200 chemical properties, perform quantitative descriptive sensory analysis with a trained tasting panel and map data from over 180,000 consumer reviews to train 10 different machine learning models. The best-performing algorithm, Gradient Boosting, yields models that significantly outperform predictions based on conventional statistics and accurately predict complex food features and consumer appreciation from chemical profiles. Model dissection allows identifying specific and unexpected compounds as drivers of beer flavor and appreciation. Adding these compounds results in variants of commercial alcoholic and non-alcoholic beers with improved consumer appreciation. Together, our study reveals how big data and machine learning uncover complex links between food chemistry, flavor and consumer perception, and lays the foundation to develop novel, tailored foods with superior flavors.

Similar content being viewed by others

BitterSweet: Building machine learning models for predicting the bitter and sweet taste of small molecules

Rudraksh Tuwani, Somin Wadhwa & Ganesh Bagler

Sensory lexicon and aroma volatiles analysis of brewing malt

Xiaoxia Su, Miao Yu, … Tianyi Du

Predicting odor from molecular structure: a multi-label classification approach

Kushagra Saini & Venkatnarayan Ramanathan

Introduction

Predicting and understanding food perception and appreciation is one of the major challenges in food science. Accurate modeling of food flavor and appreciation could yield important opportunities for both producers and consumers, including quality control, product fingerprinting, counterfeit detection, spoilage detection, and the development of new products and product combinations (food pairing) 1 , 2 , 3 , 4 , 5 , 6 . Accurate models for flavor and consumer appreciation would contribute greatly to our scientific understanding of how humans perceive and appreciate flavor. Moreover, accurate predictive models would also facilitate and standardize existing food assessment methods and could supplement or replace assessments by trained and consumer tasting panels, which are variable, expensive and time-consuming 7 , 8 , 9 . Lastly, apart from providing objective, quantitative, accurate and contextual information that can help producers, models can also guide consumers in understanding their personal preferences 10 .

Despite the myriad of applications, predicting food flavor and appreciation from its chemical properties remains a largely elusive goal in sensory science, especially for complex food and beverages 11 , 12 . A key obstacle is the immense number of flavor-active chemicals underlying food flavor. Flavor compounds can vary widely in chemical structure and concentration, making them technically challenging and labor-intensive to quantify, even in the face of innovations in metabolomics, such as non-targeted metabolic fingerprinting 13 , 14 . Moreover, sensory analysis is perhaps even more complicated. Flavor perception is highly complex, resulting from hundreds of different molecules interacting at the physiochemical and sensorial level. Sensory perception is often non-linear, characterized by complex and concentration-dependent synergistic and antagonistic effects 15 , 16 , 17 , 18 , 19 , 20 , 21 that are further convoluted by the genetics, environment, culture and psychology of consumers 22 , 23 , 24 . Perceived flavor is therefore difficult to measure, with problems of sensitivity, accuracy, and reproducibility that can only be resolved by gathering sufficiently large datasets 25 . Trained tasting panels are considered the prime source of quality sensory data, but require meticulous training, are low throughput and high cost. Public databases containing consumer reviews of food products could provide a valuable alternative, especially for studying appreciation scores, which do not require formal training 25 . Public databases offer the advantage of amassing large amounts of data, increasing the statistical power to identify potential drivers of appreciation. However, public datasets suffer from biases, including a bias in the volunteers that contribute to the database, as well as confounding factors such as price, cult status and psychological conformity towards previous ratings of the product.

Classical multivariate statistics and machine learning methods have been used to predict flavor of specific compounds by, for example, linking structural properties of a compound to its potential biological activities or linking concentrations of specific compounds to sensory profiles 1 , 26 . Importantly, most previous studies focused on predicting organoleptic properties of single compounds (often based on their chemical structure) 27 , 28 , 29 , 30 , 31 , 32 , 33 , thus ignoring the fact that these compounds are present in a complex matrix in food or beverages and excluding complex interactions between compounds. Moreover, the classical statistics commonly used in sensory science 34 , 35 , 36 , 37 , 38 , 39 require a large sample size and sufficient variance amongst predictors to create accurate models. They are not fit for studying an extensive set of hundreds of interacting flavor compounds, since they are sensitive to outliers, have a high tendency to overfit and are less suited for non-linear and discontinuous relationships 40 .

In this study, we combine extensive chemical analyses and sensory data of a set of different commercial beers with machine learning approaches to develop models that predict taste, smell, mouthfeel and appreciation from compound concentrations. Beer is particularly suited to model the relationship between chemistry, flavor and appreciation. First, beer is a complex product, consisting of thousands of flavor compounds that partake in complex sensory interactions 41 , 42 , 43 . This chemical diversity arises from the raw materials (malt, yeast, hops, water and spices) and biochemical conversions during the brewing process (kilning, mashing, boiling, fermentation, maturation and aging) 44 , 45 . Second, the advent of the internet saw beer consumers embrace online review platforms, such as RateBeer (ZX Ventures, Anheuser-Busch InBev SA/NV) and BeerAdvocate (Next Glass, inc.). In this way, the beer community provides massive data sets of beer flavor and appreciation scores, creating extraordinarily large sensory databases to complement the analyses of our professional sensory panel. Specifically, we characterize over 200 chemical properties of 250 commercial beers, spread across 22 beer styles, and link these to the descriptive sensory profiling data of a 16-person in-house trained tasting panel and data acquired from over 180,000 public consumer reviews. These unique and extensive datasets enable us to train a suite of machine learning models to predict flavor and appreciation from a beer’s chemical profile. Dissection of the best-performing models allows us to pinpoint specific compounds as potential drivers of beer flavor and appreciation. Follow-up experiments confirm the importance of these compounds and ultimately allow us to significantly improve the flavor and appreciation of selected commercial beers. Together, our study represents a significant step towards understanding complex flavors and reinforces the value of machine learning to develop and refine complex foods. In this way, it represents a stepping stone for further computer-aided food engineering applications 46 .

To generate a comprehensive dataset on beer flavor, we selected 250 commercial Belgian beers across 22 different beer styles (Supplementary Fig.  S1 ). Beers with ≤ 4.2% alcohol by volume (ABV) were classified as non-alcoholic and low-alcoholic. Blonds and Tripels constitute a significant portion of the dataset (12.4% and 11.2%, respectively) reflecting their presence on the Belgian beer market and the heterogeneity of beers within these styles. By contrast, lager beers are less diverse and dominated by a handful of brands. Rare styles such as Brut or Faro make up only a small fraction of the dataset (2% and 1%, respectively) because fewer of these beers are produced and because they are dominated by distinct characteristics in terms of flavor and chemical composition.

Extensive analysis identifies relationships between chemical compounds in beer

For each beer, we measured 226 different chemical properties, including common brewing parameters such as alcohol content, iso-alpha acids, pH, sugar concentration 47 , and over 200 flavor compounds (Methods, Supplementary Table  S1 ). A large portion (37.2%) are terpenoids arising from hopping, responsible for herbal and fruity flavors 16 , 48 . A second major category are yeast metabolites, such as esters and alcohols, that result in fruity and solvent notes 48 , 49 , 50 . Other measured compounds are primarily derived from malt, or other microbes such as non- Saccharomyces yeasts and bacteria (‘wild flora’). Compounds that arise from spices or staling are labeled under ‘Others’. Five attributes (caloric value, total acids and total ester, hop aroma and sulfur compounds) are calculated from multiple individually measured compounds.

As a first step in identifying relationships between chemical properties, we determined correlations between the concentrations of the compounds (Fig.  1 , upper panel, Supplementary Data  1 and 2 , and Supplementary Fig.  S2 . For the sake of clarity, only a subset of the measured compounds is shown in Fig.  1 ). Compounds of the same origin typically show a positive correlation, while absence of correlation hints at parameters varying independently. For example, the hop aroma compounds citronellol, and alpha-terpineol show moderate correlations with each other (Spearman’s rho=0.39 and 0.57), but not with the bittering hop component iso-alpha acids (Spearman’s rho=0.16 and −0.07). This illustrates how brewers can independently modify hop aroma and bitterness by selecting hop varieties and dosage time. If hops are added early in the boiling phase, chemical conversions increase bitterness while aromas evaporate, conversely, late addition of hops preserves aroma but limits bitterness 51 . Similarly, hop-derived iso-alpha acids show a strong anti-correlation with lactic acid and acetic acid, likely reflecting growth inhibition of lactic acid and acetic acid bacteria, or the consequent use of fewer hops in sour beer styles, such as West Flanders ales and Fruit beers, that rely on these bacteria for their distinct flavors 52 . Finally, yeast-derived esters (ethyl acetate, ethyl decanoate, ethyl hexanoate, ethyl octanoate) and alcohols (ethanol, isoamyl alcohol, isobutanol, and glycerol), correlate with Spearman coefficients above 0.5, suggesting that these secondary metabolites are correlated with the yeast genetic background and/or fermentation parameters and may be difficult to influence individually, although the choice of yeast strain may offer some control 53 .

Spearman rank correlations are shown. Descriptors are grouped according to their origin (malt (blue), hops (green), yeast (red), wild flora (yellow), Others (black)), and sensory aspect (aroma, taste, palate, and overall appreciation). Please note that for the chemical compounds, for the sake of clarity, only a subset of the total number of measured compounds is shown, with an emphasis on the key compounds for each source. For more details, see the main text and Methods section. Chemical data can be found in Supplementary Data  1 , correlations between all chemical compounds are depicted in Supplementary Fig.  S2 and correlation values can be found in Supplementary Data  2 . See Supplementary Data  4 for sensory panel assessments and Supplementary Data  5 for correlation values between all sensory descriptors.

Interestingly, different beer styles show distinct patterns for some flavor compounds (Supplementary Fig.  S3 ). These observations agree with expectations for key beer styles, and serve as a control for our measurements. For instance, Stouts generally show high values for color (darker), while hoppy beers contain elevated levels of iso-alpha acids, compounds associated with bitter hop taste. Acetic and lactic acid are not prevalent in most beers, with notable exceptions such as Kriek, Lambic, Faro, West Flanders ales and Flanders Old Brown, which use acid-producing bacteria ( Lactobacillus and Pediococcus ) or unconventional yeast ( Brettanomyces ) 54 , 55 . Glycerol, ethanol and esters show similar distributions across all beer styles, reflecting their common origin as products of yeast metabolism during fermentation 45 , 53 . Finally, low/no-alcohol beers contain low concentrations of glycerol and esters. This is in line with the production process for most of the low/no-alcohol beers in our dataset, which are produced through limiting fermentation or by stripping away alcohol via evaporation or dialysis, with both methods having the unintended side-effect of reducing the amount of flavor compounds in the final beer 56 , 57 .

