Latent Class Analysis: A Guide to Best Practice
Open accessThe DOI matches its Crossref record; 51 of the 51 references it cites were found in a registry or an index.
Abstract
Latent class analysis (LCA) is a statistical procedure used to identify qualitatively different subgroups within populations who often share certain outward characteristics. The assumption underlying LCA is that membership in unobserved groups (or classes) can be explained by patterns of scores across survey questions, assessment indicators, or scales. The application of LCA is an active area of research and continues to evolve. As more researchers begin to apply the approach, detailed information on key considerations in conducting LCA is needed. In the present article, we describe LCA, review key elements to consider when conducting LCA, and provide an example of its application.
Overview
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
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?
The cited source says what the claim attributes to it.
In the citing paper
“However, LCA and cluster analysis make different assumptions about the data and use different statistical procedures [43].”
From Recommendations for empirical syndemics analyses: A stepwise methodological guide (2024) · cites this paper as [43]
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The self-check needs the paper’s full text.
Citations
Cites
Principles and Practice of Structural Equation Modeling
Tabri, N., & Elliott, C. M. (2012) · Canadian Graduate Journal of Sociology and Criminology · doi:10.15353/cgjsc-rcessc.v1i1.25 · Cited by 38,891
Nylund, K. L., Asparouhov, T., & Muthén, B. O. (2007) · Structural Equation Modeling: A Multidisciplinary Journal · doi:10.1080/10705510701575396 · Cited by 11,415 · cited 9×
“…(BIC) may be the most reliable fit statistic and should routinely be reported (Nylund et al., 2007; Vermunt, 2002); and (c) theoretical interpretability should be considered in choosing a solution (B. O. Muthén & Muthén, 2000; Nylund et al., 2007; Shanahan et al., 2013; Stringaris et al., 2013).”
BackgroundMethodInfluentialMcLachlan, G., & Peel, D. (2000) · Wiley Series in Probability and Statistics · doi:10.1002/0471721182 · Cited by 10,991
Likelihood Ratio Tests for Model Selection and Non-Nested Hypotheses
Vuong, Q. H. (1989) · Econometrica · doi:10.2307/1912557 · Cited by 5,929 · cited 2×
“Other fit statistics can also be used to select a final class, such as likelihood tests (i.e., Vuong-Lo-Mendell-Rubin adjusted likelihood ratio test (Lo et al., 2001; Vuong, 1989) and the bootstrapped likelihood ratio test (McLachlan & Peel, 2000).”
MethodTesting the number of components in a normal mixture
Lo, Y. (2001) · Biometrika · doi:10.1093/biomet/88.3.767 · Cited by 4,640 · cited 2×
“Other fit statistics can also be used to select a final class, such as likelihood tests (i.e., Vuong-Lo-Mendell-Rubin adjusted likelihood ratio test (Lo et al., 2001; Vuong, 1989) and the bootstrapped likelihood ratio test (McLachlan & Peel, 2000).”
MethodAuxiliary Variables in Mixture Modeling: Three-Step Approaches Using M plus
Asparouhov, T., & Muthén, B. (2014) · Structural Equation Modeling: A Multidisciplinary Journal · doi:10.1080/10705511.2014.915181 · Cited by 3,535
Muthén, B., & Muthén, L. K. (2000) · Alcoholism: Clinical and Experimental Research · doi:10.1111/j.1530-0277.2000.tb02070.x · Cited by 3,149 · cited 9×
“The average latent class posterior probability is the average probability of the class model accurately predicting class membership for individuals (B. O. Muthén & Muthén, 2000).”
BackgroundMethodInfluentialA practical solution to the pervasive problems ofp values
Wagenmakers, E.-J. (2007) · Psychonomic Bulletin & Review · doi:10.3758/bf03194105 · Cited by 2,490 · cited 1×
“The BF indicated a four- or five-class model were in the moderately acceptable range (Wagenmakers, 2007).”
