Pushing the Limits
The Performance of Maximum Likelihood and Bayesian Estimation With Small and Unbalanced Samples in a Latent Growth Model
Abstract
Abstract. Longitudinal developmental research is often focused on patterns of change or growth across different (sub)groups of individuals. Particular to some research contexts, developmental inquiries may involve one or more (sub)groups that are small in nature and therefore difficult to properly capture through statistical analysis. The current study explores the lower-bound limits of subsample sizes in a multiple group latent growth modeling by means of a simulation study. We particularly focus on how the maximum likelihood (ML) and Bayesian estimation approaches differ when (sub)sample sizes are small. The results show that Bayesian estimation resolves computational issues that occur with ML estimation and that the addition of prior information can be the key to detect a difference between groups when sample and effect sizes are expected to be limited. The acquisition of prior information with respect to the smaller group is especially influential in this context.
References
2010, September). Bayesian analysis of latent variable models using Mplus. Retrieved from https://www.statmodel.com/techappen.shtml
(2001).
(The robustness of LISREL modeling revisited . In R. CudeckK. G. JöreskogD. SörbomEds., Structural equation models: Present and future. A festschrift in honor of Karl Jöreskog (pp. 139–168). Lincolnwood, IL: Scientific Software International.2015). Collinear latent variables in multilevel confirmatory factor analysis a comparison of maximum likelihood and Bayesian estimations. Educational and Psychological Measurement, 75, 406–427. https://doi.org/10.1177/0013164414547959
(1988). Statistical power analysis for the behavioral sciences (2nd ed.). Hillsdale, NJ: Erlbaum. https://doi.org/10.4324/9780203771587
(2013). Mixture class recovery in GMM under varying degrees of class separation: Frequentist versus Bayesian estimation. Psychological Methods, 18, 186. https://doi.org/10.1037/a0031609
(2015). Improving transparency and replication in Bayesian statistics: The WAMBS-checklist. Psychological Methods, 22(2), 240–261. https://doi.org/10.1037/met0000065
(1992). Inference from iterative simulation using multiple sequences. Statistical Science, 7, 457–472. https://doi.org/10.1214/ss/1177011136
(2013, October). MplusAutomation: Automating Mplus model estimation and interpretation. Package MplusAutomation. Retrieved from https://cran.r-project.org/web/packages/MplusAutomation/MplusAutomation.pdf
(2017). Analyzing organizational growth in repeated cross-sectional designs using multilevel structural equation modeling. Methodology, 13, 83–97. https://doi.org/10.1027/1614-2241/a000133
(2001). The accuracy of multilevel structural equation modeling with pseudobalanced groups and small samples. Structural Equation Modeling, 8, 157–174. https://doi.org/10.1207/S15328007SEM08021
(2014). Analyzing indirect effects in cluster randomized trials. The effect of estimation method, number of groups and group sizes on accuracy and power. Frontiers in Psychology, 5, 78. https://doi.org/10.3389/fpsyg.2014.00078
(2012). How few countries will do? Comparative survey analysis from a Bayesian perspective. Survey Research Association, 6, 87–93. https://doi.org/10.18148/srm/2012.v6i2.5033
(2009). Functional consequences of marijuana use in adolescents. Pharmacology, Biochemistry and Behavior, 4, 559–565. https://doi.org/10.1016/j.pbb.2009.04.001
(2011). Introduction to special section on Bayesian data analysis. Perspectives on Psychological Science, 6, 272–273. https://doi.org/10.1177/1745691611406926
(2014). Doing Bayesian data analysis: A tutorial with R, JAGS, and Stan (2nd ed.). San Diego, CA: Academic Press.
(2004). Evaluation of the Bayesian and maximum likelihood approaches in analyzing structural equation models with small sample sizes. Multivariate Behavioral Research, 39, 653–686. https://doi.org/10.1207/s15327906mbr3904_4
(2013). Longitudinal structural equation modeling. New York, NY: Guilford Press.
(2011). A 2 × 2 taxonomy of multilevel latent contextual models: Accuracy-bias trade-offs in full and partial error correction models. Psychological Methods, 16, 444. https://doi.org/10.1037/a0024376
(2005). Sufficient sample sizes for multilevel modeling. Methodology, 1, 86–92. https://doi.org/10.1027/1614-1881.1.3.86
(2016a). On using Bayesian methods to address small sample problems. Structural Equation Modeling, 23, 750–773. https://doi.org/10.1080/10705511.2016.1186549
(2016b). Using data-dependent priors to mitigate small sample bias in latent growth models: A discussion and illustration using Mplus. Journal of Educational and Behavioral Statistics, 41, 27–56. https://doi.org/10.3102/1076998615621299
(2009). A Monte Carlo sample size study: How many countries are needed for accurate multilevel SEM? Survey Research Methods, 3, 45–58. https://doi.org/10.18148/srm/2009.v3i1.666
(1997). General longitudinal modeling of individual differences in experimental designs: A latent variable framework for analysis and power estimation. Psychological Methods, 2, 371–402. https://doi.org/10.1037/1082-989X.2.4.371
(1998–2012). Mplus user’s guide (7th ed.). Los Angeles, CA: Muthén & Muthén.
(2002). How to use a Monte Carlo study to decide on sample size and determine power. Structural Equation Modeling, 9, 599–620. https://doi.org/10.1207/S15328007SEM0904_8
(2014). Working memory and alcohol use in at-risk adolescents: A 2-year follow-up. Alcoholism: Clinical and Experimental Research, 38, 1176–1183. https://doi.org/10.1111/acer.12339
(2015). R: A language and environment for statistical computing [Computer software manual]. Vienna, Austria. Retrieved from https://www.R-project.org/
. (2012). lavaan: An R package for structural equation modeling. Journal of Statistical Software, 48, 1–36. Retrieved from https://www.jstatsoft.org/article/view/v048i02
(2000). Latenttien kasvukäyrä- ja simplex-mallien teoriaa ja sovelluksia pitkittäisaineistoissa kehityksen ja muutoksen analysointiin
([Latent growth and simplex models: Theory and applications in longitudinal models for analysis of development and change] . Jyväskylä, Finland: Department of Statistics, University of Jyväskylä.2015). Analyzing small data sets using Bayesian estimation: The case of posttraumatic stress symptoms following mechanical ventilation in burn survivors. European Journal of Psychotraumatology, 6, 25216. https://doi.org/10.3402/ejpt.v6.25216
(2013). A gentle introduction to Bayesian analysis: Applications to developmental research. Child Development, 85, 842–860. https://doi.org/10.1111/cdev.12169
(2017). A systematic review of Bayesian articles in psychology: The last 25 years. Psychological Methods, 22, 217–239. https://doi.org/10.1037/met0000100
(2017). Where do priors come from? Applying guidelines to construct informative priors in small sample research. Research in Human Development, 14, 305–320. https://doi.org/10.1080/15427609.2017.1370966
(