The world of academic psychology was sent into a tailspin earlier today after it emerged that over half of psychological experiments could not be replicated.
The disturbing findings, which were published in the journal Science, concluded a year-long process of reviewing 100 studies that had been previously published in academic journals. These studies formed part of the core knowledge by which psychologists understand people and their relationship to the world.
While repeating these experiments, over 60% of the results could not be replicated. What makes this so surprising is that the team of experimenters worked closely with the original researchers to reproduce the conditions and methods of the original studies.
It does not appear that data in the original studies was deliberately falsified. Writing about these results in an article ‘A Lot of Published Psychology Results are Bullshit‘, Maddie Stone commented,
But most of the published works examined in the new Science paper were not maliciously manipulated. Rather, on re-analysis, the findings simply weren’t as strong or convincing as the authors had originally claimed. There are a lot of possible explanations here, including issues with sample size and experimental design, or misinterpretation and misuse of statistical tools. These are problems that have long plagued experimental, data-driven sciences, from economics to cancer research:
The news is likely to add momentum to the growing skepticism about the social sciences among the non-academic public. However, the news comes as a charge for psychologists to strengthen standard research methods as well as to move to greater acknowledgement that there can be flexibility in the way data is interpreted. Moreover, the limitations of quantitative research may lead to greater receptivity to other models of understanding behavior, such as grounded theory.