Veronica Cole

My program of research focuses on the measurement of social, emotional, and academic development in adolescents. Most recently, this has entailed two major questions:

1. How do we measure peer connectedness? For a long time, researchers have been concerned about whether and how an adolescent’s overall level of connectedness to their peers relates to their risk behavior. But this general construct, peer connectedness, encompasses many different things, including (among many others) the number of friends an adolescent has, the amount of social support they receive, the amount of influence they have over their peer network, the extent to which others hold them in high esteem. This makes it very difficult to generalize, and it underscores the need to be precise with our language when considering what aspects of the peer ecosystem are related to risky behaviors such as substance misuse.

2. How much does measurement invariance matter? When we are measuring a latent construct using some set of items, we generally assume that the items measure that construct equally well across individuals, regardless of other differences between them. Fortunately, we have some models which can be used to accommodate violations of this assumption. But can these models still yield accurate estimates even if we misspecify them? And how do we know when our model is “good enough” to yield unbiased estimates of individuals’ scores? My lab has tried to assess these questions through simulation and empirical data analyses — we often fit multiple models to a given dataset and examine whether and how our results change, which can be very informative!

Representative Publications

• Cole, V. T., Hussong, A. M., McNeish, D. M., Ennett, S. T., Rothenberg, W. A., Gottfredson, N. C., Faris, R. W. (2021). The Self-Medication Pathway to Smoking for Adolescents: Interactions between Depressive Symptoms, Coping Motives for Smoking, and Social Standing. In press at Journal of Studies on Alcohol and Drugs.
• Cole, V. T., & Hussong, A. M. (2020). Psychosocial functioning among college students who misuse stimulants versus other drugs. Addictive behaviors, 105, 106290.
• Bauer, D. J., Belzak, W., Cole, V. T. (2020). Simplifying the Assessment of Measurement Invariance over Multiple Background Variables: Using Moderated Nonlinear Factor Analysis to Detect Differential Item Functioning. Under review at Structural Equation Modeling: A Multidisciplinary Journal, 27(1), 43-55.
• Hussong, A. M., Ennett, S. T., McNeish, D. M., Cole, V. T., Gottfredson, N. C., Rothenberg, W. A., & Faris, R. J. (2020). Social network isolation mediates associations between risky symptoms and substance use in the high school transition. Development and psychopathology, 32(2), 615-630.
• Cole. V. T., Bauer, D. J., & Hussong, A. M. (2019) Assessing the Robustness of Mixture Models to Measurement Non-Invariance. Multivariate Behavioral Research, 24(6).
• Cole, V. T., Hussong, A. M., Faris, R. W., Rothenberg, W. A., Gottfredson, N. C., & Ennett, S. T. (2019). A Latent Variable Approach to Modeling Social Dynamics in Adolescence. Journal of Research on Adolescence.