Bayesian Model Selection in Social Research
The DOI matches its Crossref record; 44 of the 44 references it cites were found in a registry or an index.
Abstract
It is argued that P-values and the tests based upon them give unsatisfactory results, especially in large samples. It is shown that, in regression, when there are many candidate independent variables, standard variable selection procedures can give very misleading results. Also, by selecting a single model, they ignore model uncertainty and so underestimate the uncertainty about quantities of interest. The Bayesian approach to hypothesis testing, model selection, and accounting for model uncertainty is presented. Implementing this is straightforward through the use of the simple and accurate BIC approximation, and it can be done using the output from standard software. Specific results are presented for most of the types of model commonly used in sociology. It is shown that this approach overcomes the difficulties with P-values and standard model selection procedures based on them. It also allows easy comparison of nonnested models, and permits the quantification of the evidence for a null hypothesis of interest, such as a convergence theory or a hypothesis about societal norms.
Overview
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
In the citing paper
“The Laplace method (Tierney and Kadane, 1986) can provide an excellent approximation to pr lparenOSCASB D vertSHT Mk rparenOSCASB ; in certain circumstances this yields the very simple BIC approximation (Schwarz, 1978; Kass and Wasserman 1995; Raftery, 1995).”
From Bayesian model averaging: a tutorial (with comments by M. Clyde, David Draper and E. I. George, and a rejoinder by the authors (1999) · cites this paper as Raftery, 1995
In the citing paper
“Under regularity conditions, this yields the simple BIC approximation given by equation (13) in our paper (Kass and Wasserman, 1995; Raftery, 1995).”
From Bayesian model averaging: a tutorial (with comments by M. Clyde, David Draper and E. I. George, and a rejoinder by the authors (1999) · cites this paper as Raftery, 1995
In the citing paper
“The limitations of NHST methods are well documented and need not be rehashed here in their entirety”
From Reevaluating the effectiveness of n-back training on transfer through the Bayesian lens: Support for the null (2016) · cites this paper as Raftery (1995)
In the citing paper
“An advantage of the Bayesian approach is that it provides an opportunity to analytically determine whether the null hypothesis is more likely than the alternative(s) given the observed data (e.g., Raftery, 1995; Gallistel, 2009).”
From Cognitive Mapping Without Vision: Comparing Wayfinding Performance After Learning From Digital Touchscreen-Based Multimodal Maps vs. Embossed Tactile Overlays (2020) · cites this paper as Raftery, 1995
In the citing paper
“A change in 10 points or more suggests a significantly improved fit (Raftery, 1995).”
From Children overcoming picky eating (COPE) - A cluster randomised controlled trial. (2020) · cites this paper as Raftery, 1995
3 more claims about this paper
Create a free account to see every claim other papers make about this one.
Citations
Cites
Estimating the Dimension of a Model
Schwarz, G. (1978) · The Annals of Statistics · doi:10.1214/aos/1176344136 · Cited by 47,033
(2005) · Statistics for Biology and Health · doi:10.1007/0-387-27255-0_9 · Cited by 20,133
Structural Equations with Latent Variables.
Clogg, C. C., & Bollen, K. A. (1991) · Contemporary Sociology · doi:10.2307/2072165 · Cited by 17,442
Crime and Punishment: An Economic Approach
Becker, G. S. (1968) · Journal of Political Economy · doi:10.1086/259394 · Cited by 13,999
Beckman, R. J. (1987) · Technometrics · doi:10.1080/00401706.1987.10488195 · Cited by 2,835
Participation in Illegitimate Activities: A Theoretical and Empirical Investigation
Ehrlich, I. (1973) · Journal of Political Economy · doi:10.1086/260058 · Cited by 2,492
The Comparative Method: Moving Beyond Qualitative and Quantitative Strategies
Ragin, C. C. (2014) · doi:10.1525/9780520957350 · Cited by 2,130
Accurate Approximations for Posterior Moments and Marginal Densities
Tierney, L., & Kadane, J. B. (1986) · Journal of the American Statistical Association · doi:10.1080/01621459.1986.10478240 · Cited by 2,129
Discrete Multivariate Analysis
(1977) · Technometrics · doi:10.1080/00401706.1977.10489605 · Cited by 1,757
Bayesian statistical inference for psychological research.
