I agree with you, the effect size is relatively small, here's the main table: http://www.nature.com/mp/journal/vaop/ncurrent/fig_tab/mp2016213t3.html#figure-ti

I agree with you, the effect size is relatively small, here's the main table: http://www.nature.com/mp/journal/vaop/ncurrent/fig_tab/mp2016213t3.html#figure-title http://www.nature.com/mp/journal/vaop/ncurrent/fig_tab/mp201... Beta is the effect size here, which ranges from 0.07 to 0.03, not exactly high. The p-values may be tiny but that's just a measure of the likelihood of getting these effect sizes by chance. >These results imply that, after correcting for confounding variables (see notes under Table 3), those who were deficient at time of birth had on average 0.06 point higher gender-weighted SRS score, which corresponds to an increase of approximately one fourth of the s.d. (s.d. of the SRS in the total sample=0.23 Now is 0.06 a large growth for SRS? Let's have a look at the scale of that value: >The SRS-A mean total score was highest in the ASD sample [autism disorder] (M = 78.5, SD = 13.7), intermediate in the CLIN [other disorders] (M = 63.4, SD = 15.4), and lowest in the TD [typically developing, 'normal ] (M = 55.5, SD = 9.9) https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3425739/ https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3425739/ So you have about 10 points on average between these groups, an effect size of 0.06 doesn't seem extreme to me (but it is nevertheless linked!)