An academic study of academic studies recently found that management research isn't always honest, the Wall Street Journal reports.

The study measured discrepancies between management studies when they are written as dissertations and when they are eventually published in academic journals. It found that at the dissertation level, papers' hypotheses were disproven about 56 percent of the time. However, by the time the study advanced to publication, hypotheses were disproven just 34 percent of the time.

The findings don't just suggest that prideful researchers might switch their hypothesis to make them look smarter when they publish a piece. A broader implication is that it indicates that academics are willing to shift the narrative of their study to fit their narrative.

One Size Does Not Fit All

Here's the thing about that, though: Management science is always messy. While shfting hypotheses aren't a great look for an academic, neither is buying wholesale into what any study says for a leader.

Consider, if you will, that research--be it management or any other discipline--rarely creates one size that fits all. In general, if a study shows that a given tactic or strategy works at 60 percent of the time, that means the tactic had an effect.

But while that shows a pretty high success rate, it also means the strategy at hand didn't work at 40 percent of the companies that were included in that study. The reasons why might vary, from industry to company culture to dumb luck. For all the virtues of data and research, a little bit of human intuition still goes a long way.

There are myriad reasons to take academic research with a little bit of salt as you consider whether a given finding would actually hold true at your company. It doesn't mean research isn't useful. It just means its findings aren't always prescriptive, regardless of whether or not a hypothesis was toyed with.