Besides expected associations, our data also reveals less trivial associations between beer styles and specific parameters. For example, geraniol and citronellol, two monoterpenoids responsible for citrus, floral and rose flavors and characteristic of Citra hops, are found in relatively high amounts in Christmas, Saison, and Brett/co-fermented beers, where they may originate from terpenoid-rich spices such as coriander seeds instead of hops 58 .

Tasting panel assessments reveal sensorial relationships in beer

To assess the sensory profile of each beer, a trained tasting panel evaluated each of the 250 beers for 50 sensory attributes, including different hop, malt and yeast flavors, off-flavors and spices. Panelists used a tasting sheet (Supplementary Data  3 ) to score the different attributes. Panel consistency was evaluated by repeating 12 samples across different sessions and performing ANOVA. In 95% of cases no significant difference was found across sessions ( p  > 0.05), indicating good panel consistency (Supplementary Table  S2 ).

Aroma and taste perception reported by the trained panel are often linked (Fig.  1 , bottom left panel and Supplementary Data  4 and 5 ), with high correlations between hops aroma and taste (Spearman’s rho=0.83). Bitter taste was found to correlate with hop aroma and taste in general (Spearman’s rho=0.80 and 0.69), and particularly with “grassy” noble hops (Spearman’s rho=0.75). Barnyard flavor, most often associated with sour beers, is identified together with stale hops (Spearman’s rho=0.97) that are used in these beers. Lactic and acetic acid, which often co-occur, are correlated (Spearman’s rho=0.66). Interestingly, sweetness and bitterness are anti-correlated (Spearman’s rho = −0.48), confirming the hypothesis that they mask each other 59 , 60 . Beer body is highly correlated with alcohol (Spearman’s rho = 0.79), and overall appreciation is found to correlate with multiple aspects that describe beer mouthfeel (alcohol, carbonation; Spearman’s rho= 0.32, 0.39), as well as with hop and ester aroma intensity (Spearman’s rho=0.39 and 0.35).

Similar to the chemical analyses, sensorial analyses confirmed typical features of specific beer styles (Supplementary Fig.  S4 ). For example, sour beers (Faro, Flanders Old Brown, Fruit beer, Kriek, Lambic, West Flanders ale) were rated acidic, with flavors of both acetic and lactic acid. Hoppy beers were found to be bitter and showed hop-associated aromas like citrus and tropical fruit. Malt taste is most detected among scotch, stout/porters, and strong ales, while low/no-alcohol beers, which often have a reputation for being ‘worty’ (reminiscent of unfermented, sweet malt extract) appear in the middle. Unsurprisingly, hop aromas are most strongly detected among hoppy beers. Like its chemical counterpart (Supplementary Fig.  S3 ), acidity shows a right-skewed distribution, with the most acidic beers being Krieks, Lambics, and West Flanders ales.

Tasting panel assessments of specific flavors correlate with chemical composition

We find that the concentrations of several chemical compounds strongly correlate with specific aroma or taste, as evaluated by the tasting panel (Fig.  2 , Supplementary Fig.  S5 , Supplementary Data  6 ). In some cases, these correlations confirm expectations and serve as a useful control for data quality. For example, iso-alpha acids, the bittering compounds in hops, strongly correlate with bitterness (Spearman’s rho=0.68), while ethanol and glycerol correlate with tasters’ perceptions of alcohol and body, the mouthfeel sensation of fullness (Spearman’s rho=0.82/0.62 and 0.72/0.57 respectively) and darker color from roasted malts is a good indication of malt perception (Spearman’s rho=0.54).

Heatmap colors indicate Spearman’s Rho. Axes are organized according to sensory categories (aroma, taste, mouthfeel, overall), chemical categories and chemical sources in beer (malt (blue), hops (green), yeast (red), wild flora (yellow), Others (black)). See Supplementary Data  6 for all correlation values.

Interestingly, for some relationships between chemical compounds and perceived flavor, correlations are weaker than expected. For example, the rose-smelling phenethyl acetate only weakly correlates with floral aroma. This hints at more complex relationships and interactions between compounds and suggests a need for a more complex model than simple correlations. Lastly, we uncovered unexpected correlations. For instance, the esters ethyl decanoate and ethyl octanoate appear to correlate slightly with hop perception and bitterness, possibly due to their fruity flavor. Iron is anti-correlated with hop aromas and bitterness, most likely because it is also anti-correlated with iso-alpha acids. This could be a sign of metal chelation of hop acids 61 , given that our analyses measure unbound hop acids and total iron content, or could result from the higher iron content in dark and Fruit beers, which typically have less hoppy and bitter flavors 62 .

Public consumer reviews complement expert panel data

To complement and expand the sensory data of our trained tasting panel, we collected 180,000 reviews of our 250 beers from the online consumer review platform RateBeer. This provided numerical scores for beer appearance, aroma, taste, palate, overall quality as well as the average overall score.

Public datasets are known to suffer from biases, such as price, cult status and psychological conformity towards previous ratings of a product. For example, prices correlate with appreciation scores for these online consumer reviews (rho=0.49, Supplementary Fig.  S6 ), but not for our trained tasting panel (rho=0.19). This suggests that prices affect consumer appreciation, which has been reported in wine 63 , while blind tastings are unaffected. Moreover, we observe that some beer styles, like lagers and non-alcoholic beers, generally receive lower scores, reflecting that online reviewers are mostly beer aficionados with a preference for specialty beers over lager beers. In general, we find a modest correlation between our trained panel’s overall appreciation score and the online consumer appreciation scores (Fig.  3 , rho=0.29). Apart from the aforementioned biases in the online datasets, serving temperature, sample freshness and surroundings, which are all tightly controlled during the tasting panel sessions, can vary tremendously across online consumers and can further contribute to (among others, appreciation) differences between the two categories of tasters. Importantly, in contrast to the overall appreciation scores, for many sensory aspects the results from the professional panel correlated well with results obtained from RateBeer reviews. Correlations were highest for features that are relatively easy to recognize even for untrained tasters, like bitterness, sweetness, alcohol and malt aroma (Fig.  3 and below).

RateBeer text mining results can be found in Supplementary Data  7 . Rho values shown are Spearman correlation values, with asterisks indicating significant correlations ( p  < 0.05, two-sided). All p values were smaller than 0.001, except for Esters aroma (0.0553), Esters taste (0.3275), Esters aroma—banana (0.0019), Coriander (0.0508) and Diacetyl (0.0134).

Besides collecting consumer appreciation from these online reviews, we developed automated text analysis tools to gather additional data from review texts (Supplementary Data  7 ). Processing review texts on the RateBeer database yielded comparable results to the scores given by the trained panel for many common sensory aspects, including acidity, bitterness, sweetness, alcohol, malt, and hop tastes (Fig.  3 ). This is in line with what would be expected, since these attributes require less training for accurate assessment and are less influenced by environmental factors such as temperature, serving glass and odors in the environment. Consumer reviews also correlate well with our trained panel for 4-vinyl guaiacol, a compound associated with a very characteristic aroma. By contrast, correlations for more specific aromas like ester, coriander or diacetyl are underrepresented in the online reviews, underscoring the importance of using a trained tasting panel and standardized tasting sheets with explicit factors to be scored for evaluating specific aspects of a beer. Taken together, our results suggest that public reviews are trustworthy for some, but not all, flavor features and can complement or substitute taste panel data for these sensory aspects.

Models can predict beer sensory profiles from chemical data

The rich datasets of chemical analyses, tasting panel assessments and public reviews gathered in the first part of this study provided us with a unique opportunity to develop predictive models that link chemical data to sensorial features. Given the complexity of beer flavor, basic statistical tools such as correlations or linear regression may not always be the most suitable for making accurate predictions. Instead, we applied different machine learning models that can model both simple linear and complex interactive relationships. Specifically, we constructed a set of regression models to predict (a) trained panel scores for beer flavor and quality and (b) public reviews’ appreciation scores from beer chemical profiles. We trained and tested 10 different models (Methods), 3 linear regression-based models (simple linear regression with first-order interactions (LR), lasso regression with first-order interactions (Lasso), partial least squares regressor (PLSR)), 5 decision tree models (AdaBoost regressor (ABR), extra trees (ET), gradient boosting regressor (GBR), random forest (RF) and XGBoost regressor (XGBR)), 1 support vector regression (SVR), and 1 artificial neural network (ANN) model.

To compare the performance of our machine learning models, the dataset was randomly split into a training and test set, stratified by beer style. After a model was trained on data in the training set, its performance was evaluated on its ability to predict the test dataset obtained from multi-output models (based on the coefficient of determination, see Methods). Additionally, individual-attribute models were ranked per descriptor and the average rank was calculated, as proposed by Korneva et al. 64 . Importantly, both ways of evaluating the models’ performance agreed in general. Performance of the different models varied (Table  1 ). It should be noted that all models perform better at predicting RateBeer results than results from our trained tasting panel. One reason could be that sensory data is inherently variable, and this variability is averaged out with the large number of public reviews from RateBeer. Additionally, all tree-based models perform better at predicting taste than aroma. Linear models (LR) performed particularly poorly, with negative R 2 values, due to severe overfitting (training set R 2  = 1). Overfitting is a common issue in linear models with many parameters and limited samples, especially with interaction terms further amplifying the number of parameters. L1 regularization (Lasso) successfully overcomes this overfitting, out-competing multiple tree-based models on the RateBeer dataset. Similarly, the dimensionality reduction of PLSR avoids overfitting and improves performance, to some extent. Still, tree-based models (ABR, ET, GBR, RF and XGBR) show the best performance, out-competing the linear models (LR, Lasso, PLSR) commonly used in sensory science 65 .

GBR models showed the best overall performance in predicting sensory responses from chemical information, with R 2 values up to 0.75 depending on the predicted sensory feature (Supplementary Table  S4 ). The GBR models predict consumer appreciation (RateBeer) better than our trained panel’s appreciation (R 2 value of 0.67 compared to R 2 value of 0.09) (Supplementary Table  S3 and Supplementary Table  S4 ). ANN models showed intermediate performance, likely because neural networks typically perform best with larger datasets 66 . The SVR shows intermediate performance, mostly due to the weak predictions of specific attributes that lower the overall performance (Supplementary Table  S4 ).

