Dynamic modeling and the individual
State of the art talk by Ellen Hamaker, Department of Methodology and Statistics, Utrecht University, Netherlands
Chair: Sy-Miin Chow, Tuesday 21st July, 17.15 - 18.00, Palmeston Lecture Theatre, Fisher Building.
It is common practice in psychological research to consider large representative samples. While this approach ensures that the results can be generalized to the population from which the sample was obtained, such results are not necessarily informative if one is interested in psychological processes that take place at the level of the individual. An alternative approach which is especially useful for modeling processes is time series analysis.
In this presentation I begin with giving a brief historic account of why psychologists and psychometricians have become so focused on populations instead of the individual. Next I discuss three applications of time series analysis in psychological research, which illustrate the usefulness of this technique for the study of processes. These applications consist of modeling: a) the symptomatology of children who are treated with neurofeedback; b) dyadic interactions; and c) mood swings in patients with bipolar disorder. I end by discussing several recent developments that help bridge the gap between the standard large sample approach and dynamic modeling.