Key points are not available for this paper at this time.
Gervais & Norenzayan (2012) reported in Science a series of 4 experiments in which manipulations intended to foster analytic thinking decreased religious belief. We conducted a precise, large, multi-site pre-registered replication of one of these experiments. We observed little to no effect of the experimental manipulation on religious belief (d = 0.07 in the wrong direction, 95% CI-0.12, 0.25, N = 941). The original finding does not seem to provide reliable or valid evidence that analytic thinking causes a decrease in religious belief.
Sanchez et al. (2017) studied this question.