I guess I wasn’t as interested in the p value as the direction of sub field difference, which I have replicated pretty much every year in my methods course. (Confounded with within/btwn design, num items, etc). One old example here