Why Do You Always Make Me Feel (insert emotion here)?

From the movie office space

From the movie office space

Are there certain people in your life that elicit strong emotions every time you see them, positive or negative?  Have you ever wondered if that same person drives everyone else crazy too, or just you?   Today, everydata delves into two recent statistical studies on what is called "affective emotions"--the idea that the same people elicit the same emotional response from many different individuals.

In a study highlighted in this recent article on Yahoo Health,  researchers at the University of Sheffield conducted an experiment involving 40 people who took part in a speed-dating event.  At the end of 4-minute speed dates, each participant described the emotions that were invoked by their dating partner.   The study found, among other things, that positive emotional responses tended to be correlated across individuals.  The strongest consistent emotions evoked were boredom and enthusiasm.

In a prior study, researchers at the University of Pennsylvania and Washington University studied the first year class of MBA students at a major University taking an Organizational Behavior class.  MBA students in the same work group "enrolled in all the same classes, worked together on every group project, and frequently socialized outside of class."  The study found that individuals own trait effects (how they felt about themselves) was in some cases as large as how they felt about others consistently (particularly for negative emotions).

As a consumer of data, what does one make of these studies?  I am not a psychology researcher, so in some respects, my reading and summarizing these studies is very similar to that of a lay person.  A few things to consider as you assess what to make of these studies:

First, I would flag what economists call external validity.  The two studies are conducted on very selected samples of individuals-- MBAs and college students in a bar in London.  That doesn't invalidate the studies, but it does suggest that one should think carefully about how these studies extend to the more general population.   For example, a speed dating setting may be one in which people are particularly attune to how someone else makes them feel as they are assessing potential romantic interests (something the authors address in the studies).

Second, these studies, by necessity, are conducted on somewhat subjective measures of emotions.   It is important to think about what that might mean--does the fact that participants chose among a limited set of possible emotional responses somehow influence the results or the correlation in emotional responses?

One other more subtle point--both studies are trying to capture variation in emotional response.   Both studies are reporting how much of the overall variation observed in the survey responses are explained by one's emotional reaction to others.  What constitutes a high or low level of explanatory power is an interesting question as well, particularly in studies with relative small sample sizes.

Overall, my favorite part of these studies is they are academic studies tackling challenging and interesting questions in a rigorous way.  But, notice the actual issues are far more complicated and far less definitive than the Yahoo Health story might lead you to believe.