Analysis of Nonlinear Patterns of Change with Random Coefficient Models

Abstract

Nonlinear patterns of change arise frequently in the analysis of repeated measures from longitudinal studies in psychology. The main feature of nonlinear development is that change is more rapid in some periods than in others. There generally also are strong individual differences, so although there is a general similarity of patterns for different persons over time, individuals exhibit substantial heterogeneity in their particular response. To describe data of this kind, researchers have extended the random coefficient model to accommodate nonlinear trajectories of change. It can often produce a statistically satisfying account of subject-specific development. In this review we describe and illustrate the main ideas of the nonlinear random coefficient model with concrete examples.

Predictors of change in postmerger identification during a merger process: A longitudinal study.

“…Although we did not predict such a non-linear change in identification, the result is not surprising given the fact that many behavioral processes exhibit differential rates of change (Cudek & Harring, 2007). The non-linear increase of post-merger identification points to the fact that change is not uniform over time.…”

Pigeon and human performance in a multi-armed bandit task in response to changes in variable interval schedules

“…The equation generates eight probabilities, one for each of the eight disks, that sum to 1.0. To fit Luce's decision rule to behavior, we used nonlinear mixed-effects modeling and identified the maximum likelihood best-fitting parameter values (Cudeck & Harring, 2007;Davidian & Giltinan, 2003). Mixed-effects modeling is used to simultaneously generate estimates of parameter estimates for each subject and as a function of the independent variables (e.g., Laird & Ware, 1982;Pinheiro & Bates, 2004).…”

Fitting correlated residual error structures in nonlinear mixed-effects models using SAS PROC NLMIXED

“…Toward that end, a submodel may be specified for each. At the population level (level 2), a general specification of an individual coefficient is a potentially nonlinear function of fixed parameters, (β ), covariates (z i ), and random effects (b i ) (see, e.g., Cudeck & Harring, 2007). For individual coefficient k, this is…”