Modeling reciprocal effects in medical research: Critical discussion on the current practices and potential alternative models
Autoři:
Satoshi Usami aff001; Naoya Todo aff002; Kou Murayama aff003
Působiště autorů:
Department of Education, University of Tokyo, Bunkyo-ku, Tokyo, Japan
aff001; Department of Psychology, University of Tsukuba, Tsukuba, Ibaraki, Japan
aff002; Department of Psychology, University of Reading, Reading Berkshire, United Kingdom
aff003; Research Institute, Kochi University of Technology, Kami, Kochi, Japan
aff004
Vyšlo v časopise:
PLoS ONE 14(9)
Kategorie:
Research Article
doi:
https://doi.org/10.1371/journal.pone.0209133
Souhrn
Longitudinal designs provide a strong inferential basis for uncovering reciprocal effects or causality between variables. For this analytic purpose, a cross-lagged panel model (CLPM) has been widely used in medical research, but the use of the CLPM has recently been criticized in methodological literature because parameter estimates in the CLPM conflate between-person and within-person processes. The aim of this study is to present some alternative models of the CLPM that can be used to examine reciprocal effects, and to illustrate potential consequences of ignoring the issue. A literature search, case studies, and simulation studies are used for this purpose. We examined more than 300 medical papers published since 2009 that applied cross-lagged longitudinal models, finding that in all studies only a single model (typically the CLPM) was performed and potential alternative models were not considered to test reciprocal effects. In 49% of the studies, only two time points were used, which makes it impossible to test alternative models. Case studies and simulation studies showed that the CLPM and alternative models often produce different (or even inconsistent) parameter estimates for reciprocal effects, suggesting that research that relies only on the CLPM may draw erroneous conclusions about the presence, predominance, and sign of reciprocal effects. Simulation studies also showed that alternative models are sometimes susceptible to improper solutions, even when reseachers do not misspecify the model.
Klíčová slova:
Adolescents – Behavior – Child development – Medicine and health sciences – Mental health and psychiatry – Simulation and modeling – Child psychiatry – Adolescent psychiatry
Zdroje
1. Nesselroade JR, Baltes PB. Longitudinal research in the study of behavior and development. New York: Academic Press; 1979.
2. Hsiao C. Analysis of Panel Data. London: Cambridge University Press; 2014.
3. Duncan OD. Some linear models for two-wave, two-variable panel analysis. Psychol. Bull 1969;72:177–182. doi: 10.1037/h0027876
4. Finkel SE. Causal Analysis with Panel Data. Thousand Oaks, CA: Sage; 1995.
5. Marsh HW, Yeung AS. Causal effects of academic self-concept on academic achievement—structural equation models of longitudinal data. J. Educ. Psychol. 1997;89:41–54. doi: 10.1037/0022-0663.89.1.41
6. Hamaker EL, Kuiper RM, Grasman RPPP. A critique of the cross-lagged panel model. Psychol. Methods 2015;20:102–116. http://dx.doi.org/10.1037/a0038889 25822208
7. Curran PJ, Bauer DJ. The disaggregation of within-person and between-person effects in longitudinal models of change. Annu. Rev. Psychol. 2011;62:583–619. doi: 10.1146/annurev.psych.093008.100356 19575624
8. Hamaker EL. Why researchers should think “within-person”: A paradigmatic rationale. in Handbook of research methods for studying daily life (ed. Matthias M and Tamlin C.) Guilford Press; 2012;43–61.
9. Hoffman L, Stawski RS. Persons as contexts: Evaluating between-person and within-person effects in longitudinal analysis. Res. Hum. Dev. 2009;6: 97–120. doi: 10.1080/15427600902911189
10. Kenny DA, Zautra A. The trait-state-error model for multiwave data. J. Consult. Clin. Psychol. 1995;63:52–59. doi: 10.1037//0022-006x.63.1.52 7896990
11. Kenny DA. Zautra A. Trait-state models for longitudinal data in New methods for the analysis of change (ed. Collins L. and Sayer A.) Washington, DC: American Psychological Association; 2001:243–263.
