© 2024
Virginia's Public Radio
Play Live Radio
Next Up:
0:00
0:00
0:00 0:00
Available On Air Stations

Study: Findings Might Not Be Reliable Over Time

Kaiser Health News

Year after year, we discover new things about human behavior, the ways we learn, think and behave, but how accurate are those studies?  A team in Virginia set out to answer that question -to see if the study were done again, would findings be the same.  What they found has shaken the world of social science.

The Center for Open Science recruited colleagues from around the world to try and replicate a hundred studies published in three scientific journals in 2008 - studies that looked at personality, relationships, learning and memory.  270 scientists from 17 countries signed up, and Johanna Cohoon says the subjects varied widely.

“Some involved toddlers.  Some were eye tracking studies: where do you look under certain circumstances, and another looked at the correlation between SAT scores and participants’ response to a puzzle.”

Three and a half years later, in the journal Science, they reported that 75% of social psychology experiments and half of studies looking at learning and memory came up with very different findings, but the center’s Mallory Kidwell does not conclude the original studies were wrong.

“I think that in order to understand the phenomena that we’re testing, we need to look at these phenomena over a long period of time, using several tests and several studies to examine them.”

Some social scientists described the findings as devastating, and they worried that therapists and educators rely on these studies to guide them, but the man who launched The Reproducibility Project - UVA Professor Brian Nosek - hopes it will spur social scientists to better understand the research process and to fix problems. 

Next up for the Center - a study of widely cited results in cancer.

Sandy Hausman is Radio IQ's Charlottesville Bureau Chief
Related Content