Researchers are split over guilt. Many of them think that guilt is negative—it feels bad, it’s related to poor functioning, and it’s something we should reduce in our lives. (That may be your assessment, too.) 

But another group of researchers suggests that guilt is good. It leads people to take actions to repair relationships and engage in prosocial behavior. Our attempts to get rid of guilt lead to good behavior—and ultimately the guilt experience and response is positive. 

In her talk “Connecting Affective Traits to their Affective States: A Tale of Two Guilts” at the symposium “Bridging Personality and Social Perspectives: Connections Between Traits and Emotions,” Stefanie Tignor tried to unravel this puzzle by attending to what some might consider the “soft underbelly” of social psychology: measurement. 

The constructs and processes that social psychologists research are complex and myriad, and new theoretical insights or hypotheses often lead people to develop measurement tools specifically suited to their needs. However, when these measurement tools are described in the literature, researchers label them according to the higher order concepts they are meant to tap into. 

For example, in Baumeister, Bratslavsky, Muraven, and Tice’s 1998 paper on ego depletion, the authors measured willpower four different ways in four different studies. One of them involved time spent on an unsolvable problem after having to eat a bowl of radishes—while looking longingly at a nearby bowl of cookies. Another involved number of anagram puzzles solved after watching a funny or sad movie without being allowed to express emotions. Both effects were labeled “ego depletion” because they intuitively seem to tap a similar ability: the willingness to perform a difficult action after having had to inhibit a dominant response. 

However, when a recent large-scale replication of an ego depletion study found a null effect, Baumeister was correct to point out that only one of many ways of measuring ego depletion showed no effect. It could be possible that other versions of the task lead to other conclusions. Really resolving debates in the psychological literature like this involve not just examining the concept that the researcher wanted to measure, but the different ways researchers actually measure them. 

In dealing with the conflicting accounts of guilt, Tignor saw a pattern in what types of measures different researchers were using to assess guilt. Those who said guilt was bad tended to use checklist items. These measures asked people how they tend to feel, devoid of context. 

Those who said guilt was good, on the other hand, tended to use scenario based measures. These emphasized what a participant would do in a given situation, emphasizing responses to feelings. 

In her studies, Tignor asked participants to fill out both kinds of questionnaires, as well as questionnaires about a suite of other adaptive and maladaptive traits. She found that checklist scales of guilt correlated most highly with maladaptive qualities like depression, low self-esteem, or low self-acceptance. Scenario scales, on the other hand, correlated most highly with adaptive qualities like agreeableness, caring about personal growth, and having high empathic concern for others. 

Further, the traits that correlated highly with one measure of guilt didn’t correlate highly with the other. The different ways of measuring guilt were pointing to different personal profiles. 

Finally, Tignor examined how the different ways of measuring guilt predicted day-to-day experiences. The checklist scales were associated with lower day to day ratings of well-being and more negative feelings. The scenario scales, on the other hand, were associated with high well-being and higher frequency of performing prosocial behaviors like being nurturing to a friend or volunteering. 

Overall, these findings suggest that the argument about guilt wasn’t about a single trait—the tendency to feel guilty—at all. Instead, they suggest that researchers have been measuring two different things. One is centered around a generalized emotional response, and this construct—feeling guilty all the time—is related to negative outcomes. Another is centered around specific responses to situations where you’re guilty and this construct—guilting yourself into helping others—is related to positive outcomes. 

The argument in the literature wasn’t about guilt at all. It was about people try to use the same intuitive label—guilt—for two different things. 

While it may be tempting to assume that because two researchers say they are measuring the same thing, we should treat this claim with skepticism. We need psychological research precisely because our lay intuitions about thinking and feeling don’t always map onto what’s really happening in the mind. Instead of arguing about whose perspective is right, it might be better to take a step back and ask what researchers are really measuring.

Alex Danvers is a graduate student in social psychology at Arizona State University. He studies emotions and social interaction from a dynamic systems perspective.