Model dissection identifies specific, unexpected compounds as drivers of consumer appreciation

Next, we leveraged our models to infer important contributors to sensory perception and consumer appreciation. Consumer preference is a crucial sensory aspects, because a product that shows low consumer appreciation scores often does not succeed commercially 25 . Additionally, the requirement for a large number of representative evaluators makes consumer trials one of the more costly and time-consuming aspects of product development. Hence, a model for predicting chemical drivers of overall appreciation would be a welcome addition to the available toolbox for food development and optimization.

Since GBR models on our RateBeer dataset showed the best overall performance, we focused on these models. Specifically, we used two approaches to identify important contributors. First, rankings of the most important predictors for each sensorial trait in the GBR models were obtained based on impurity-based feature importance (mean decrease in impurity). High-ranked parameters were hypothesized to be either the true causal chemical properties underlying the trait, to correlate with the actual causal properties, or to take part in sensory interactions affecting the trait 67 (Fig.  4A ). In a second approach, we used SHAP 68 to determine which parameters contributed most to the model for making predictions of consumer appreciation (Fig.  4B ). SHAP calculates parameter contributions to model predictions on a per-sample basis, which can be aggregated into an importance score.

A The impurity-based feature importance (mean deviance in impurity, MDI) calculated from the Gradient Boosting Regression (GBR) model predicting RateBeer appreciation scores. The top 15 highest ranked chemical properties are shown. B SHAP summary plot for the top 15 parameters contributing to our GBR model. Each point on the graph represents a sample from our dataset. The color represents the concentration of that parameter, with bluer colors representing low values and redder colors representing higher values. Greater absolute values on the horizontal axis indicate a higher impact of the parameter on the prediction of the model. C Spearman correlations between the 15 most important chemical properties and consumer overall appreciation. Numbers indicate the Spearman Rho correlation coefficient, and the rank of this correlation compared to all other correlations. The top 15 important compounds were determined using SHAP (panel B).

Both approaches identified ethyl acetate as the most predictive parameter for beer appreciation (Fig.  4 ). Ethyl acetate is the most abundant ester in beer with a typical ‘fruity’, ‘solvent’ and ‘alcoholic’ flavor, but is often considered less important than other esters like isoamyl acetate. The second most important parameter identified by SHAP is ethanol, the most abundant beer compound after water. Apart from directly contributing to beer flavor and mouthfeel, ethanol drastically influences the physical properties of beer, dictating how easily volatile compounds escape the beer matrix to contribute to beer aroma 69 . Importantly, it should also be noted that the importance of ethanol for appreciation is likely inflated by the very low appreciation scores of non-alcoholic beers (Supplementary Fig.  S4 ). Despite not often being considered a driver of beer appreciation, protein level also ranks highly in both approaches, possibly due to its effect on mouthfeel and body 70 . Lactic acid, which contributes to the tart taste of sour beers, is the fourth most important parameter identified by SHAP, possibly due to the generally high appreciation of sour beers in our dataset.

Interestingly, some of the most important predictive parameters for our model are not well-established as beer flavors or are even commonly regarded as being negative for beer quality. For example, our models identify methanethiol and ethyl phenyl acetate, an ester commonly linked to beer staling 71 , as a key factor contributing to beer appreciation. Although there is no doubt that high concentrations of these compounds are considered unpleasant, the positive effects of modest concentrations are not yet known 72 , 73 .

To compare our approach to conventional statistics, we evaluated how well the 15 most important SHAP-derived parameters correlate with consumer appreciation (Fig.  4C ). Interestingly, only 6 of the properties derived by SHAP rank amongst the top 15 most correlated parameters. For some chemical compounds, the correlations are so low that they would have likely been considered unimportant. For example, lactic acid, the fourth most important parameter, shows a bimodal distribution for appreciation, with sour beers forming a separate cluster, that is missed entirely by the Spearman correlation. Additionally, the correlation plots reveal outliers, emphasizing the need for robust analysis tools. Together, this highlights the need for alternative models, like the Gradient Boosting model, that better grasp the complexity of (beer) flavor.

Finally, to observe the relationships between these chemical properties and their predicted targets, partial dependence plots were constructed for the six most important predictors of consumer appreciation 74 , 75 , 76 (Supplementary Fig.  S7 ). One-way partial dependence plots show how a change in concentration affects the predicted appreciation. These plots reveal an important limitation of our models: appreciation predictions remain constant at ever-increasing concentrations. This implies that once a threshold concentration is reached, further increasing the concentration does not affect appreciation. This is false, as it is well-documented that certain compounds become unpleasant at high concentrations, including ethyl acetate (‘nail polish’) 77 and methanethiol (‘sulfury’ and ‘rotten cabbage’) 78 . The inability of our models to grasp that flavor compounds have optimal levels, above which they become negative, is a consequence of working with commercial beer brands where (off-)flavors are rarely too high to negatively impact the product. The two-way partial dependence plots show how changing the concentration of two compounds influences predicted appreciation, visualizing their interactions (Supplementary Fig.  S7 ). In our case, the top 5 parameters are dominated by additive or synergistic interactions, with high concentrations for both compounds resulting in the highest predicted appreciation.

To assess the robustness of our best-performing models and model predictions, we performed 100 iterations of the GBR, RF and ET models. In general, all iterations of the models yielded similar performance (Supplementary Fig.  S8 ). Moreover, the main predictors (including the top predictors ethanol and ethyl acetate) remained virtually the same, especially for GBR and RF. For the iterations of the ET model, we did observe more variation in the top predictors, which is likely a consequence of the model’s inherent random architecture in combination with co-correlations between certain predictors. However, even in this case, several of the top predictors (ethanol and ethyl acetate) remain unchanged, although their rank in importance changes (Supplementary Fig.  S8 ).

Next, we investigated if a combination of RateBeer and trained panel data into one consolidated dataset would lead to stronger models, under the hypothesis that such a model would suffer less from bias in the datasets. A GBR model was trained to predict appreciation on the combined dataset. This model underperformed compared to the RateBeer model, both in the native case and when including a dataset identifier (R 2  = 0.67, 0.26 and 0.42 respectively). For the latter, the dataset identifier is the most important feature (Supplementary Fig.  S9 ), while most of the feature importance remains unchanged, with ethyl acetate and ethanol ranking highest, like in the original model trained only on RateBeer data. It seems that the large variation in the panel dataset introduces noise, weakening the models’ performances and reliability. In addition, it seems reasonable to assume that both datasets are fundamentally different, with the panel dataset obtained by blind tastings by a trained professional panel.

Lastly, we evaluated whether beer style identifiers would further enhance the model’s performance. A GBR model was trained with parameters that explicitly encoded the styles of the samples. This did not improve model performance (R2 = 0.66 with style information vs R2 = 0.67). The most important chemical features are consistent with the model trained without style information (eg. ethanol and ethyl acetate), and with the exception of the most preferred (strong ale) and least preferred (low/no-alcohol) styles, none of the styles were among the most important features (Supplementary Fig.  S9 , Supplementary Table  S5 and S6 ). This is likely due to a combination of style-specific chemical signatures, such as iso-alpha acids and lactic acid, that implicitly convey style information to the original models, as well as the low number of samples belonging to some styles, making it difficult for the model to learn style-specific patterns. Moreover, beer styles are not rigorously defined, with some styles overlapping in features and some beers being misattributed to a specific style, all of which leads to more noise in models that use style parameters.

Model validation

To test if our predictive models give insight into beer appreciation, we set up experiments aimed at improving existing commercial beers. We specifically selected overall appreciation as the trait to be examined because of its complexity and commercial relevance. Beer flavor comprises a complex bouquet rather than single aromas and tastes 53 . Hence, adding a single compound to the extent that a difference is noticeable may lead to an unbalanced, artificial flavor. Therefore, we evaluated the effect of combinations of compounds. Because Blond beers represent the most extensive style in our dataset, we selected a beer from this style as the starting material for these experiments (Beer 64 in Supplementary Data  1 ).

In the first set of experiments, we adjusted the concentrations of compounds that made up the most important predictors of overall appreciation (ethyl acetate, ethanol, lactic acid, ethyl phenyl acetate) together with correlated compounds (ethyl hexanoate, isoamyl acetate, glycerol), bringing them up to 95 th percentile ethanol-normalized concentrations (Methods) within the Blond group (‘Spiked’ concentration in Fig.  5A ). Compared to controls, the spiked beers were found to have significantly improved overall appreciation among trained panelists, with panelist noting increased intensity of ester flavors, sweetness, alcohol, and body fullness (Fig.  5B ). To disentangle the contribution of ethanol to these results, a second experiment was performed without the addition of ethanol. This resulted in a similar outcome, including increased perception of alcohol and overall appreciation.

Adding the top chemical compounds, identified as best predictors of appreciation by our model, into poorly appreciated beers results in increased appreciation from our trained panel. Results of sensory tests between base beers and those spiked with compounds identified as the best predictors by the model. A Blond and Non/Low-alcohol (0.0% ABV) base beers were brought up to 95th-percentile ethanol-normalized concentrations within each style. B For each sensory attribute, tasters indicated the more intense sample and selected the sample they preferred. The numbers above the bars correspond to the p values that indicate significant changes in perceived flavor (two-sided binomial test: alpha 0.05, n  = 20 or 13).

In a last experiment, we tested whether using the model’s predictions can boost the appreciation of a non-alcoholic beer (beer 223 in Supplementary Data  1 ). Again, the addition of a mixture of predicted compounds (omitting ethanol, in this case) resulted in a significant increase in appreciation, body, ester flavor and sweetness.

Predicting flavor and consumer appreciation from chemical composition is one of the ultimate goals of sensory science. A reliable, systematic and unbiased way to link chemical profiles to flavor and food appreciation would be a significant asset to the food and beverage industry. Such tools would substantially aid in quality control and recipe development, offer an efficient and cost-effective alternative to pilot studies and consumer trials and would ultimately allow food manufacturers to produce superior, tailor-made products that better meet the demands of specific consumer groups more efficiently.