BackgroundAn entropy criterion for assessing the number of clusters in a mixture model
Celeux, G., & Soromenho, G. (1996) · Journal of Classification · doi:10.1007/bf01246098 · Cited by 2,448 · cited 1×
“In general, an entropy value close to 1 is ideal (Celeux & Soromenho, 1996) and above .8 is acceptable.”
BackgroundHow to Use a Monte Carlo Study to Decide on Sample Size and Determine Power
Muthén, L. K., & Muthén, B. O. (2002) · Structural Equation Modeling: A Multidisciplinary Journal · doi:10.1207/s15328007sem0904_8 · Cited by 2,362
Latent Class Modeling with Covariates: Two Improved Three-Step Approaches
Vermunt, J. K. (2010) · Political Analysis · doi:10.1093/pan/mpq025 · Cited by 2,343 · cited 3×
“Previously, researchers would include covariates in the same model as the model used to identify the class solution (Vermunt, 2010).”
MethodTen frequently asked questions about latent class analysis
Nylund-Gibson, K., & Choi, A. Y. (2018) · Translational Issues in Psychological Science · doi:10.1037/tps0000176 · Cited by 2,206 · cited 6×
“It should be noted that historically some of the diagnostics were used to select a final model (B. O. Muthén, 1998-2004); however, suggestions on their use have recently changed (Nylund-Gibson & Choi, 2018).”
BackgroundMethodInfluential(2002) · Cambridge University Press eBooks · doi:10.1017/cbo9780511499531 · Cited by 1,958
Collins, L. M., & Lanza, S. T. (2009) · Wiley Series in Probability and Statistics · doi:10.1002/9780470567333 · Cited by 1,545
Latent Class Analysis and Finite Mixture Modeling
Masyn, K. E. (2013) · Oxford Handbooks Online · doi:10.1093/oxfordhb/9780199934898.013.0025 · Cited by 1,458 · cited 4×
“In addition to evaluating fit, researchers need to review classification diagnostics (Masyn, 2013).”
BackgroundMethodInfluentialAppelbaum, M., Cooper, H., Kline, R. B., Mayo-Wilson, E., Nezu, A. M., & Rao, S. M. (2018) · American Psychologist · doi:10.1037/amp0000191 · Cited by 1,449 · cited 1×
“To present clear results and justified conclusions, LCA reports need to detail study procedures and results with clarity and coherence (Appelbaum et al., 2018).”
BackgroundEstimating latent structure models with categorical variables: one-step versus three-step estimators
Bolck, A., Croon, M., & Hagenaars, J. (2004) · Political Analysis · doi:10.1093/pan/mph001 · Cited by 1,406
NAGIN, D. S., & LAND, K. C. (1993) · Criminology · doi:10.1111/j.1745-9125.1993.tb01133.x · Cited by 1,127 · cited 1×
“We further sought to have the lowest average latent class posterior probability be .80 or higher (Nagin & Land, 1993; Weden & Zabin, 2005).”
BackgroundImpact of Bullying in Childhood on Adult Health, Wealth, Crime, and Social Outcomes
Wolke, D., Copeland, W. E., Angold, A., & Costello, E. J. (2013) · Psychological Science · doi:10.1177/0956797613481608 · Cited by 640 · cited 1×
“The assumption underlying LCA is that membership in unobserved classes can cause or explain patterns of scores across survey questions, assessment indicators, or scales (B. O. Muthén & Muthén, 2000; Wolke et al., 2013).”
BackgroundClogg, C. C. (1995) · Handbook of Statistical Modeling for the Social and Behavioral Sciences · doi:10.1007/978-1-4899-1292-3_6 · Cited by 621 · cited 1×
“LCA was first introduced in 1950 (Lazarsfeld, 1950) and has since undergone a number of revisions and advancements (for a full summary, see Clogg, 1981, 1995; Hagenaars, 1990; Vermunt, 1997; Nylund-Gibson & Choi, 2018).”
BackgroundWurpts, I. C., & Geiser, C. (2014) · Frontiers in Psychology · doi:10.3389/fpsyg.2014.00920 · Cited by 354 · cited 1×
“Currently, no consensus exists on the number of indicator variables to include in a model, but generally more indicator variables lead to better results (Wurpts & Geiser, 2014).”
