The Replication Crisis, and What It Means

Share this...
Share on FacebookShare on Google+Tweet about this on TwitterPin on PinterestShare on StumbleUponShare on TumblrShare on LinkedIn

Have you heard about the crisis?  It sounds dire, right?  The crisis is the replication crisis, meaning that previously accepted psychological studies are unable to be replicated.  What does this mean for the results and conclusions previously accepted as true?  What does this mean for the field of psychology?  Do we know anything about humans, and the way they think, feel, love, and belong?

The advancement of our field comes about through research study, with statistical significance powerful enough to make an impact.  Hypotheses are generated, and with controlled clinical studies, the outcome becomes a stepping stone for greater understanding, additional questions, and more research.  An assumption with research is that when the conditions are controlled enough, that research studies should be able to be replicated with exactly the same outcomes.  Yet as more and more researchers discover, only about half (or even less) of the psychological studies attempted to be replicated could be found to have the same outcome.

The reasons for this are as diverse as the studies themselves.  While accusations of faked results, biased reporting, and statistically too small sample sizes have been and should be taken seriously, this doesn’t account for the crisis itself.  As with many things, a more nuanced understanding is in order.  For you see, with psychology, the study subjects are almost always human beings.  And human beings are complex, diverse, and the variations between humans are hard to account for.

For example, lets say that a survey study conducted in 1960 was asking about trust in authority figures, such as bosses.  If that study were to be replicated today, many factors would no longer hold true.  Gender differences in participants would be dramatic; more women would be expected in the present day study than the 1960 one.  Additionally, the technology advancements, as well as cultural expectations and norms, have shifted as well.  Perhaps there is increased trust, as workers are permitted more flexibility, the autonomy to work remotely, and collaboration is encouraged within the workplace.  Perhaps there is decreased trust, as technology advancements have made it possible for bosses and authority figures to trace our browsing history and email correspondence at work.  Regardless, the results would not be the same as the 1960 study.  Another example is the famous “Marshmallow Test” in which young children were told they could have two marshmallows if they were able to wait in a room with just one, for up to 15 minutes.  The goal of the study was to measure self-control over temptation, with the idea that this would predict self-control for the future.  In truth, the ability to resist temptation at age 5 has little to do with adult self-control, and the children had diverse and varied means of reducing the temptation (or giving in to it).  While this is a cute study, the human variance was too much to make broad interpretations from.

Good has come from this replication crisis.  Researchers are more forthcoming, open, and collaborative in their research efforts and in sharing their results.  Publishers are looking to “pre-register” analysis, which aims to prevent interpretations of data that are twisted to support a hypothesis.  Efforts are being made to increase the control over the many variabilities within human subject study; additionally, efforts are being made to better understand, and account for these variations.

As with any other science, rigorous study and understanding is essential for making factual claims.  Psychology has long stood against accusations of being a “psuedoscience” and it is with honesty, careful and controlled study, and acknowledgement of our weaknesses that we stand among other researchers and scientists.  We belong in the scientific community, as there is a true desire to fully understand and explain human condition.  There is no real “replication crisis,” but rather an acknowledgement that humans change, hypotheses change, and we need psychological researchers to continue to understand these changes.

Share this...
Share on FacebookShare on Google+Tweet about this on TwitterPin on PinterestShare on StumbleUponShare on TumblrShare on LinkedIn