A limited set of studies have previously tried, to varying degrees of success, to predict beer flavor and beer popularity based on (a limited set of) chemical compounds and flavors 79 , 80 . Current sensitive, high-throughput technologies allow measuring an unprecedented number of chemical compounds and properties in a large set of samples, yielding a dataset that can train models that help close the gaps between chemistry and flavor, even for a complex natural product like beer. To our knowledge, no previous research gathered data at this scale (250 samples, 226 chemical parameters, 50 sensory attributes and 5 consumer scores) to disentangle and validate the chemical aspects driving beer preference using various machine-learning techniques. We find that modern machine learning models outperform conventional statistical tools, such as correlations and linear models, and can successfully predict flavor appreciation from chemical composition. This could be attributed to the natural incorporation of interactions and non-linear or discontinuous effects in machine learning models, which are not easily grasped by the linear model architecture. While linear models and partial least squares regression represent the most widespread statistical approaches in sensory science, in part because they allow interpretation 65 , 81 , 82 , modern machine learning methods allow for building better predictive models while preserving the possibility to dissect and exploit the underlying patterns. Of the 10 different models we trained, tree-based models, such as our best performing GBR, showed the best overall performance in predicting sensory responses from chemical information, outcompeting artificial neural networks. This agrees with previous reports for models trained on tabular data 83 . Our results are in line with the findings of Colantonio et al. who also identified the gradient boosting architecture as performing best at predicting appreciation and flavor (of tomatoes and blueberries, in their specific study) 26 . Importantly, besides our larger experimental scale, we were able to directly confirm our models’ predictions in vivo.

Our study confirms that flavor compound concentration does not always correlate with perception, suggesting complex interactions that are often missed by more conventional statistics and simple models. Specifically, we find that tree-based algorithms may perform best in developing models that link complex food chemistry with aroma. Furthermore, we show that massive datasets of untrained consumer reviews provide a valuable source of data, that can complement or even replace trained tasting panels, especially for appreciation and basic flavors, such as sweetness and bitterness. This holds despite biases that are known to occur in such datasets, such as price or conformity bias. Moreover, GBR models predict taste better than aroma. This is likely because taste (e.g. bitterness) often directly relates to the corresponding chemical measurements (e.g., iso-alpha acids), whereas such a link is less clear for aromas, which often result from the interplay between multiple volatile compounds. We also find that our models are best at predicting acidity and alcohol, likely because there is a direct relation between the measured chemical compounds (acids and ethanol) and the corresponding perceived sensorial attribute (acidity and alcohol), and because even untrained consumers are generally able to recognize these flavors and aromas.

The predictions of our final models, trained on review data, hold even for blind tastings with small groups of trained tasters, as demonstrated by our ability to validate specific compounds as drivers of beer flavor and appreciation. Since adding a single compound to the extent of a noticeable difference may result in an unbalanced flavor profile, we specifically tested our identified key drivers as a combination of compounds. While this approach does not allow us to validate if a particular single compound would affect flavor and/or appreciation, our experiments do show that this combination of compounds increases consumer appreciation.

It is important to stress that, while it represents an important step forward, our approach still has several major limitations. A key weakness of the GBR model architecture is that amongst co-correlating variables, the largest main effect is consistently preferred for model building. As a result, co-correlating variables often have artificially low importance scores, both for impurity and SHAP-based methods, like we observed in the comparison to the more randomized Extra Trees models. This implies that chemicals identified as key drivers of a specific sensory feature by GBR might not be the true causative compounds, but rather co-correlate with the actual causative chemical. For example, the high importance of ethyl acetate could be (partially) attributed to the total ester content, ethanol or ethyl hexanoate (rho=0.77, rho=0.72 and rho=0.68), while ethyl phenylacetate could hide the importance of prenyl isobutyrate and ethyl benzoate (rho=0.77 and rho=0.76). Expanding our GBR model to include beer style as a parameter did not yield additional power or insight. This is likely due to style-specific chemical signatures, such as iso-alpha acids and lactic acid, that implicitly convey style information to the original model, as well as the smaller sample size per style, limiting the power to uncover style-specific patterns. This can be partly attributed to the curse of dimensionality, where the high number of parameters results in the models mainly incorporating single parameter effects, rather than complex interactions such as style-dependent effects 67 . A larger number of samples may overcome some of these limitations and offer more insight into style-specific effects. On the other hand, beer style is not a rigid scientific classification, and beers within one style often differ a lot, which further complicates the analysis of style as a model factor.

Our study is limited to beers from Belgian breweries. Although these beers cover a large portion of the beer styles available globally, some beer styles and consumer patterns may be missing, while other features might be overrepresented. For example, many Belgian ales exhibit yeast-driven flavor profiles, which is reflected in the chemical drivers of appreciation discovered by this study. In future work, expanding the scope to include diverse markets and beer styles could lead to the identification of even more drivers of appreciation and better models for special niche products that were not present in our beer set.

In addition to inherent limitations of GBR models, there are also some limitations associated with studying food aroma. Even if our chemical analyses measured most of the known aroma compounds, the total number of flavor compounds in complex foods like beer is still larger than the subset we were able to measure in this study. For example, hop-derived thiols, that influence flavor at very low concentrations, are notoriously difficult to measure in a high-throughput experiment. Moreover, consumer perception remains subjective and prone to biases that are difficult to avoid. It is also important to stress that the models are still immature and that more extensive datasets will be crucial for developing more complete models in the future. Besides more samples and parameters, our dataset does not include any demographic information about the tasters. Including such data could lead to better models that grasp external factors like age and culture. Another limitation is that our set of beers consists of high-quality end-products and lacks beers that are unfit for sale, which limits the current model in accurately predicting products that are appreciated very badly. Finally, while models could be readily applied in quality control, their use in sensory science and product development is restrained by their inability to discern causal relationships. Given that the models cannot distinguish compounds that genuinely drive consumer perception from those that merely correlate, validation experiments are essential to identify true causative compounds.

Despite the inherent limitations, dissection of our models enabled us to pinpoint specific molecules as potential drivers of beer aroma and consumer appreciation, including compounds that were unexpected and would not have been identified using standard approaches. Important drivers of beer appreciation uncovered by our models include protein levels, ethyl acetate, ethyl phenyl acetate and lactic acid. Currently, many brewers already use lactic acid to acidify their brewing water and ensure optimal pH for enzymatic activity during the mashing process. Our results suggest that adding lactic acid can also improve beer appreciation, although its individual effect remains to be tested. Interestingly, ethanol appears to be unnecessary to improve beer appreciation, both for blond beer and alcohol-free beer. Given the growing consumer interest in alcohol-free beer, with a predicted annual market growth of >7% 84 , it is relevant for brewers to know what compounds can further increase consumer appreciation of these beers. Hence, our model may readily provide avenues to further improve the flavor and consumer appreciation of both alcoholic and non-alcoholic beers, which is generally considered one of the key challenges for future beer production.

Whereas we see a direct implementation of our results for the development of superior alcohol-free beverages and other food products, our study can also serve as a stepping stone for the development of novel alcohol-containing beverages. We want to echo the growing body of scientific evidence for the negative effects of alcohol consumption, both on the individual level by the mutagenic, teratogenic and carcinogenic effects of ethanol 85 , 86 , as well as the burden on society caused by alcohol abuse and addiction. We encourage the use of our results for the production of healthier, tastier products, including novel and improved beverages with lower alcohol contents. Furthermore, we strongly discourage the use of these technologies to improve the appreciation or addictive properties of harmful substances.

The present work demonstrates that despite some important remaining hurdles, combining the latest developments in chemical analyses, sensory analysis and modern machine learning methods offers exciting avenues for food chemistry and engineering. Soon, these tools may provide solutions in quality control and recipe development, as well as new approaches to sensory science and flavor research.

Beer selection

250 commercial Belgian beers were selected to cover the broad diversity of beer styles and corresponding diversity in chemical composition and aroma. See Supplementary Fig.  S1 .

Chemical dataset

Sample preparation.

Beers within their expiration date were purchased from commercial retailers. Samples were prepared in biological duplicates at room temperature, unless explicitly stated otherwise. Bottle pressure was measured with a manual pressure device (Steinfurth Mess-Systeme GmbH) and used to calculate CO 2 concentration. The beer was poured through two filter papers (Macherey-Nagel, 500713032 MN 713 ¼) to remove carbon dioxide and prevent spontaneous foaming. Samples were then prepared for measurements by targeted Headspace-Gas Chromatography-Flame Ionization Detector/Flame Photometric Detector (HS-GC-FID/FPD), Headspace-Solid Phase Microextraction-Gas Chromatography-Mass Spectrometry (HS-SPME-GC-MS), colorimetric analysis, enzymatic analysis, Near-Infrared (NIR) analysis, as described in the sections below. The mean values of biological duplicates are reported for each compound.

HS-GC-FID/FPD

HS-GC-FID/FPD (Shimadzu GC 2010 Plus) was used to measure higher alcohols, acetaldehyde, esters, 4-vinyl guaicol, and sulfur compounds. Each measurement comprised 5 ml of sample pipetted into a 20 ml glass vial containing 1.75 g NaCl (VWR, 27810.295). 100 µl of 2-heptanol (Sigma-Aldrich, H3003) (internal standard) solution in ethanol (Fisher Chemical, E/0650DF/C17) was added for a final concentration of 2.44 mg/L. Samples were flushed with nitrogen for 10 s, sealed with a silicone septum, stored at −80 °C and analyzed in batches of 20.

The GC was equipped with a DB-WAXetr column (length, 30 m; internal diameter, 0.32 mm; layer thickness, 0.50 µm; Agilent Technologies, Santa Clara, CA, USA) to the FID and an HP-5 column (length, 30 m; internal diameter, 0.25 mm; layer thickness, 0.25 µm; Agilent Technologies, Santa Clara, CA, USA) to the FPD. N 2 was used as the carrier gas. Samples were incubated for 20 min at 70 °C in the headspace autosampler (Flow rate, 35 cm/s; Injection volume, 1000 µL; Injection mode, split; Combi PAL autosampler, CTC analytics, Switzerland). The injector, FID and FPD temperatures were kept at 250 °C. The GC oven temperature was first held at 50 °C for 5 min and then allowed to rise to 80 °C at a rate of 5 °C/min, followed by a second ramp of 4 °C/min until 200 °C kept for 3 min and a final ramp of (4 °C/min) until 230 °C for 1 min. Results were analyzed with the GCSolution software version 2.4 (Shimadzu, Kyoto, Japan). The GC was calibrated with a 5% EtOH solution (VWR International) containing the volatiles under study (Supplementary Table  S7 ).