BackgroundWANG, M.-C., DENG, Q., BI, X., YE, H., & YANG, W. (2017) · Acta Psychologica Sinica · doi:10.3724/sp.j.1041.2017.01473 · Cited by 286 · cited 1×
“Entropy is another diagnostic statistic (Wang et al., 2017).”
Nylund-Gibson, K., & Masyn, K. E. (2016) · Structural Equation Modeling: A Multidisciplinary Journal · doi:10.1080/10705511.2016.1221313 · Cited by 280 · cited 2×
“Because nuances of including covariates in LCA models vary by statistical software and the approaches used, researchers are encouraged to consult articles with details on the topic (e.g., Asparouhov & Muthén 2014; Nylund-Gibson & Choi, 2018; Nylund-Gibson & Masyn, 2016; Vermunt & Magidson, 2016).”
MethodLatent Class Analysis: An example for reporting results
Schreiber, J. B. (2017) · Research in Social and Administrative Pharmacy · doi:10.1016/j.sapharm.2016.11.011 · Cited by 203 · cited 2×
“Despite variation in researchers’ perspectives (e.g., Nylund et al., 2007; Schreiber, 2017) on which model statistics to report, we recommend reporting the BIC and at least two additional fit indices, as well as entropy, the percentage and size of the smallest class, and the smallest off-diagonal value of the average…”
BackgroundPetersen, K. J., Qualter, P., & Humphrey, N. (2019) · Frontiers in Psychology · doi:10.3389/fpsyg.2019.01214 · Cited by 195 · cited 10×
“Because of these debates, a number of recent systematic reviews have been conducted to summarize the LCA literature (e.g., Killian et al., 2019; Petersen et al., 2019; Ulbricht et al., 2018).”
BackgroundInfluentialLatent Class Analysis: A Method for Capturing Heterogeneity
Scotto Rosato, N., & Baer, J. C. (2012) · Social Work Research · doi:10.1093/swr/svs006 · Cited by 182 · cited 1×
“Studies have used as few as four indicator variables (Travis & CombsOrme, 2007), whereas other studies have used more than 20 indicators (Rosato & Baer, 2012).”
MethodFactor Analysis and Measurement in Sociological Research.
Muller, K. E., Jackson, D. J., & Borgatta, E. R. (1983) · Journal of the American Statistical Association · doi:10.2307/2288232 · Cited by 182
A Review of Mental Health and Mental Health Care Disparities Research: 2011-2014
Cook, B. L., Hou, S. S.-Y., Lee-Tauler, S. Y., Progovac, A. M., Samson, F., & Sanchez, M. J. (2019) · Medical Care Research and Review · doi:10.1177/1077558718780592 · Cited by 178 · cited 1×
“…caregiver-reported versus youth-reported mental health outcomes (e.g., Breland-Noble & Weller, 2012) and disparities in access to providers (e.g., Cook et al., 2019), the prevalence of mental health conditions may be underreported in this study, which, by extension, may indicate that the current…”
BackgroundA systematic review of person-centered approaches to investigating patterns of trauma exposure
O'Donnell, M. L., Schaefer, I., Varker, T., Kartal, D., Forbes, D., Bryant, R. A., Silove, D. M., Creamer, M. C., McFARLANE, A. C., Malhi, G. S., et al. (2017) · Clinical Psychology Review · doi:10.1016/j.cpr.2017.08.009 · Cited by 175 · cited 1×
“However, these suggestions have been relaxed and a number of publications have included class sizes smaller than 5% or 50 cases (e.g., O’Donnell et al., 2017).”
BackgroundStringaris, A., Maughan, B., Copeland, W. S., Costello, E. J., & Angold, A. (2013) · Journal of the American Academy of Child & Adolescent Psychiatry · doi:10.1016/j.jaac.2013.05.017 · Cited by 169 · cited 3×
“…(BIC) may be the most reliable fit statistic and should routinely be reported (Nylund et al., 2007; Vermunt, 2002); and (c) theoretical interpretability should be considered in choosing a solution (B. O. Muthén & Muthén, 2000; Nylund et al., 2007; Shanahan et al., 2013; Stringaris et al., 2013).”