Edwards, W., Lindman, H., & Savage, L. J. (1963) · Psychological Review · doi:10.1037/h0044139 · Cited by 1,734
Cited by
Items 301–368 of 368
Wang, L., Long, D., Li, Z., & Armour, C. (2011) · Journal of Abnormal Child Psychology · doi:10.1007/s10802-011-9508-4 · Cited by 65
“72 output, and was thus calculated separately using the following formula: BIC 1⁄4 S B#2 þ lnðNÞ»t, where N = sample size and t = number of parameters estimated in the model (Raftery 1995).”
MethodInfluentialMay, H. (2006) · Journal of Educational and Behavioral Statistics · doi:10.3102/10769986031001063 · Cited by 65
“Raftery (1995, Equation 20) shows how the BIC difference can be calculated directly from the likelihood test above using the formula χ2 − df[ln(n)]. The difference in BIC statistics for these two models is 3,085. Raftery (1995) suggests a BIC larger than 10 “very strongly” favors the more complex model.”
BackgroundMethodInfluentialI Matter to My Friend, Therefore I am Happy: Friendship, Mattering, and Happiness
Demir, M., Özen, A., Doğan, A., Bilyk, N. A., & Tyrell, F. A. (2011) · Journal of Happiness Studies · doi:10.1007/s10902-010-9240-8 · Cited by 65
Ishida, K., Stupp, P., Melian, M., Serbanescu, F., & Goodwin, M. (2010) · Social Science & Medicine · doi:10.1016/j.socscimed.2010.08.007 · Cited by 63
Gamelike features might not improve data
Hawkins, G. E., Rae, B., Nesbitt, K. V., & Brown, S. D. (2013) · Behavior Research Methods · doi:10.3758/s13428-012-0264-3 · Cited by 61
Reorganisation of brain networks in frontotemporal dementia and progressive supranuclear palsy
Hughes, L. E., Ghosh, B. C. P., & Rowe, J. B. (2013) · NeuroImage: Clinical · doi:10.1016/j.nicl.2013.03.009 · Cited by 60
“…statement about the probabilities of models (within the set of 15 models tested) given the data, and do not make frequentist statistical inferences about the (un)likelihood of the data given a null hypotheses (Kass and Raftery, 1995; Raftery, 1995), obviating multiple comparison corrections.”
Giudice, N. A., Guenther, B. A., Jensen, N. A., & Haase, K. N. (2020) · Frontiers in Human Neuroscience · doi:10.3389/fnhum.2020.00087 · Cited by 60
“An advantage of the Bayesian approach is that it provides an opportunity to analytically determine whether the null hypothesis is more likely than the alternative(s) given the observed data (e.g., Raftery, 1995; Gallistel, 2009).”
BackgroundGood-enough level and dose-effect models: variation among outcomes and therapists
Owen, J. J., Adelson, J., Budge, S., Kopta, S. M., & Reese, R. J. (2016) · Psychotherapy Research · doi:10.1080/10503307.2014.966346 · Cited by 58
“Although there are no clear cut-offs for how much lower the better fitting model should be, Raftery (1995) suggests that differences over 10 points is strong evidence for the better fitting model.”
BackgroundMahoney, A., Li, I., Haskelberg, H., Millard, M., & Newby, J. M. (2021) · Journal of Affective Disorders · doi:10.1016/j.jad.2021.05.116 · Cited by 57
A-Tjak, J. G. L., Morina, N., Topper, M., & Emmelkamp, P. M. G. (2021) · BMC Psychiatry · doi:10.1186/s12888-020-03020-1 · Cited by 57
“Raftery A. Bayesian model selection in social research.”
