On the robustness of reciprocal associations between personality and religiosity in a German sample.

Abstract

Entringer etal. used longitudinal data from a German panel study to examine reciprocal causal effects between personality and religiosity, along with cultural moderators of these effects. The current paper examines the robustness of the original effects to alternative model specifications. We reanalyzed the same four-wave data spanning 12 years (total N = 46,316), first replicating the original cross-lagged panel analyses and then extending these analyses in three ways: Using a random-intercept cross-lagged panel model, using observed rather than latent variables, and modeling each trait individually rather than simultaneously. Correlations between personality and religiosity were all small in size, even when aggregating over 12 years. Lagged effects were very small, and none was robust across all model specifications. Cultural moderators also depended on model specifications. The very small size of these reciprocal effects, along with their sensitivity to model specifications, suggest that conclusions about causal effects of personality and religiosity should be drawn very cautiously.