BackgroundThe use of latent class analysis for identifying subtypes of depression: a systematic review
Ulbricht, C. M., Chrysanthopoulou, S. A., Levin, L., & Lapane, K. L. (2018) · Psychiatry Research · doi:10.1016/j.psychres.2018.03.003 · Cited by 138 · cited 1×
“Because of these debates, a number of recent systematic reviews have been conducted to summarize the LCA literature (e.g., Killian et al., 2019; Petersen et al., 2019; Ulbricht et al., 2018).”
BackgroundGender and Ethnic Differences in the Co-occurrence of Adolescent Risk Behaviors
Weden, M. M., & Zabin, L. S. (2005) · Ethnicity & Health · doi:10.1080/13557850500115744 · Cited by 117 · cited 2×
“We further sought to have the lowest average latent class posterior probability be .80 or higher (Nagin & Land, 1993; Weden & Zabin, 2005).”
BackgroundMethodLatent Class Analysis of Complex Sample Survey Data
Patterson, B. H., Dayton, C. M., & Graubard, B. I. (2002) · Journal of the American Statistical Association · doi:10.1198/016214502388618465 · Cited by 103
Venables, W. N., & Smith, D. (2002) · Cited by 98
Shanahan, L., Copeland, W. E., Worthman, C. M., Erkanli, A., Angold, A., & Costello, E. J. (2013) · Psychoneuroendocrinology · doi:10.1016/j.psyneuen.2013.04.010 · Cited by 96 · cited 5×
“…(BIC) may be the most reliable fit statistic and should routinely be reported (Nylund et al., 2007; Vermunt, 2002); and (c) theoretical interpretability should be considered in choosing a solution (B. O. Muthén & Muthén, 2000; Nylund et al., 2007; Shanahan et al., 2013; Stringaris et al., 2013).”
BackgroundInfluentialByrd, C. M., & Andrews, D. J. C. (2016) · Journal of School Psychology · doi:10.1016/j.jsp.2016.05.001 · Cited by 89 · cited 1×
“LCA has been used to study a variety of issues and vulnerable populations, such as mental health among Black youth (Rose et al., 2017), adolescent perceptions of in-school discrimination (Byrd & Carter-Andrews, 2016), and young Malawaian adults with or at-risk for HIV (Weller & Small, 2015).”
MethodMental Health and Educational Experiences Among Black Youth: A Latent Class Analysis
Rose, T., Lindsey, M. A., Xiao, Y., Finigan‐Carr, N., & Joe, S. (2017) · Journal of Youth and Adolescence · doi:10.1007/s10964-017-0723-3 · Cited by 86
A systematic review of latent variable mixture modeling research in social work journals
Killian, M. O., Cimino, A. N., Weller, B. E., & Hyun Seo, C. (2019) · Journal of Evidence-Based Social Work · doi:10.1080/23761407.2019.1577783 · Cited by 65 · cited 3×
“The most reported is the Bayesian Information Criterion (BIC; Killian et al., 2019).”
BackgroundSelf-citationResilient Parenting: Overcoming Poor Parental Bonding
Travis, W. J., & Combs-Orme, T. (2007) · Social Work Research · doi:10.1093/swr/31.3.135 · Cited by 49
Weller, B. E., Blanford, K. L., & Butler, A. M. (2018) · Journal of Adolescent Health · doi:10.1016/j.jadohealth.2017.12.020 · Cited by 38 · cited 1×
“Consistent with previous research using data from the NSCH, we included the following control variables: gender, income based on federal poverty level, highest level of education, and primary language spoken in the home (e.g., Butler et al., 2015; Weller et al., 2018; Weller et al., 2019).”
MethodSelf-citationBowen, N. K., Lee, J., & Weller, B. E. (2007) · Children & Schools · doi:10.1093/cs/29.4.229 · Cited by 30 · cited 1×
“We have found it helpful to use theory or previous research to inform the number of classes we anticipated finding (Bowen et al., 2007; Weller et al., 2013).”