BackgroundRetrieval-based model accounts for striking profile of episodic memory and generalization
Banino, A., Koster, R., Hassabis, D., & Kumaran, D. (2016) · Scientific Reports · doi:10.1038/srep31330 · Cited by 57
Tick, B., Colvert, E., McEwen, F., Stewart, C., Woodhouse, E., Gillan, N., Hallett, V., Lietz, S., Garnett, T., Simonoff, E., et al. (2016) · Journal of the American Academy of Child & Adolescent Psychiatry · doi:10.1016/j.jaac.2015.11.013 · Cited by 56
De Baene, W., Kühn, S., & Brass, M. (2012) · Human Brain Mapping · doi:10.1002/hbm.21234 · Cited by 52
“From a Bayesian perspective, a model is selected on the basis of the model evidence, i.e., the probability P(y|m) of the data y given a particular model m [Raftery, 1995].”
BackgroundMethodBiglan, A., Gau, J. M., Jones, L. B., Hinds, E., Rusby, J. C., Cody, C., & Sprague, J. (2015) · Journal of Contextual Behavioral Science · doi:10.1016/j.jcbs.2014.12.001 · Cited by 52
Neiss, M. B., Stevenson, J., Legrand, L. N., Iacono, W. G., & Sedikides, C. (2009) · Journal of Personality · doi:10.1111/j.1467-6494.2008.00549.x · Cited by 50
“Lower BIC values indicate better fit (Raftery, 1995).”
ResultFamily Transmission and Heritability of Externalizing Disorders: A Twin-Family Study
Hicks, B. M., Krueger, R. F., Iacono, W. G., McGue, M., & Patrick, C. J. (2004) · Archives of General Psychiatry · doi:10.1001/archpsyc.61.9.922 · Cited by 46
Salvadó, G., Larsson, V., Cody, K. A., Cullen, N. C., Jonaitis, E. M., Stomrud, E., Kollmorgen, G., Wild, N., Palmqvist, S., Janelidze, S., et al. (2023) · Alzheimer's & Dementia · doi:10.1002/alz.12907 · Cited by 46
“CSF levels of Aβ4215 as well as Total-Tau and Phospho-Tau(181P)16 were measured using the Elecsys β-Amyloid (1-42), Total-Tau, and Phospho-Tau(181P) CSF electrochemiluminescence immunoassays on a fully automated cobas e 601 instrument (Roche Diagnostics International Ltd.”
MethodMordeno, I. G., Carpio, J. G. E., Nalipay, M. J. N., & Saavedra, R. L. J. (2017) · Psychiatric Quarterly · doi:10.1007/s11126-016-9429-z · Cited by 46
Autin, F., Batruch, A., & Butera, F. (2015) · Frontiers in Psychology · doi:10.3389/fpsyg.2015.00707 · Cited by 45
Farmer, R. F., Seeley, J. R., Kosty, D. B., & Lewinsohn, P. M. (2009) · Journal of Abnormal Psychology · doi:10.1037/a0017205 · Cited by 44
James, D. (2017) · Journal of Racial and Ethnic Health Disparities · doi:10.1007/s40615-016-0269-1 · Cited by 44
Interpretation of main effects in the presence of non-significant interaction effects.
Lorah, J. A. (2020) · The Quantitative Methods for Psychology · doi:10.20982/tqmp.16.1.p033 · Cited by 44
Vanroelen, C., Levecque, K., & Louckx, F. (2009) · International Archives of Occupational and Environmental Health · doi:10.1007/s00420-008-0340-2 · Cited by 44
“BIC makes a correction to the L²-statistic for the number of model parameters and the sample size (Singer and Willett 2003): the lower the values of BIC, the better a model Wts the data ( Raftery 1995 ).”
MethodSemenza, D. C., & Stansfield, R. (2021) · Health & Place · doi:10.1016/j.healthplace.2021.102625 · Cited by 44
Bezdjian, S., Tuvblad, C., Wang, P., Raine, A., & Baker, L. A. (2014) · Developmental Psychology · doi:10.1037/a0038037 · Cited by 42
“(BIC; Raftery, 1995), in which smaller values correspond to better fitting models.”
BackgroundScandola, M., Aglioti, S. M., Lazzeri, G., Avesani, R., Ionta, S., & Moro, V. (2020) · Scientific Reports · doi:10.1038/s41598-020-62080-1 · Cited by 38
Online measurement of appraisals of students faced with curricular tasks.
