University Park, PA 16802
2012-Present: Associate Professor, Human Development and Family Studies, Pennsylvania State University
2007-2012: Assistant Professor, Department of Psychology, University of North Carolina
2005-2007: Assistant Professor, Department of Psychology, University of Notre Dame
2005-2006: Visiting Scientist, Max Planck Institute for Human Development, Berlin, Germany
2004-2005: Assistant Research Professor, Department of Psychology, University of Notre Dame
Sy-Miin Chow is a Professor in the Department of Human Development and Family Studies at the Pennsylvania State University. The focus of her work has been on developing and testing longitudinal and dynamic models, including differential equation models, time series models and state-space models. She has developed novel methods that serve as practical alternatives for addressing common data analytic problems (e.g., Bayesian variable selection methods for factor analysis and structural equation models, nonparametric approaches, missing data, and nonlinearities/nonstationarities in dynamics). She has also collaborated with scholars in emotion, aging, child development, family dynamics and prevention research to bring newer methodological developments to these areas. She is a winner of the Cattell Award from the Society of Multivariate Experimental Psychology (SMEP) as well as the Early Career Award from the Psychometric Society.
Descriptions of Ongoing Research
- Innovative Methods for Handling Data Analytic Issues in Intensive Longitudinal Data. Dynamic models are longitudinal models that are designed to describe more complex change processes. Due to these models’ explicit focus on process and dynamics, the associated data typically extend over substantially longer time spans (e.g., with greater than 35 measurement occasions) than those implicated in conventional panel models (typically, with less than 10 measurement occasions). One of my key research foci resides in developing methods for handling methodological issues encountered in the analysis of such data. Some of my representative work in this area involves fitting models with time-varying parameters (e.g., cyclic dynamics with time-varying amplitude; Chow, Hamaker, Fujita, & Boker, 2009; Chow, & Zhang, 2013; Chow, Zu, Shifren, & Zhang, 2011), developing continuous-time dynamic models for use with irregularly spaced data (obtained e.g., from experience sampling designs; Chow & Zhang, 2008), outlier detection in dynamic models (Chow, Hamaker, & Allaire, 2009), comparing the state-space modeling approach—one common modeling framework for formulating dynamic models—with other well-known modeling frameworks such as structural equation modeling (Chow, Ho, Hamaker, & Dolan, 2010), and developing methods for handling sudden shifts (regime switches) in human dynamics (Chow, Grimm, Guillaume, Dolan, & McArdle, 2013; Chow, & Fileau, 2010; Chow, & Zhang, 2013).
- Methods for Fitting Nonlinear Dynamical Systems Models. With few exceptions, dynamic models in the behavioral sciences have focused almost exclusively on linear changes. As such, most of the existing software programs in social and behavioral sciences cannot readily handle models with nonlinear relationships among latent variables. Often, many nonlinear constraints have to be explicitly specified and some of the commonly utilized approaches can quickly become cumbersome. Such issues are especially salient in fitting differential equation models with nonlinear changes at the latent level. While differential equation models in general have special utility in accommodating irregular measurement intervals, the mathematical constraints involved are highly complex. To this end, my collaborators and I have extended existing techniques as well as developing new ones for fitting nonlinear dynamical systems models (Chow, Bendezú, Cole, & Ram, in press; Chow, Lu, Sherwood, & Zhu, 2016; Chow, Tang, Yuan, Song, and Zhu, 2011; Chow, Witkiewitz, Grasman, & Maisto, 2015).
- Dynamic Models of Emotions, Lifespan Development and Family Dynamics. My methodological interests are motivated in part by empirical data analytic problems. There has been an emerging consensus that more sophisticated dynamic modeling tools are needed to better capture the complexities of different change processes. To this end, I have presented several novel applications of dynamic modeling techniques to represent affective processes (Chow, Ram, Boker, Fujita, & Clore, 2005; Chow, Nesselroade, Shifren, & McArdle, 2004; Yang, & Chow, 2010), interaction between parent-infant dyads (Chow, Haltigan, & Messinger, 2010; Messinger, Mahoor, Chow, & Cohn, 2009), family dynamics (Feinberg, Xia, Fosco, & Chow, under review; Schermerhorn, Chow, & Cummings, 2010) and lifespan development (Chow, Hamagami, & Nesselroade, 2008; Cole, Bendezú, Ram, & Chow, under review).
*Lu, Z-H., Chow, S-M., & Loken, E. (in press). Bayesian factor analysis as a variable selection problem. Multivariate Behavioral Research.
Chow, S-M., *Bendezú, J. J., Cole, P. M., & Ram, N. (in press). A Comparison of Two- Stage Approaches for Fitting Nonlinear Ordinary Differential Equation (ODE) Models with Mixed Effects. Multivariate Behavioral Research.
Chow, S-M., *Ou, O. Cohn, J. F., & Messinger D. S. (in press). Representing Self-Organization and Non-Stationarities in Dyadic Interaction Processes Using Dynamic Systems Modeling Techniques. In Von Davier, A. Kyllonen, P. C., & Zhu, M. Innovative Assessment of Collaboration. New York: Springer.
Chow, S-M., *Lu, Z., Sherwood, A. & Zhu, H. (2016). Fitting linear and nonlinear differential equation models with random effects and unknown initial conditions using the stochastic approximation expectation-maximization (SAEM) algorithm. Psychometrika, 81(1), 102-134. Doi:10.1007/s11336-014-9431-z. PubMed #: 25416456
*Lu, Z-H., Chow, S-M., Sherwood, A, & Zhu, H. (2015). Bayesian analysis of ambulatory cardiovascular dynamics with application to irregularly spaced sparse data. Annals of Applied Statistics, 9(3), 1601-1620. Doi: 10.1214/15-AOAS846.
Chow, S-M., Witkiewitz, K. Grasman, R., & Maisto, S. (2015). The Cusp Catastrophe Model as Cross-Sectional and Longitudinal Mixture Structural Equation Models. Psychological Methods, 20, 142-164. PubMed # 25822209 NIHMSID 667553
*Hutton, R. S., & Chow, S-M. (2014). Longitudinal Multi-Trait-State-Method model using ordinal data. Multivariate Behavioral Research, 49, 269-282.
Zhang, G., Browne, W. M., Ong, A. D., & Chow, S.-M. (2014). Analytic standard errors for exploratory process factor analysis. Psychometrika,79(3), 444-469.
Chow, S-M., Witkiewitz, K., Grasman, R., Hutton, R. S. & Maisto, Stephen. (2014). A regime-switching longitudinal model of alcohol lapse-relapse. In P. C. M. Molenaar, K. M. Newell, & R. M. Lerner, Handbook of Relational Developmental Systems: Emerging Methods and Concepts (pp. 397-422). New York: Guilford Publications, Inc.
Messinger, Daniel S., Mahoor, M. H., Chow, S-M., Haltigan, J. D., Vadavid, S., & Cohn,J. F. (2013). Early emotional communication: Novel approaches to interaction. Social Emotions in Nature and Artifact, 162
Chow, S-M., Grimm, K. J., *Guillaume, F., Dolan, C. V, & McArdle, J. J. (2013). Regime-switching bivariate dual change score model.Multivariate Behavioral Research, 48(4), 463-502.
Chow, S-M, & Zhang, G. (2013). Regime-switching nonlinear dynamic factor analysis models. Psychometrika, 78(4), 740-768.
Development and adaptation of modeling and analysis tools that are suited to evaluating linear and nonlinear dynamical systems models, including longitudinal structural equation models and state-space modeling techniques.