“…However, low alphas tend to underestimate effects, thus effects found in this study could have been stronger, not weaker, if the scores had demonstrated higher reliability (e.g., Furr & Bacharach, ). Finally, the lack of direct association between religiousness at Time 1 and risk‐taking at Time 3 is surprising, given the prior finding of modest yet robust association between religiousness and externalizing behaviors (e.g., Holmes & Kim‐Spoon, 2016a). In this community sample of early to middle adolescents, risk‐taking behavior prevalence was relatively low, resulting in limited variance, which may have contributed to the dampened longitudinal effect of religiousness.…”