Crombach, M. J., Boekaerts, M., & Voeten, M. J. M. (2003) · Educational and Psychological Measurement · doi:10.1177/0013164402239319 · Cited by 40
Duivis, H. E., Kupper, N., Vermunt, J. K., Penninx, B. W., Bosch, N. M., Riese, H., Oldehinkel, A. J., & de Jonge, P. (2015) · Health Psychology · doi:10.1037/hea0000210 · Cited by 40
“A BIC difference of 10 is strong evidence that the models with the lower BIC are better (Kass & Raftery, 1995; Raftery, 1995).”
BackgroundRelationship satisfaction and commitment in the transition to parenthood: A couple-centered approach
Leonhardt, N. D., Rosen, N. O., Dawson, S. J., Kim, J. J., Johnson, M. D., & Impett, E. A. (2022) · Journal of Marriage and Family · doi:10.1111/jomf.12785 · Cited by 39
Aharonovich, E., Sarvet, A., Stohl, M., DesJarlais, D., Tross, S., Hurst, T., Urbina, A., & Hasin, D. (2017) · Journal of Substance Abuse Treatment · doi:10.1016/j.jsat.2016.12.009 · Cited by 30
Analyzing paleomagnetic data: To anchor or not to anchor?
Heslop, D., & Roberts, A. P. (2016) · Journal of Geophysical Research: Solid Earth · doi:10.1002/2016jb013387 · Cited by 34
Profiles in health: multiple roles and health lifestyles in early adulthood
Mize, T. D. (2017) · Social Science & Medicine · doi:10.1016/j.socscimed.2017.02.017 · Cited by 33
Frost, R., Murphy, J., Hyland, P., Shevlin, M., Ben-Ezra, M., Hansen, M., Armour, C., McCarthy, A., Cunningham, T., & McDonagh, T. (2020) · European Journal of Psychotraumatology · doi:10.1080/20008198.2020.1836864 · Cited by 33
“The model with the lowest BIC value is considered the best fit and a difference of 10 points provides evidence in favour of the model with the lower value (Raftery, 1995).”
ResultChu, B. C., Skriner, L. C., & Zandberg, L. J. (2013) · Journal of Consulting and Clinical Psychology · doi:10.1037/a0033390 · Cited by 33
Herrera, M. S., Galkuté, M., Fernández, M. B., & Elgueta, R. (2022) · Social Science & Medicine · doi:10.1016/j.socscimed.2022.114786 · Cited by 33
Longitudinal Associations among Child Support Debt, Employment, and Recidivism after Prison
Link, N. W., & Roman, C. G. (2017) · The Sociological Quarterly · doi:10.1080/00380253.2016.1246892 · Cited by 30
The Multifaceted Role of Self‐Generated Question Asking in Curiosity‐Driven Learning
Kedrick, K., Schrater, P., & Koutstaal, W. (2023) · Cognitive Science · doi:10.1111/cogs.13253 · Cited by 29
Law (1998) · doi:10.3310/phr02050 · Cited by 29
Orthorexia Nervosa and Dieting in a Non-Clinical Sample: A Prospective Study
Novara, C., Pardini, S., Visioli, F., & Meda, N. (2022) · Eating and Weight Disorders - Studies on Anorexia, Bulimia and Obesity · doi:10.1007/s40519-021-01353-8 · Cited by 28
Skowronski, M., Busching, R., & Krahé, B. (2021) · Sex Roles · doi:10.1007/s11199-020-01187-1 · Cited by 27
Haspolat, N. K., & Yalçın, İ. (2023) · Psychology in the Schools · doi:10.1002/pits.23012 · Cited by 28
Nakano, S., & Watkins, M. W. (2013) · Psychology in the Schools · doi:10.1002/pits.21724 · Cited by 27
“015 (Chen, 2007), and BIC > +2 (Raftery, 1995) as standards.”
