Isn't that because the replications only looked at a selected subset of all the possible literature? You can be almost sure that if an article's conclusion hing

Isn't that because the replications only looked at a selected subset of all the possible literature? You can be almost sure that if an article's conclusion hinges on a wide interpretation of the experimental result, or the stimuli haven't been sampled properly (and who knows the distribution of stimuli?) or the subjects are first year psych students, and the proof is a rejection of the null hypothesis, that it cannot be replicated. The worst offenders are those that conclude their theory is true because the theory they argue against is rejected.