Networks and directed acyclic graphs : Initial steps to efficiently examine causal relations between expectancies, values, and prior achievement

Abstract

The present study explored a set of plausible directed acyclic graphs (DAGs) of constructs involved in Situated Expectancy–Value Theory (SEVT) using cross-sectional data. To do so, three datasets (n = 1,540; 1,867; and 103) with expectancy, values, and prior achievement constructs were used. First, networks showed a consistent magnitude of associations between SEVT constructs across datasets. Expectancy was associated with achievement (r = .21; 0.29; 0.37) and intrinsic value (r = .35; 0.68; 0.60); intrinsic value was associated with attainment value (r = .31; 0.79; 0.61). We reason through a set of plausible DAGs, or hypothesized directional models. Across datasets, DAGs revealed that prior achievement relates to expectancy, intrinsic value relates to expectancy, and utility relates to attainment value. This study highlights that DAGs can be used as a step to help triangulate, in conjunction with experimental and longitudinal work, on the underlying processes shaping relations between constructs.