MethodSelf-citationButler, A. M., Weller, B. E., & Titus, C. (2015) · Administration and Policy in Mental Health and Mental Health Services Research · doi:10.1007/s10488-014-0612-y · Cited by 30 · cited 1×
“Consistent with previous research using data from the NSCH, we included the following control variables: gender, income based on federal poverty level, highest level of education, and primary language spoken in the home (e.g., Butler et al., 2015; Weller et al., 2018; Weller et al., 2019).”
MethodSelf-citationBreland‐Noble, A. M., & Weller, B. E. (2011) · Journal of Child and Family Studies · doi:10.1007/s10826-011-9547-z · Cited by 27 · cited 1×
“…interval; ADHD = attention deficit/hyperactivity disorder. discrepancies in caregiver-reported versus youth-reported mental health outcomes (e.g., Breland-Noble & Weller, 2012) and disparities in access to providers (e.g., Cook et al., 2019), the prevalence of mental health conditions may be…”
BackgroundSelf-citationLythgoe, D. T., García‐Fiñana, M., & Cox, T. F. (2018) · Structural Equation Modeling A Multidisciplinary Journal · doi:10.1080/10705511.2018.1495081 · Cited by 25 · cited 1×
“However, this approach is most commonly used in LCA methodological papers (e.g., Kim & Wang, 2019; Lythgoe et al., 2019; Shin et al., 2019).”
MethodYouth Access to Medical Homes and Medical Home Components by Race and Ethnicity
Weller, B. E., Faubert, S. J., & Ault, A. K. (2019) · Maternal and Child Health Journal · doi:10.1007/s10995-019-02831-3 · Cited by 25 · cited 1×
“Consistent with previous research using data from the NSCH, we included the following control variables: gender, income based on federal poverty level, highest level of education, and primary language spoken in the home (e.g., Butler et al., 2015; Weller et al., 2018; Weller et al., 2019).”
MethodSelf-citationLEM: A general program for the analysis of categorical data. Users manual
Vermunt, J. K. (1997) · Data Archiving and Networked Services (DANS) · Cited by 24
Comparing the Robustness of Stepwise Mixture Modeling With Continuous Nonnormal Distal Outcomes
Shin, M., No, U., & Hong, S. (2019) · Educational and Psychological Measurement · doi:10.1177/0013164419839770 · Cited by 20 · cited 1×
“However, this approach is most commonly used in LCA methodological papers (e.g., Kim & Wang, 2019; Lythgoe et al., 2019; Shin et al., 2019).”
MethodKim, E., & Wang, Y. (2018) · Structural Equation Modeling A Multidisciplinary Journal · doi:10.1080/10705511.2018.1521284 · Cited by 19 · cited 1×
“However, this approach is most commonly used in LCA methodological papers (e.g., Kim & Wang, 2019; Lythgoe et al., 2019; Shin et al., 2019).”
MethodLiu, S. R., Kia‐Keating, M., & Nylund-Gibson, K. K. (2019) · Prevention Science · doi:10.1007/s11121-019-01021-5 · Cited by 17 · cited 2×
“Results are somewhat consistent with the few other studies that have identified latent profiles of promotive factors among Black adolescent (e.g., Liu et al., 2019).”
ResultWeller, B. E., Bowen, N. K., & Bowen, G. L. (2012) · Journal of Social Work · doi:10.1177/1468017311435446 · Cited by 16 · cited 2×
“We have found it helpful to use theory or previous research to inform the number of classes we anticipated finding (Bowen et al., 2007; Weller et al., 2013).”
MethodSelf-citationProfiles of Risk for HIV/AIDS Among Young Malawian Adults: Understanding Behavioral Intentions
Weller, B. E., & Small, E. (2015) · Sexuality & Culture · doi:10.1007/s12119-015-9289-y · Cited by 4 · cited 2×
“Profiles of risk for HIV/AIDS among young Malawian adults: Understanding behavioral intentions.”