Method(2020) · USENIX Security Symposium · Cited by 27
Impulsivity influences betting under stress in laboratory gambling
Canale, N., Rubaltelli, E., Vieno, A., Pittarello, A., & Billieux, J. (2017) · Scientific Reports · doi:10.1038/s41598-017-10745-9 · Cited by 26
Promotional language and the adoption of innovative ideas in science
Peng, H., Qiu, H. S., Fosse, H. B., & Uzzi, B. (2024) · Proceedings of the National Academy of Sciences · doi:10.1073/pnas.2320066121 · Cited by 26
The Relationship Between English Language Learner Status and Music Ensemble Participation
Lorah, J. A., Sanders, E. A., & Morrison, S. J. (2014) · Journal of Research in Music Education · doi:10.1177/0022429414542301 · Cited by 26
McCullough, C., Harding, H. G., Shaffer, A., Han, R. Z., & Bright, M. (2014) · Journal of Family Violence · doi:10.1007/s10896-014-9593-6 · Cited by 25
Surrogate Decision Making: Do We Have to Trade Off Accuracy and Procedural Satisfaction?
Frey, R., Hertwig, R., & Herzog, S. M. (2014) · Medical Decision Making · doi:10.1177/0272989x12471729 · Cited by 24
Confirmatory factor analysis of the WISC-IV in a hospital referral sample.
Devena, S. E., Gay, C. E., & Watkins, M. W. (2013) · Journal of Psychoeducational Assessment · doi:10.1177/0734282913483981 · Cited by 22
“01 and ΔBIC > 5 indicated meaningful change (Cheung & Rensvold, 2002; Raftery, 1995).”
MethodMcCarthy, E., DeViva, J. C., Norman, S. B., Southwick, S. M., & Pietrzak, R. H. (2019) · Psychological Trauma Theory Research Practice and Policy · doi:10.1037/tra0000436 · Cited by 21
Wolpe, N., Nombela, C., & Rowe, J. B. (2015) · Frontiers in Psychology · doi:10.3389/fpsyg.2015.01514 · Cited by 18
Kuss, O., & Kuehn, A. (2008) · Acta Obstetricia Et Gynecologica Scandinavica · doi:10.1080/00016340802478174 · Cited by 17
PTSD factor structure differences between veterans with and without a PTSD diagnosis
Biehn, T. L., Elhai, J. D., Fine, T. H., Seligman, L. D., & Richardson, J. D. (2012) · Journal of Anxiety Disorders · doi:10.1016/j.janxdis.2012.01.008 · Cited by 17
Children overcoming picky eating (COPE) - A cluster randomised controlled trial.
Bennett, C., Copello, A., Jones, C., & Blissett, J. (2020) · Appetite · doi:10.1016/j.appet.2020.104791 · Cited by 17
Rutherford, M., Forsyth, K., McKenzie, K., McClure, I., Murray, A., McCartney, D., Irvine, L., & O’Hare, A. (2018) · Journal of Autism and Developmental Disorders · doi:10.1007/s10803-018-3501-5 · Cited by 15
“Here we judged ‘better fitting’ to be a model in which parsimony corrected fit indexes Akaike’s Information Criteria (AIC; Akaike 1987) and Bayesian Information Criterion (BIC; Rafferty 1995) were smaller than the comparison model.”
Method(1998) · doi:10.1007/3-540-26577-5_17 · Cited by 15
Integrative Bayesian analysis of neuroimaging-genetic data with application to cocaine dependence
Azadeh, S., Hobbs, B. P., Ma, L., Nielsen, D. A., Gerard Moeller, F., & Baladandayuthapani, V. (2016) · NeuroImage · doi:10.1016/j.neuroimage.2015.10.033 · Cited by 12
Subcellular mRNA kinetic modeling reveals nuclear retention as rate-limiting
(2024) · doi:10.1038/s44320-024-00073-2 · Cited by 14
Antoniou, F., Alkhadim, G., Mouzaki, A., & Simos, P. (2022) · Journal of Intelligence · doi:10.3390/jintelligence10010006 · Cited by 12
Postoperative serum cortisol and Cushing disease recurrence in patients with corticotroph adenomas
Catalino, M. P., Moore, D. T., Ironside, N., Munoz, A. R., Coley, J., Jonas, R., Kearns, K., Min, L., Vance, M. L., Jane, J. A., et al. (2023) · The Journal of Clinical Endocrinology & Metabolism · doi:10.1210/clinem/dgad347 · Cited by 11
The etiology of social aggression: a nuclear twin family study
Slawinski, B. L., Klump, K. L., & Burt, S. A. (2019) · Psychological Medicine · doi:10.1017/s0033291718000697 · Cited by 10
Almansa, J., Vermunt, J. K., Forero, C. G., Vilagut, G., De Graaf, R., De Girolamo, G., & Alonso, J. (2011) · International Journal of Methods in Psychiatric Research · doi:10.1002/mpr.334 · Cited by 9
“Using these indexes it is possible to choose the best‐fitting model among a set of nested or non‐nested alternative models (Raftery, 1995).”
