Modeling Retest Effects in a Longitudinal Measurement Burst Study of Memory.

ABSTRACT: Longitudinal designs must deal with the confound between increasing age and increasing task experience (i.e., retest effects). Most existing methods for disentangling these factors rely on large sample sizes and are impractical for smaller scale projects. Here, we show that a measurement burst design combined with a model of retest effects can be used to study age-related change with modest sample sizes. A combined model of age-related change and retest-related effects was developed. In a simulation experiment, we show that with sample sizes as small as n = 8, the model can reliably detect age effects of the size reported in the longitudinal literature while avoiding false positives when there is no age effect. We applied the model to data from a measurement burst study in which eig

SUBMITTER: Broitman AW 

PROVIDER: S-EPMC7717555 | biostudies-literature | 2020 Jun

REPOSITORIES: biostudies-literature