The news media in 2008 warmed to a research study that found that warm temperatures led to interpersonal warmth. Not so fast.
In science, a single study’s findings, even the newsworthy, are considered just that: a single study’s findings. Repeating the study, a process called replication, is a rule of the research enterprise.
The journal Social Psychology recently invited researchers to submit study-replication proposals for a special issue dedicated to this fundamental exercise. Katherine Corker, assistant professor of psychology, co-authored one of 15 papers that were pre-approved for the issue published in May.
“This was the first journal issue in any scientific discipline to be composed entirely of pre-registered replications,” Corker said. “It’s important for us in psychology to have reproducible and cumulative science, and I wanted to do what I could to help be a part of that process.”
Corker and her colleagues ran three replication efforts of the popular 2008 study that suggested that exposure to warmth led to warm relations. Handling a warm beverage or therapeutic gel pack was associated in that study with generous and caring behavior.
Multiple media outlets embraced the study, capitalizing on the “warm-hands, warm-heart” theme of the results.
But in experiments led by Corker and her collaborators, those findings did not hold up. In the new study, researchers asked participants to evaluate a product – either a hot or cold therapeutic gel pack. Subjects then were asked to select a reward for their participation, and they could choose between a voucher for a drink or a cupcake for themselves or a voucher for a treat to give to a friend.
In three different cities, with almost 900 participants, the temperature of the gel pack made no difference in whether participants selected a reward for themselves or for a friend, meaning that exposure to physical warmth was not be linked to generosity.
Corker noted that while independent confirmation of initial results lends credibility to that earlier finding, a single replication is not the last word in research.
“The failure to replicate doesn’t necessarily mean the original hypothesis is not correct,” Corker said. “It just means that in a larger sample conducted independently across three sites, we found no support for the hypothesis. So the evidence for this effect does not seem particularly strong. We hope our effort continues to ignite interest in this area and that our peers will continue to explore it.”
The effort created a valuable research opportunity for four Kenyon psychology majors who assisted, including Katie Finnigan ’15 of Valencia, California; Maureen Hirt ’14 of Bloomington, Indiana; and Melek Spinel ’14 of Bogota, Columbia, who were Summer Science Scholars, and Olivia Siulagi ’14 of Portland, Oregon, a Summer Legal Scholar.
The special journal issue has generated debate among some in the field, but Corker is focused on the bigger picture. “Ultimately,” she said, “we all want the same thing – for our science to be reproducible so we can get a better understanding of human nature.”