12. Usami S, Murayama K, Hamaker EL. A unified framework of longitudinal models to examine reciprocal relations. Psychol. Methods in press. doi: 10.1037/met0000210 30998041
13. Granger CWJ. Investigating causal relations by econometric models and cross-spectral methods. Econometrica 1969;37:424–438. doi: 10.2307/1912791
14. Cole DA, Martin NC, Steiger JH. Empirical and conceptual problems with longitudinal trait-state models: Introducing a trait-state-occasion model. Psychol. Methods 2005;10:3–20. doi: 10.1037/1082-989X.10.1.3 15810866
15. Luhmann M, Schimmack U, Eid M. Stability and variability in the relationship between subjective well-being and income. J. Res. Pers. 2011;45:186–197. doi: 10.1016/j.jrp.2011.01.004
16. Lüdtke O, Robitzsch A, Wagner J. More stable estimation of the STARTS model: A Bayesian approach using Markov Chain Monte Carlo techniques. Psychol. Methods 2018;23:570–593. doi: 10.1037/met0000155 29172612
17. Muthén LK, Muthén BO. Mplus User’s Guide. Los Angeles, CA: Muthén Muthén: 1998-2017.
18. Robinson WS. Ecological correlations and the behavior of individuals. Am. Sociol. Rev. 1950;15:351–357. doi: 10.2307/2087176
19. R Core Team. R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria; 2016. https://www.R-project.org/.
20. Rosseel Y. lavaan: An R package for structural equation modeling. J. Stat. Softw. 2012;48:1–36. http://www.jstatsoft.org/v48/i02/
21. Dormann C. & Griffin M. Optimal time lags in panel studies. Psychol. Methods 2015;20:489–505. doi: 10.1037/met0000041 26322999
22. Bringmann LF, et al. A network approach to psychopathology: New insights into clinical longitudinal data. PLoS ONE. 2013;8:e60188. doi: 10.1371/journal.pone.0060188 23593171
23. Schuurman NK, Ferrer E, de Boer-Sonnenschein M. & Hamaker E.L. How to compare cross-lagged associations in a multilevel autoregressive model. Psychol. Methods 2016;21:206–221. doi: 10.1037/met0000062 27045851
24. Richiardi L, Bellocco R, Zugna D. Mediation analysis in epidemiology: methods, interpretation and bias. Int. J. Epidemiol. 2013;42:1511–1519. doi: 10.1093/ije/dyt127 24019424
25. Ten Have TR, Joffe MM. A review of causal estimation of effects in mediation analyses. Stat. Methods Med. Res. 2012;21:77–107. doi: 10.1177/0962280210391076 21163849
26. VanderWeele TJ. Mediation analysis: A practitioner’s guide. Annu. Rev. Public Health 2016;37:17–32. doi: 10.1146/annurev-publhealth-032315-021402 26653405
27. Huang J, Yuan Y. Bayesian dynamic mediation analysis. Psychol. Methods 2017;22:667–686. doi: 10.1037/met0000073 27123750
28. Preacher K. Advances in mediation analysis: a survey and synthesis of new developments. Annu. Rev. Psychol. 2015;66:825–852. doi: 10.1146/annurev-psych-010814-015258 25148853
29. Maxwell SE, Cole DA. & Mitchell M.A. Bias in crosssectional analyses of longitudinal mediation: Partial and complete mediation under an autoregressive model. Multivariate Behav. Res. 2011;46:816–841. doi: 10.1080/00273171.2011.606716 26736047
Článek vyšel v časopise
PLOS One
2019 Číslo 9
- Proč jsou nemocnice nepřítelem spánku? A jak to změnit?
- Dlouhodobá ketodieta může poškozovat naše orgány
- „Jednohubky“ z klinického výzkumu – 2024/42
- Metamizol jako analgetikum první volby: kdy, pro koho, jak a proč?
- MUDr. Jana Horáková: Remise již dosahujeme u více než 80 % pacientů s myastenií