In a large representative sample of Swedish male conscripts (N = 49,246), we demonstrate a positive association between measured intelligence and response consistency on items measuring endorsement of a strong military defense. This association is accentuated, to some degree, by an apparent difficulty to handle reversed items among those with low measured intelligence. Consequently, due to regression toward the mean, reversed items in measurement instruments would tend to have a negative effect on statistical power when analyzing the association between whatever the instrument is measuring and intelligence. On the other hand, the presence of reversed items could mitigate bias in the measurement due to an acquiescent response style among those with low measured intelligence. Lack of motivation/carelessness and inadequate reading ability among some subjects are two possible confounders for the found association in the present study.