How do you interpret scientific evidence?
Just yesterday, one of my colleagues posted a link to a paper from the American Journal of Clinical Nutrition
published in 2010 that analyzed data from a number of studies involving almost 350,000 people. The analysis suggests there is no significant relationship between heart disease and eating saturated fats. He seemed excited about this result, presumably because it supported his desire to eat fatty foods.
It is always nice to discover that something you hoped were true really is true. But, can your desire for an answer affect the way you evaluate the evidence?
This question was explored in a clever study by Anthony Bastardi, Eric Uhlmann, and Lee Ross published in the June, 2011 issue of Psychological Science. They examined how people evaluated new evidence when what they believed to be true conflicted with what they wanted to be true.
In this study, participants were people who expected to have children in the near future. All of them believed that caring for young children at home was better for the child than sending them to an outside daycare. Of these participants, half were people who expected they would send their child to daycare some day, while the other half were people who expected they would keep their child at home.
The experiment was conducted in a different session from when the participants expressed their beliefs about daycare and home care, and so it was not obvious to participants that this study was intended to be related to their existing beliefs or plans for the future.
In the experiment, everyone read one study that supported the conclusion that home care really is better than daycare. The other study supported the conclusion that daycare is better than home-care. The methods of the two studies were different. People were asked to evaluate the studies for whether the methods were valid and whether the studies were convincing.
Not surprisingly, the people who believed that home care is better and planned to care for their children at home believed that studies demonstrating that home care is best were more convincing than those demonstrating that daycare is best.
Those who planned to care for their children using daycare showed the opposite pattern. Even though they originally believed that home care is best, they found the study demonstrating daycare to be best to be more convincing than the study demonstrating home care to be best.
In many real-world situations, there is conflicting evidence from different studies. So, it is important to make judgments about which evidence is strongest. But, these results suggest that people are biased to interpret the evidence in ways that are consistent with their desires. That means that people may ultimately come to believe that the weight of evidence supports the position that they already wanted to believe was true. And they will believe this without recognizing that their own desires influenced the evaluation of the evidence.
Follow me on Twitter.