HS-SPME-GC-MS

HS-SPME-GC-MS (Shimadzu GCMS-QP-2010 Ultra) was used to measure additional volatile compounds, mainly comprising terpenoids and esters. Samples were analyzed by HS-SPME using a triphase DVB/Carboxen/PDMS 50/30 μm SPME fiber (Supelco Co., Bellefonte, PA, USA) followed by gas chromatography (Thermo Fisher Scientific Trace 1300 series, USA) coupled to a mass spectrometer (Thermo Fisher Scientific ISQ series MS) equipped with a TriPlus RSH autosampler. 5 ml of degassed beer sample was placed in 20 ml vials containing 1.75 g NaCl (VWR, 27810.295). 5 µl internal standard mix was added, containing 2-heptanol (1 g/L) (Sigma-Aldrich, H3003), 4-fluorobenzaldehyde (1 g/L) (Sigma-Aldrich, 128376), 2,3-hexanedione (1 g/L) (Sigma-Aldrich, 144169) and guaiacol (1 g/L) (Sigma-Aldrich, W253200) in ethanol (Fisher Chemical, E/0650DF/C17). Each sample was incubated at 60 °C in the autosampler oven with constant agitation. After 5 min equilibration, the SPME fiber was exposed to the sample headspace for 30 min. The compounds trapped on the fiber were thermally desorbed in the injection port of the chromatograph by heating the fiber for 15 min at 270 °C.

The GC-MS was equipped with a low polarity RXi-5Sil MS column (length, 20 m; internal diameter, 0.18 mm; layer thickness, 0.18 µm; Restek, Bellefonte, PA, USA). Injection was performed in splitless mode at 320 °C, a split flow of 9 ml/min, a purge flow of 5 ml/min and an open valve time of 3 min. To obtain a pulsed injection, a programmed gas flow was used whereby the helium gas flow was set at 2.7 mL/min for 0.1 min, followed by a decrease in flow of 20 ml/min to the normal 0.9 mL/min. The temperature was first held at 30 °C for 3 min and then allowed to rise to 80 °C at a rate of 7 °C/min, followed by a second ramp of 2 °C/min till 125 °C and a final ramp of 8 °C/min with a final temperature of 270 °C.

Mass acquisition range was 33 to 550 amu at a scan rate of 5 scans/s. Electron impact ionization energy was 70 eV. The interface and ion source were kept at 275 °C and 250 °C, respectively. A mix of linear n-alkanes (from C7 to C40, Supelco Co.) was injected into the GC-MS under identical conditions to serve as external retention index markers. Identification and quantification of the compounds were performed using an in-house developed R script as described in Goelen et al. and Reher et al. 87 , 88 (for package information, see Supplementary Table  S8 ). Briefly, chromatograms were analyzed using AMDIS (v2.71) 89 to separate overlapping peaks and obtain pure compound spectra. The NIST MS Search software (v2.0 g) in combination with the NIST2017, FFNSC3 and Adams4 libraries were used to manually identify the empirical spectra, taking into account the expected retention time. After background subtraction and correcting for retention time shifts between samples run on different days based on alkane ladders, compound elution profiles were extracted and integrated using a file with 284 target compounds of interest, which were either recovered in our identified AMDIS list of spectra or were known to occur in beer. Compound elution profiles were estimated for every peak in every chromatogram over a time-restricted window using weighted non-negative least square analysis after which peak areas were integrated 87 , 88 . Batch effect correction was performed by normalizing against the most stable internal standard compound, 4-fluorobenzaldehyde. Out of all 284 target compounds that were analyzed, 167 were visually judged to have reliable elution profiles and were used for final analysis.

Discrete photometric and enzymatic analysis

Discrete photometric and enzymatic analysis (Thermo Scientific TM Gallery TM Plus Beermaster Discrete Analyzer) was used to measure acetic acid, ammonia, beta-glucan, iso-alpha acids, color, sugars, glycerol, iron, pH, protein, and sulfite. 2 ml of sample volume was used for the analyses. Information regarding the reagents and standard solutions used for analyses and calibrations is included in Supplementary Table  S7 and Supplementary Table  S9 .

NIR analyses

NIR analysis (Anton Paar Alcolyzer Beer ME System) was used to measure ethanol. Measurements comprised 50 ml of sample, and a 10% EtOH solution was used for calibration.

Correlation calculations

Pairwise Spearman Rank correlations were calculated between all chemical properties.

Sensory dataset

Trained panel.

Our trained tasting panel consisted of volunteers who gave prior verbal informed consent. All compounds used for the validation experiment were of food-grade quality. The tasting sessions were approved by the Social and Societal Ethics Committee of the KU Leuven (G-2022-5677-R2(MAR)). All online reviewers agreed to the Terms and Conditions of the RateBeer website.

Sensory analysis was performed according to the American Society of Brewing Chemists (ASBC) Sensory Analysis Methods 90 . 30 volunteers were screened through a series of triangle tests. The sixteen most sensitive and consistent tasters were retained as taste panel members. The resulting panel was diverse in age [22–42, mean: 29], sex [56% male] and nationality [7 different countries]. The panel developed a consensus vocabulary to describe beer aroma, taste and mouthfeel. Panelists were trained to identify and score 50 different attributes, using a 7-point scale to rate attributes’ intensity. The scoring sheet is included as Supplementary Data  3 . Sensory assessments took place between 10–12 a.m. The beers were served in black-colored glasses. Per session, between 5 and 12 beers of the same style were tasted at 12 °C to 16 °C. Two reference beers were added to each set and indicated as ‘Reference 1 & 2’, allowing panel members to calibrate their ratings. Not all panelists were present at every tasting. Scores were scaled by standard deviation and mean-centered per taster. Values are represented as z-scores and clustered by Euclidean distance. Pairwise Spearman correlations were calculated between taste and aroma sensory attributes. Panel consistency was evaluated by repeating samples on different sessions and performing ANOVA to identify differences, using the ‘stats’ package (v4.2.2) in R (for package information, see Supplementary Table  S8 ).

Online reviews from a public database

The ‘scrapy’ package in Python (v3.6) (for package information, see Supplementary Table  S8 ). was used to collect 232,288 online reviews (mean=922, min=6, max=5343) from RateBeer, an online beer review database. Each review entry comprised 5 numerical scores (appearance, aroma, taste, palate and overall quality) and an optional review text. The total number of reviews per reviewer was collected separately. Numerical scores were scaled and centered per rater, and mean scores were calculated per beer.

For the review texts, the language was estimated using the packages ‘langdetect’ and ‘langid’ in Python. Reviews that were classified as English by both packages were kept. Reviewers with fewer than 100 entries overall were discarded. 181,025 reviews from >6000 reviewers from >40 countries remained. Text processing was done using the ‘nltk’ package in Python. Texts were corrected for slang and misspellings; proper nouns and rare words that are relevant to the beer context were specified and kept as-is (‘Chimay’,’Lambic’, etc.). A dictionary of semantically similar sensorial terms, for example ‘floral’ and ‘flower’, was created and collapsed together into one term. Words were stemmed and lemmatized to avoid identifying words such as ‘acid’ and ‘acidity’ as separate terms. Numbers and punctuation were removed.

Sentences from up to 50 randomly chosen reviews per beer were manually categorized according to the aspect of beer they describe (appearance, aroma, taste, palate, overall quality—not to be confused with the 5 numerical scores described above) or flagged as irrelevant if they contained no useful information. If a beer contained fewer than 50 reviews, all reviews were manually classified. This labeled data set was used to train a model that classified the rest of the sentences for all beers 91 . Sentences describing taste and aroma were extracted, and term frequency–inverse document frequency (TFIDF) was implemented to calculate enrichment scores for sensorial words per beer.

The sex of the tasting subject was not considered when building our sensory database. Instead, results from different panelists were averaged, both for our trained panel (56% male, 44% female) and the RateBeer reviews (70% male, 30% female for RateBeer as a whole).

Beer price collection and processing

Beer prices were collected from the following stores: Colruyt, Delhaize, Total Wine, BeerHawk, The Belgian Beer Shop, The Belgian Shop, and Beer of Belgium. Where applicable, prices were converted to Euros and normalized per liter. Spearman correlations were calculated between these prices and mean overall appreciation scores from RateBeer and the taste panel, respectively.

Pairwise Spearman Rank correlations were calculated between all sensory properties.

Machine learning models

Predictive modeling of sensory profiles from chemical data.

Regression models were constructed to predict (a) trained panel scores for beer flavors and quality from beer chemical profiles and (b) public reviews’ appreciation scores from beer chemical profiles. Z-scores were used to represent sensory attributes in both data sets. Chemical properties with log-normal distributions (Shapiro-Wilk test, p  <  0.05 ) were log-transformed. Missing chemical measurements (0.1% of all data) were replaced with mean values per attribute. Observations from 250 beers were randomly separated into a training set (70%, 175 beers) and a test set (30%, 75 beers), stratified per beer style. Chemical measurements (p = 231) were normalized based on the training set average and standard deviation. In total, three linear regression-based models: linear regression with first-order interaction terms (LR), lasso regression with first-order interaction terms (Lasso) and partial least squares regression (PLSR); five decision tree models, Adaboost regressor (ABR), Extra Trees (ET), Gradient Boosting regressor (GBR), Random Forest (RF) and XGBoost regressor (XGBR); one support vector machine model (SVR) and one artificial neural network model (ANN) were trained. The models were implemented using the ‘scikit-learn’ package (v1.2.2) and ‘xgboost’ package (v1.7.3) in Python (v3.9.16). Models were trained, and hyperparameters optimized, using five-fold cross-validated grid search with the coefficient of determination (R 2 ) as the evaluation metric. The ANN (scikit-learn’s MLPRegressor) was optimized using Bayesian Tree-Structured Parzen Estimator optimization with the ‘Optuna’ Python package (v3.2.0). Individual models were trained per attribute, and a multi-output model was trained on all attributes simultaneously.

Model dissection

GBR was found to outperform other methods, resulting in models with the highest average R 2 values in both trained panel and public review data sets. Impurity-based rankings of the most important predictors for each predicted sensorial trait were obtained using the ‘scikit-learn’ package. To observe the relationships between these chemical properties and their predicted targets, partial dependence plots (PDP) were constructed for the six most important predictors of consumer appreciation 74 , 75 .

The ‘SHAP’ package in Python (v0.41.0) was implemented to provide an alternative ranking of predictor importance and to visualize the predictors’ effects as a function of their concentration 68 .

Validation of causal chemical properties

To validate the effects of the most important model features on predicted sensory attributes, beers were spiked with the chemical compounds identified by the models and descriptive sensory analyses were carried out according to the American Society of Brewing Chemists (ASBC) protocol 90 .

