Short Communication - Journal of Neuroinformatics and Neuroimaging (2022) Volume 7, Issue 1

Neuroimaging: Capturing nonlinear relationships with longitudinal studies.

Abstract

There are two types of variables in statistical modeling. The first type is categorical (e.g., sex, group, condition), and its possible categories are treated as the levels of a factor. The other type is quantitative, in which numerical values represent measurements. Sometimes a variable can be treated one way or another depending on the research focus and hypothesis. For example, when a group of subjects are scanned once during each of five consecutive years, the five time points can be modeled as a factor with 5 levels when the differences among them, irrespective of order, are of interest.