Lou, C., Zhou, X., Lebowitz, E. R., Williams, L. L., & Storch, E. A. (2020) · Frontiers in Psychology · doi:10.3389/fpsyg.2020.01018 · Cited by 9
“A BIC difference of 10 represents a 150:1 likelihood (p < 0.05) that the model with lower BIC value fits better; the difference between 6 to 10 provides strong support; a difference of more than 10 indicates very strong support (Raftery, 1995).”
BackgroundInfluentialFacing stress: No effect of acute stress at encoding or retrieval on face recognition memory
Marr, C., Quaedflieg, C. W. E. M., Otgaar, H., Hope, L., & Sauerland, M. (2021) · Acta Psychologica · doi:10.1016/j.actpsy.2021.103376 · Cited by 8
The Structure and Stability of Common Mental Disorders: The NEMESIS Study
Vollebergh, W. A. M., Iedema, J., Bijl, R. V., de Graaf, R., Smit, F., & Ormel, J. (2001) · Archives of General Psychiatry · doi:10.1001/archpsyc.58.6.597 · Cited by 6
Identifying subgroups of urge suppression in Obsessive-Compulsive Disorder using machine learning
Eng, G. K., De Nadai, A. S., Collins, K. A., Recchia, N., Tobe, R. H., Bragdon, L. B., & Stern, E. R. (2024) · Journal of Psychiatric Research · doi:10.1016/j.jpsychires.2024.06.052 · Cited by 6
Enhanced strong-lensing model of MACS J0138.0–2155 based on new JWST and VLT/MUSE observations
Acebron, A., Bergamini, P., Rosati, P., Tozzi, P., Meneghetti, M., Caminha, G. B., Ertl, S., Granata, G., Koekemoer, A. M., Grillo, C., et al. (2025) · Astronomy & Astrophysics · doi:10.1051/0004-6361/202554468 · Cited by 6
Wladis, C., Sencindiver, B., Hachey, A. C., Myszkowski, N., Offenholley, K., Samuels, J., & Aly, G. (2026) · AERA Open · doi:10.1177/23328584261419497 · Cited by 1
Metadata
- Authors
- Raftery, A. E.
- Affiliations
- University of Washington (US)
- Corresponding author
- Raftery, Adrian E.
- ORCID
- Raftery, Adrian E.
- Venue
- Sociological Methodology
- Publisher
- JSTOR
- Published
- 1995
- Volume
- 25
- Pages
- 111
- ISSN
- 0081-1750 (ISSN-L)
- OpenAlex
- W2089763487
- Cited by
- 6,823
- References
- 44
- Citations by year
- 2026: 219 · 2025: 305 · 2024: 321 · 2023: 344 · 2022: 383 · 2021: 423 · 2020: 433 · 2019: 447 · 2018: 430 · 2017: 351 · 2016: 372 · 2015: 363 · 2014: 307 · 2013: 305 · 2012: 251
- FWCI
- 27.28
- Type
- journal article
- Language
- en
- Open access
- closed
- Locations
- Sociological Methodology (publishedVersion)
- Topics
- Physical Sciences › Mathematics › Statistics and Probability › Statistical Methods and Bayesian Inference
- Physical Sciences › Mathematics › Statistics and Probability › Statistical Methods and Inference
- Physical Sciences › Mathematics › Statistics and Probability › Statistical Methods and Applications
- Keywords
- Bayesian model selection, model uncertainty, p-values, feature selection, Bayesian information criterion, non-nested models, hypothesis testing, convergence theory, social norms, large-sample inference, regression analysis
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.