MethodSelf-citation
Research works printed but not linked: 8
- Collins L. M., 2010, Latent class and latent transition analysis
- Hagenaars J. A., 1990, Categorical longitudinal data: Log-linear analysis of panel, trend and cohort data
- Lazarsfeld P. F., 1950, Studies in social psychology World War II: Measurement and prediction, 4, 361
- Muthén B. O., 1998, Mplus technical appendices
- Muthén L. K., 1998, Mplus user’s guide
- Norusis M. J., 1990, SPSS/PC+ Statistics 4.0
- SAS Institute Inc. (2016). SAS/SHARE® 9.4: User’s guide (2nd ed.). http://documentation.sas.com/api/docsets/shrref/9.4/content/shrref.pdf?locale=en#nameddest=bookinfo
- StataCorp LLC. (1985-2019). Stata user’s guide: Release 16. Stata Press. https://www.stata.com/manuals/u.pdf
Cited by
Symptoms and risk factors for long COVID in non-hospitalized adults
Subramanian, A., Nirantharakumar, K., Hughes, S., Myles, P., Williams, T., Gokhale, K. M., Taverner, T., Chandan, J. S., Brown, K., Simms-Williams, N., et al. (2022) · Nature Medicine · doi:10.1038/s41591-022-01909-w · Cited by 1,065
Whose AI? How different publics think about AI and its social impacts
Bao, L., Krause, N. M., Calice, M. N., Scheufele, D. A., Wirz, C. D., Brossard, D., Newman, T. P., & Xenos, M. A. (2022) · Computers in Human Behavior · doi:10.1016/j.chb.2022.107182 · Cited by 162
Recommended Practices in Latent Class Analysis Using the Open-Source R-Package tidySEM
Lissa, C. J. V., Garnier‐Villarreal, M., & Anadria, D. (2023) · Structural Equation Modeling A Multidisciplinary Journal · doi:10.1080/10705511.2023.2250920 · Cited by 143
“According to the naming fallacy , overly simplistic and general class names can be misleading (Weller et al., 2020).”
BackgroundMethodSaqr, M., López-Pernas, S., Helske, S., & Hrastinski, S. (2023) · Computers & Education · doi:10.1016/j.compedu.2023.104787 · Cited by 95
Wang, Y., Wu, H., & Wang, Y. (2025) · Journal of Multilingual and Multicultural Development · doi:10.1080/01434632.2024.2379534 · Cited by 104
“This approach tends to overlook the heterogeneity inherent in learner populations, focusing on mean effects and disregarding the nuanced individual differences in reactions to educational stimuli (Weller, Bowen, and Faubert 2020).”
Metadata
- Authors
- Weller, B. E., Bowen, N. K., & Faubert, S. J.
- Affiliations
- Western Michigan University (US)
- The Ohio State University (US)
- Grand Valley State University (US)
- Corresponding author
- Weller, Bridget E.
- Venue
- Journal of Black Psychology
- Publisher
- SAGE Publications
- Published
- 2020-05
- Volume
- 46
- Issue
- 4
- Pages
- 287-311
- ISSN
- 0095-7984 (ISSN-L)
- 1552-4558
- OpenAlex
- W3032942076
- Cited by
- 2,223
- References
- 51
- Citations by year
- 2026: 524 · 2025: 605 · 2024: 526 · 2023: 351 · 2022: 195 · 2021: 20 · 2020: 1
- FWCI
- 90.10
- Type
- journal article
- Language
- en
- Open access
- bronze
- doi.org
- Licence
- journals.sagepub.com/page/policies/text-and-data-mining-license (tdm)
- Locations
- Journal of Black Psychology (publishedVersion)
- Topics
- Social Sciences › Decision Sciences › Management Science and Operations Research › Psychometric Methodologies and Testing
- Social Sciences › Psychology › Applied Psychology › Psychological Testing and Assessment
- Social Sciences › Education › Educational Assessment and Pedagogy
- Keywords
- latent class analysis, subgroups, best practices, evaluation metrics, scale scores, group membership
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.