Compound spiking was done 30 min before tasting. Compounds were spiked into fresh beer bottles, that were immediately resealed and inverted three times. Fresh bottles of beer were opened for the same duration, resealed, and inverted thrice, to serve as controls. Pairs of spiked samples and controls were served simultaneously, chilled and in dark glasses as outlined in the Trained panel section above. Tasters were instructed to select the glass with the higher flavor intensity for each attribute (directional difference test 92 ) and to select the glass they prefer.

The final concentration after spiking was equal to the within-style average, after normalizing by ethanol concentration. This was done to ensure balanced flavor profiles in the final spiked beer. The same methods were applied to improve a non-alcoholic beer. Compounds were the following: ethyl acetate (Merck KGaA, W241415), ethyl hexanoate (Merck KGaA, W243906), isoamyl acetate (Merck KGaA, W205508), phenethyl acetate (Merck KGaA, W285706), ethanol (96%, Colruyt), glycerol (Merck KGaA, W252506), lactic acid (Merck KGaA, 261106).

Significant differences in preference or perceived intensity were determined by performing the two-sided binomial test on each attribute.

Reporting summary

Further information on research design is available in the  Nature Portfolio Reporting Summary linked to this article.

Data availability

The data that support the findings of this work are available in the Supplementary Data files and have been deposited to Zenodo under accession code 10653704 93 . The RateBeer scores data are under restricted access, they are not publicly available as they are property of RateBeer (ZX Ventures, USA). Access can be obtained from the authors upon reasonable request and with permission of RateBeer (ZX Ventures, USA).  Source data are provided with this paper.

Code availability

The code for training the machine learning models, analyzing the models, and generating the figures has been deposited to Zenodo under accession code 10653704 93 .

Tieman, D. et al. A chemical genetic roadmap to improved tomato flavor. Science 355 , 391–394 (2017).

Article   ADS   CAS   PubMed   Google Scholar  

Plutowska, B. & Wardencki, W. Application of gas chromatography–olfactometry (GC–O) in analysis and quality assessment of alcoholic beverages – A review. Food Chem. 107 , 449–463 (2008).

Article   CAS   Google Scholar  

Legin, A., Rudnitskaya, A., Seleznev, B. & Vlasov, Y. Electronic tongue for quality assessment of ethanol, vodka and eau-de-vie. Anal. Chim. Acta 534 , 129–135 (2005).

Loutfi, A., Coradeschi, S., Mani, G. K., Shankar, P. & Rayappan, J. B. B. Electronic noses for food quality: A review. J. Food Eng. 144 , 103–111 (2015).

Ahn, Y.-Y., Ahnert, S. E., Bagrow, J. P. & Barabási, A.-L. Flavor network and the principles of food pairing. Sci. Rep. 1 , 196 (2011).

Article   CAS   PubMed   PubMed Central   Google Scholar  

Bartoshuk, L. M. & Klee, H. J. Better fruits and vegetables through sensory analysis. Curr. Biol. 23 , R374–R378 (2013).

Article   CAS   PubMed   Google Scholar  

Piggott, J. R. Design questions in sensory and consumer science. Food Qual. Prefer. 3293 , 217–220 (1995).

Article   Google Scholar  

Kermit, M. & Lengard, V. Assessing the performance of a sensory panel-panellist monitoring and tracking. J. Chemom. 19 , 154–161 (2005).

Cook, D. J., Hollowood, T. A., Linforth, R. S. T. & Taylor, A. J. Correlating instrumental measurements of texture and flavour release with human perception. Int. J. Food Sci. Technol. 40 , 631–641 (2005).

Chinchanachokchai, S., Thontirawong, P. & Chinchanachokchai, P. A tale of two recommender systems: The moderating role of consumer expertise on artificial intelligence based product recommendations. J. Retail. Consum. Serv. 61 , 1–12 (2021).

Ross, C. F. Sensory science at the human-machine interface. Trends Food Sci. Technol. 20 , 63–72 (2009).

Chambers, E. IV & Koppel, K. Associations of volatile compounds with sensory aroma and flavor: The complex nature of flavor. Molecules 18 , 4887–4905 (2013).

Pinu, F. R. Metabolomics—The new frontier in food safety and quality research. Food Res. Int. 72 , 80–81 (2015).

Danezis, G. P., Tsagkaris, A. S., Brusic, V. & Georgiou, C. A. Food authentication: state of the art and prospects. Curr. Opin. Food Sci. 10 , 22–31 (2016).

Shepherd, G. M. Smell images and the flavour system in the human brain. Nature 444 , 316–321 (2006).

Meilgaard, M. C. Prediction of flavor differences between beers from their chemical composition. J. Agric. Food Chem. 30 , 1009–1017 (1982).

Xu, L. et al. Widespread receptor-driven modulation in peripheral olfactory coding. Science 368 , eaaz5390 (2020).

Kupferschmidt, K. Following the flavor. Science 340 , 808–809 (2013).

Billesbølle, C. B. et al. Structural basis of odorant recognition by a human odorant receptor. Nature 615 , 742–749 (2023).

Article   ADS   PubMed   PubMed Central   Google Scholar  

Smith, B. Perspective: Complexities of flavour. Nature 486 , S6–S6 (2012).

Pfister, P. et al. Odorant receptor inhibition is fundamental to odor encoding. Curr. Biol. 30 , 2574–2587 (2020).

Moskowitz, H. W., Kumaraiah, V., Sharma, K. N., Jacobs, H. L. & Sharma, S. D. Cross-cultural differences in simple taste preferences. Science 190 , 1217–1218 (1975).

Eriksson, N. et al. A genetic variant near olfactory receptor genes influences cilantro preference. Flavour 1 , 22 (2012).

Ferdenzi, C. et al. Variability of affective responses to odors: Culture, gender, and olfactory knowledge. Chem. Senses 38 , 175–186 (2013).

Article   PubMed   Google Scholar  

Lawless, H. T. & Heymann, H. Sensory evaluation of food: Principles and practices. (Springer, New York, NY). https://doi.org/10.1007/978-1-4419-6488-5 (2010).

Colantonio, V. et al. Metabolomic selection for enhanced fruit flavor. Proc. Natl. Acad. Sci. 119 , e2115865119 (2022).

Fritz, F., Preissner, R. & Banerjee, P. VirtualTaste: a web server for the prediction of organoleptic properties of chemical compounds. Nucleic Acids Res 49 , W679–W684 (2021).

Tuwani, R., Wadhwa, S. & Bagler, G. BitterSweet: Building machine learning models for predicting the bitter and sweet taste of small molecules. Sci. Rep. 9 , 1–13 (2019).

Dagan-Wiener, A. et al. Bitter or not? BitterPredict, a tool for predicting taste from chemical structure. Sci. Rep. 7 , 1–13 (2017).

Pallante, L. et al. Toward a general and interpretable umami taste predictor using a multi-objective machine learning approach. Sci. Rep. 12 , 1–11 (2022).

Malavolta, M. et al. A survey on computational taste predictors. Eur. Food Res. Technol. 248 , 2215–2235 (2022).

Lee, B. K. et al. A principal odor map unifies diverse tasks in olfactory perception. Science 381 , 999–1006 (2023).

Mayhew, E. J. et al. Transport features predict if a molecule is odorous. Proc. Natl. Acad. Sci. 119 , e2116576119 (2022).

Niu, Y. et al. Sensory evaluation of the synergism among ester odorants in light aroma-type liquor by odor threshold, aroma intensity and flash GC electronic nose. Food Res. Int. 113 , 102–114 (2018).

Yu, P., Low, M. Y. & Zhou, W. Design of experiments and regression modelling in food flavour and sensory analysis: A review. Trends Food Sci. Technol. 71 , 202–215 (2018).

Oladokun, O. et al. The impact of hop bitter acid and polyphenol profiles on the perceived bitterness of beer. Food Chem. 205 , 212–220 (2016).

Linforth, R., Cabannes, M., Hewson, L., Yang, N. & Taylor, A. Effect of fat content on flavor delivery during consumption: An in vivo model. J. Agric. Food Chem. 58 , 6905–6911 (2010).

Guo, S., Na Jom, K. & Ge, Y. Influence of roasting condition on flavor profile of sunflower seeds: A flavoromics approach. Sci. Rep. 9 , 11295 (2019).

Ren, Q. et al. The changes of microbial community and flavor compound in the fermentation process of Chinese rice wine using Fagopyrum tataricum grain as feedstock. Sci. Rep. 9 , 3365 (2019).

Hastie, T., Friedman, J. & Tibshirani, R. The Elements of Statistical Learning. (Springer, New York, NY). https://doi.org/10.1007/978-0-387-21606-5 (2001).

Dietz, C., Cook, D., Huismann, M., Wilson, C. & Ford, R. The multisensory perception of hop essential oil: a review. J. Inst. Brew. 126 , 320–342 (2020).

CAS   Google Scholar  

Roncoroni, Miguel & Verstrepen, Kevin Joan. Belgian Beer: Tested and Tasted. (Lannoo, 2018).

Meilgaard, M. Flavor chemistry of beer: Part II: Flavor and threshold of 239 aroma volatiles. in (1975).

Bokulich, N. A. & Bamforth, C. W. The microbiology of malting and brewing. Microbiol. Mol. Biol. Rev. MMBR 77 , 157–172 (2013).

Dzialo, M. C., Park, R., Steensels, J., Lievens, B. & Verstrepen, K. J. Physiology, ecology and industrial applications of aroma formation in yeast. FEMS Microbiol. Rev. 41 , S95–S128 (2017).

Article   PubMed   PubMed Central   Google Scholar  

Datta, A. et al. Computer-aided food engineering. Nat. Food 3 , 894–904 (2022).

American Society of Brewing Chemists. Beer Methods. (American Society of Brewing Chemists, St. Paul, MN, U.S.A.).

Olaniran, A. O., Hiralal, L., Mokoena, M. P. & Pillay, B. Flavour-active volatile compounds in beer: production, regulation and control. J. Inst. Brew. 123 , 13–23 (2017).

Verstrepen, K. J. et al. Flavor-active esters: Adding fruitiness to beer. J. Biosci. Bioeng. 96 , 110–118 (2003).

Meilgaard, M. C. Flavour chemistry of beer. part I: flavour interaction between principal volatiles. Master Brew. Assoc. Am. Tech. Q 12 , 107–117 (1975).

Briggs, D. E., Boulton, C. A., Brookes, P. A. & Stevens, R. Brewing 227–254. (Woodhead Publishing). https://doi.org/10.1533/9781855739062.227 (2004).

