No, this belief is rooted in the observation that many social science fields, like social psychology, have replication rates in the 20% [1] range and are rife with obviously broken experimental methodology - hey let's go poll Amazon Mechanical Turk penny poll pros, and extrapolate it to everybody - perfect! And note that those replication studies are not picking low hanging fruit, they're going after the some biggest, highest impact factor journals there are in these fields. And so what that tells you is that if you take a given study in these fields, took what was said and assumed the exact opposite (or at least that it was not statistically relevant) you'd be right overwhelmingly more often than somebody who took these studies and went 'wow, that's what the science says I guess.'
Furthermore, there is a difference between something being difficult and it being impossible. Many don't realize that astrology was, for many centuries a "real" science, at least as real as modern social science is. It only ended, with some irony, because of the Church. They claimed that astrology's claims to be able to effectively divine the future was heretical and blasphemous, which led to astrology to go from a staple of education and science to being a fringe pseudoscience used for entertainment purposes only. I've no doubt that many practitioners of the time did genuinely feel that what they were doing was simply very difficult. And in fact there probably are some incidental correlations one can draw based on groups across time, so they would be able to show some "real" results, even if the entire field was completely bunk.
Furthermore, there is a difference between something being difficult and it being impossible. Many don't realize that astrology was, for many centuries a "real" science, at least as real as modern social science is. It only ended, with some irony, because of the Church. They claimed that astrology's claims to be able to effectively divine the future was heretical and blasphemous, which led to astrology to go from a staple of education and science to being a fringe pseudoscience used for entertainment purposes only. I've no doubt that many practitioners of the time did genuinely feel that what they were doing was simply very difficult. And in fact there probably are some incidental correlations one can draw based on groups across time, so they would be able to show some "real" results, even if the entire field was completely bunk.
[1] - https://en.wikipedia.org/wiki/Replication_crisis#In_psycholo...