Bossaert, S., Crauwels, S., De Rouck, G. & Lievens, B. The power of sour - A review: Old traditions, new opportunities. BrewingScience 72 , 78–88 (2019).

Google Scholar  

Verstrepen, K. J. et al. Flavor active esters: Adding fruitiness to beer. J. Biosci. Bioeng. 96 , 110–118 (2003).

Snauwaert, I. et al. Microbial diversity and metabolite composition of Belgian red-brown acidic ales. Int. J. Food Microbiol. 221 , 1–11 (2016).

Spitaels, F. et al. The microbial diversity of traditional spontaneously fermented lambic beer. PLoS ONE 9 , e95384 (2014).

Blanco, C. A., Andrés-Iglesias, C. & Montero, O. Low-alcohol Beers: Flavor Compounds, Defects, and Improvement Strategies. Crit. Rev. Food Sci. Nutr. 56 , 1379–1388 (2016).

Jackowski, M. & Trusek, A. Non-Alcohol. beer Prod. – Overv. 20 , 32–38 (2018).

Takoi, K. et al. The contribution of geraniol metabolism to the citrus flavour of beer: Synergy of geraniol and β-citronellol under coexistence with excess linalool. J. Inst. Brew. 116 , 251–260 (2010).

Kroeze, J. H. & Bartoshuk, L. M. Bitterness suppression as revealed by split-tongue taste stimulation in humans. Physiol. Behav. 35 , 779–783 (1985).

Mennella, J. A. et al. A spoonful of sugar helps the medicine go down”: Bitter masking bysucrose among children and adults. Chem. Senses 40 , 17–25 (2015).

Wietstock, P., Kunz, T., Perreira, F. & Methner, F.-J. Metal chelation behavior of hop acids in buffered model systems. BrewingScience 69 , 56–63 (2016).

Sancho, D., Blanco, C. A., Caballero, I. & Pascual, A. Free iron in pale, dark and alcohol-free commercial lager beers. J. Sci. Food Agric. 91 , 1142–1147 (2011).

Rodrigues, H. & Parr, W. V. Contribution of cross-cultural studies to understanding wine appreciation: A review. Food Res. Int. 115 , 251–258 (2019).

Korneva, E. & Blockeel, H. Towards better evaluation of multi-target regression models. in ECML PKDD 2020 Workshops (eds. Koprinska, I. et al.) 353–362 (Springer International Publishing, Cham, 2020). https://doi.org/10.1007/978-3-030-65965-3_23 .

Gastón Ares. Mathematical and Statistical Methods in Food Science and Technology. (Wiley, 2013).

Grinsztajn, L., Oyallon, E. & Varoquaux, G. Why do tree-based models still outperform deep learning on tabular data? Preprint at http://arxiv.org/abs/2207.08815 (2022).

Gries, S. T. Statistics for Linguistics with R: A Practical Introduction. in Statistics for Linguistics with R (De Gruyter Mouton, 2021). https://doi.org/10.1515/9783110718256 .

Lundberg, S. M. et al. From local explanations to global understanding with explainable AI for trees. Nat. Mach. Intell. 2 , 56–67 (2020).

Ickes, C. M. & Cadwallader, K. R. Effects of ethanol on flavor perception in alcoholic beverages. Chemosens. Percept. 10 , 119–134 (2017).

Kato, M. et al. Influence of high molecular weight polypeptides on the mouthfeel of commercial beer. J. Inst. Brew. 127 , 27–40 (2021).

Wauters, R. et al. Novel Saccharomyces cerevisiae variants slow down the accumulation of staling aldehydes and improve beer shelf-life. Food Chem. 398 , 1–11 (2023).

Li, H., Jia, S. & Zhang, W. Rapid determination of low-level sulfur compounds in beer by headspace gas chromatography with a pulsed flame photometric detector. J. Am. Soc. Brew. Chem. 66 , 188–191 (2008).

Dercksen, A., Laurens, J., Torline, P., Axcell, B. C. & Rohwer, E. Quantitative analysis of volatile sulfur compounds in beer using a membrane extraction interface. J. Am. Soc. Brew. Chem. 54 , 228–233 (1996).

Molnar, C. Interpretable Machine Learning: A Guide for Making Black-Box Models Interpretable. (2020).

Zhao, Q. & Hastie, T. Causal interpretations of black-box models. J. Bus. Econ. Stat. Publ. Am. Stat. Assoc. 39 , 272–281 (2019).

Article   MathSciNet   Google Scholar  

Hastie, T., Tibshirani, R. & Friedman, J. The Elements of Statistical Learning. (Springer, 2019).

Labrado, D. et al. Identification by NMR of key compounds present in beer distillates and residual phases after dealcoholization by vacuum distillation. J. Sci. Food Agric. 100 , 3971–3978 (2020).

Lusk, L. T., Kay, S. B., Porubcan, A. & Ryder, D. S. Key olfactory cues for beer oxidation. J. Am. Soc. Brew. Chem. 70 , 257–261 (2012).

Gonzalez Viejo, C., Torrico, D. D., Dunshea, F. R. & Fuentes, S. Development of artificial neural network models to assess beer acceptability based on sensory properties using a robotic pourer: A comparative model approach to achieve an artificial intelligence system. Beverages 5 , 33 (2019).

Gonzalez Viejo, C., Fuentes, S., Torrico, D. D., Godbole, A. & Dunshea, F. R. Chemical characterization of aromas in beer and their effect on consumers liking. Food Chem. 293 , 479–485 (2019).

Gilbert, J. L. et al. Identifying breeding priorities for blueberry flavor using biochemical, sensory, and genotype by environment analyses. PLOS ONE 10 , 1–21 (2015).

Goulet, C. et al. Role of an esterase in flavor volatile variation within the tomato clade. Proc. Natl. Acad. Sci. 109 , 19009–19014 (2012).

Article   ADS   CAS   PubMed   PubMed Central   Google Scholar  

Borisov, V. et al. Deep Neural Networks and Tabular Data: A Survey. IEEE Trans. Neural Netw. Learn. Syst. 1–21 https://doi.org/10.1109/TNNLS.2022.3229161 (2022).

Statista. Statista Consumer Market Outlook: Beer - Worldwide.

Seitz, H. K. & Stickel, F. Molecular mechanisms of alcoholmediated carcinogenesis. Nat. Rev. Cancer 7 , 599–612 (2007).

Voordeckers, K. et al. Ethanol exposure increases mutation rate through error-prone polymerases. Nat. Commun. 11 , 3664 (2020).

Goelen, T. et al. Bacterial phylogeny predicts volatile organic compound composition and olfactory response of an aphid parasitoid. Oikos 129 , 1415–1428 (2020).

Article   ADS   Google Scholar  

Reher, T. et al. Evaluation of hop (Humulus lupulus) as a repellent for the management of Drosophila suzukii. Crop Prot. 124 , 104839 (2019).

Stein, S. E. An integrated method for spectrum extraction and compound identification from gas chromatography/mass spectrometry data. J. Am. Soc. Mass Spectrom. 10 , 770–781 (1999).

American Society of Brewing Chemists. Sensory Analysis Methods. (American Society of Brewing Chemists, St. Paul, MN, U.S.A., 1992).

McAuley, J., Leskovec, J. & Jurafsky, D. Learning Attitudes and Attributes from Multi-Aspect Reviews. Preprint at https://doi.org/10.48550/arXiv.1210.3926 (2012).

Meilgaard, M. C., Carr, B. T. & Carr, B. T. Sensory Evaluation Techniques. (CRC Press, Boca Raton). https://doi.org/10.1201/b16452 (2014).

Schreurs, M. et al. Data from: Predicting and improving complex beer flavor through machine learning. Zenodo https://doi.org/10.5281/zenodo.10653704 (2024).

Download references

Acknowledgements

We thank all lab members for their discussions and thank all tasting panel members for their contributions. Special thanks go out to Dr. Karin Voordeckers for her tremendous help in proofreading and improving the manuscript. M.S. was supported by a Baillet-Latour fellowship, L.C. acknowledges financial support from KU Leuven (C16/17/006), F.A.T. was supported by a PhD fellowship from FWO (1S08821N). Research in the lab of K.J.V. is supported by KU Leuven, FWO, VIB, VLAIO and the Brewing Science Serves Health Fund. Research in the lab of T.W. is supported by FWO (G.0A51.15) and KU Leuven (C16/17/006).

Author information

These authors contributed equally: Michiel Schreurs, Supinya Piampongsant, Miguel Roncoroni.

Authors and Affiliations

VIB—KU Leuven Center for Microbiology, Gaston Geenslaan 1, B-3001, Leuven, Belgium

Michiel Schreurs, Supinya Piampongsant, Miguel Roncoroni, Lloyd Cool, Beatriz Herrera-Malaver, Florian A. Theßeling & Kevin J. Verstrepen

CMPG Laboratory of Genetics and Genomics, KU Leuven, Gaston Geenslaan 1, B-3001, Leuven, Belgium

Leuven Institute for Beer Research (LIBR), Gaston Geenslaan 1, B-3001, Leuven, Belgium

Laboratory of Socioecology and Social Evolution, KU Leuven, Naamsestraat 59, B-3000, Leuven, Belgium

Lloyd Cool, Christophe Vanderaa & Tom Wenseleers

VIB Bioinformatics Core, VIB, Rijvisschestraat 120, B-9052, Ghent, Belgium

Łukasz Kreft & Alexander Botzki

AB InBev SA/NV, Brouwerijplein 1, B-3000, Leuven, Belgium

Philippe Malcorps & Luk Daenen

You can also search for this author in PubMed   Google Scholar

Contributions

S.P., M.S. and K.J.V. conceived the experiments. S.P., M.S. and K.J.V. designed the experiments. S.P., M.S., M.R., B.H. and F.A.T. performed the experiments. S.P., M.S., L.C., C.V., L.K., A.B., P.M., L.D., T.W. and K.J.V. contributed analysis ideas. S.P., M.S., L.C., C.V., T.W. and K.J.V. analyzed the data. All authors contributed to writing the manuscript.

Corresponding author

Correspondence to Kevin J. Verstrepen .

Ethics declarations

Competing interests.

K.J.V. is affiliated with bar.on. The other authors declare no competing interests.

Peer review

Peer review information.

Nature Communications thanks Florian Bauer, Andrew John Macintosh and the other, anonymous, reviewer(s) for their contribution to the peer review of this work. A peer review file is available.

Additional information

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary information

Supplementary information, peer review file, description of additional supplementary files, supplementary data 1, supplementary data 2, supplementary data 3, supplementary data 4, supplementary data 5, supplementary data 6, supplementary data 7, reporting summary, source data, source data, rights and permissions.

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/ .

Reprints and permissions

About this article

Cite this article.

Schreurs, M., Piampongsant, S., Roncoroni, M. et al. Predicting and improving complex beer flavor through machine learning. Nat Commun 15 , 2368 (2024). https://doi.org/10.1038/s41467-024-46346-0

Download citation

Received : 30 October 2023

Accepted : 21 February 2024

Published : 26 March 2024

DOI : https://doi.org/10.1038/s41467-024-46346-0

Share this article

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

By submitting a comment you agree to abide by our Terms and Community Guidelines . If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.

Quick links

• Explore articles by subject
• Guide to authors
• Editorial policies

Sign up for the Nature Briefing: Translational Research newsletter — top stories in biotechnology, drug discovery and pharma.

Read our research on: Abortion | Podcasts | Election 2024

Regions & Countries

8 in 10 americans say religion is losing influence in public life, few see biden or trump as especially religious.

Pew Research Center conducted this survey to explore Americans’ attitudes about religion’s role in public life, including politics in a presidential election year.

For this report, we surveyed 12,693 respondents from Feb. 13 to 25, 2024. Most of the respondents (10,642) are members of the American Trends Panel, an online survey panel recruited through national random sampling of residential addresses, which gives nearly all U.S. adults a chance of selection.

The remaining respondents (2,051) are members of three other panels, the Ipsos KnowledgePanel, the NORC Amerispeak panel and the SSRS opinion panel. All three are national survey panels recruited through random sampling (not “opt-in” polls). We used these additional panels to ensure that the survey would have enough Jewish and Muslim respondents to be able to report on their views.

The survey is weighted to be representative of the U.S. adult population by gender, race, ethnicity, partisan affiliation, education, religious affiliation and other categories.

For more, refer to the ATP’s Methodology and the Methodology for this report. Read the questions used in this report .

A new Pew Research Center survey finds that 80% of U.S. adults say religion’s role in American life is shrinking – a percentage that’s as high as it’s ever been in our surveys.

Most Americans who say religion’s influence is shrinking are not happy about it. Overall, 49% of U.S. adults say both that religion is losing influence and that this is a bad thing. An additional 8% of U.S. adults think religion’s influence is growing and that this is a good thing.

Together, a combined 57% of U.S adults – a clear majority – express a positive view of religion’s influence on American life.

The survey also finds that about half of U.S. adults say it’s “very” or “somewhat” important to them to have a president who has strong religious beliefs, even if those beliefs are different from their own. But relatively few Americans view either of the leading presidential candidates as very religious: 13% of Americans say they think President Joe Biden is very religious, and just 4% say this about former President Donald Trump.

Overall, there are widespread signs of unease with religion’s trajectory in American life. This dissatisfaction is not just among religious Americans. Rather, many religious and nonreligious Americans say they feel that their religious beliefs put them at odds with mainstream culture, with the people around them and with the other side of the political spectrum. For example:

• 48% of U.S. adults say there’s “a great deal” of or “some” conflict between their religious beliefs and mainstream American culture, up from 42% in 2020.
• 29% say they think of themselves as religious minorities, up from 24% in 2020.
• 41% say it’s best to avoid discussing religion at all if someone disagrees with you, up from 33% in 2019.
• 72% of religiously unaffiliated adults – those who identify, religiously, as atheist, agnostic or “nothing in particular” – say conservative Christians have gone too far in trying to control religion in the government and public schools; 63% of Christians say the same about secular liberals.

These are among the key findings of a new Pew Research Center survey, conducted Feb. 13-25, 2024, among a nationally representative sample of 12,693 U.S. adults.

This report examines:

• Religion’s role in public life
• U.S. presidential candidates and their religious engagement
• Christianity’s place in politics, and “Christian nationalism”

The survey also finds wide partisan gaps on questions about the proper role for religion in society, with Republicans more likely than Democrats to favor religious influence in governance and public life. For instance:

• 42% of Republicans and Republican-leaning independents say that when the Bible and the will of the people conflict, the Bible should have more influence on U.S. laws than the will of the people. Just 16% of Democrats and Democratic-leaning independents say this.
• 21% of Republicans and GOP leaners say the federal government should declare Christianity the official religion of the United States, compared with 7% of Democrats and Democratic leaners.

Moral and religious qualities in a president

Almost all Americans (94%) say it is “very” or “somewhat” important to have a president who personally lives a moral and ethical life. And a majority (64%) say it’s important to have a president who stands up for people with their religious beliefs.

About half of U.S. adults (48%) say it is important for the president to hold strong religious beliefs. Fewer (37%) say it’s important for the president to have the same religious beliefs as their own.

Republicans are much more likely than Democrats to value religious qualities in a president, and Christians are more likely than the religiously unaffiliated to do so. For example:

• Republicans and GOP leaners are twice as likely as Democrats and Democratic leaners to say it is important to have a president who has the same religious beliefs they do (51% vs. 25%).
• 70% of White evangelical Protestants say it is important to have a president who shares their religious beliefs. Just 11% of religiously unaffiliated Americans say this.

Views of Biden, Trump and their religious engagement

Relatively few Americans think of Biden or Trump as “very” religious. Indeed, even most Republicans don’t think Trump is very religious, and even most Democrats don’t think Biden is very religious.

• 6% of Republicans and GOP leaners say Trump is very religious, while 44% say he is “somewhat” religious. Nearly half (48%) say he is “not too” or “not at all” religious.
• 23% of Democrats and Democratic-leaning independents say Biden is very religious, while 55% say he is somewhat religious. And 21% say he is not too or not at all religious.

Though they don’t think Trump is very religious himself, most Republicans and people in religious groups that tend to favor the Republican Party do think he stands up at least to some extent for people with their religious beliefs. Two-thirds of Republicans and independents who lean toward the GOP (67%) say Trump stands up for people with their religious beliefs “a great deal,” “quite a bit” or “some.” About the same share of White evangelical Protestants (69%) say this about Trump.

Similarly, 60% of Democrats and Democratic-leaning independents, as well as 73% of Jewish Americans and 60% of Black Protestants, say Biden stands up for people with their religious beliefs a great deal, quite a bit or some.

Overall, views of both Trump and Biden are generally unfavorable.

• White evangelical Protestants – a largely Republican group – stand out as having particularly favorable views of Trump (67%) and unfavorable views of Biden (86%).
• Black Protestants and Jewish Americans – largely Democratic groups – stand out for having favorable views of Biden and unfavorable views of Trump.

Views on trying to control religious values in the government and schools

Americans are almost equally split on whether conservative Christians have gone too far in trying to push their religious values in the government and public schools, as well as on whether secular liberals have gone too far in trying to keep religious values out of these institutions.

Most religiously unaffiliated Americans (72%) and Democrats (72%) say conservative Christians have gone too far. And most Christians (63%) and Republicans (76%) say secular liberals have gone too far.

Christianity’s place in politics, and Christian nationalism

In recent years, “Christian nationalism” has received a great deal of attention as an ideology that some critics have said could threaten American democracy .

Despite growing news coverage of Christian nationalism – including reports of political leaders who seem to endorse the concept – the new survey shows that there has been no change in the share of Americans who have heard of Christian nationalism over the past year and a half. Similarly, the new survey finds no change in how favorably U.S. adults view Christian nationalism.

Overall, 45% say they have heard or read about Christian nationalism, including 25% who also have an unfavorable view of it and 5% who have a favorable view of it. Meanwhile, 54% of Americans say they haven’t heard of Christian nationalism at all.

One element often associated with Christian nationalism is the idea that church and state should not be separated, despite the Establishment Clause in the First Amendment to the U.S. Constitution.

The survey finds that about half of Americans (49%) say the Bible should have “a great deal” of or “some” influence on U.S. laws, while another half (51%) say it should have “not much” or “no influence.” And 28% of U.S. adults say the Bible should have more influence than the will of the people if the two conflict. These numbers have remained virtually unchanged over the past four years.

In the new survey, 16% of U.S. adults say the government should stop enforcing the separation of church and state. This is little changed since 2021.

In response to a separate question, 13% of U.S. adults say the federal government should declare Christianity the official religion of the U.S., and 44% say the government should not declare the country a Christian nation but should promote Christian moral values. Meanwhile, 39% say the government should not elevate Christianity in either way. 1

Overall, 3% of U.S. adults say the Bible should have more influence on U.S. laws than the will of the people; and that the government should stop enforcing separation of church and state; and that Christianity should be declared the country’s official religion. And 13% of U.S. adults endorse two of these three statements. Roughly one-fifth of the public (22%) expresses one of these three views that are often associated with Christian nationalism. The majority (62%) expresses none.

Guide to this report

The remainder of this report describes these findings in additional detail.  Chapter 1  focuses on the public’s perceptions of religion’s role in public life. Chapter 2  examines views of presidential candidates and their religious engagement. And  Chapter 3  focuses on Christian nationalism and views of the U.S. as a Christian nation.

• The share saying that the government should declare Christianity the official national religion (13%) is almost identical to the share who said the government should declare the U.S. a Christian nation in a March 2021 survey that asked a similar question (15%). ↩

Sign up for our Religion newsletter

Sent weekly on Wednesday

Report Materials

Table of contents, 5 facts about religion and americans’ views of donald trump, u.s. christians more likely than ‘nones’ to say situation at the border is a crisis, from businesses and banks to colleges and churches: americans’ views of u.s. institutions, most u.s. parents pass along their religion and politics to their children, growing share of americans see the supreme court as ‘friendly’ toward religion, most popular.

About Pew Research Center Pew Research Center is a nonpartisan fact tank that informs the public about the issues, attitudes and trends shaping the world. It conducts public opinion polling, demographic research, media content analysis and other empirical social science research. Pew Research Center does not take policy positions. It is a subsidiary of The Pew Charitable Trusts .

Search code, repositories, users, issues, pull requests...

Provide feedback.

We read every piece of feedback, and take your input very seriously.

Saved searches

Use saved searches to filter your results more quickly.

To see all available qualifiers, see our documentation .

• Notifications

T-Rex2: Towards Generic Object Detection via Text-Visual Prompt Synergy

IDEA-Research/T-Rex

Contributors 4.

• Python 100.0%