Results indicate that although high self-control is associated with a wide range of socially desirable traits, choice of social partners does not depend or depends only to a small degree on the partner's high self-control

The Upsides and Downsides of High Self-Control: Evidence for Effects of Similarity and Situation Dependency. Lukas Röseler, Jacqueline Ebert, Astrid Schütz, Roy Baumeister. Europe's Journal of Psychology, Vol 17(1). Feb 26 2021. https://ejop.psychopen.eu/index.php/ejop/article/view/2639

Abstract: High trait self-control is generally depicted as favorable. We investigated whether this holds for social perception. Using vignettes, we tested whether a person with high self-control is 1) preferred as a partner for all or only certain social situations, 2) perceived as less likeable than a person with low self-control, 3) liked more if the person is female and the behavior thus fits the sex-stereotype, and 4) perceived differently from a person with low self-control with respect to a wide range of adjectives used to describe personality. Competing theories are presented for each area. Results indicate that although high self-control is associated with a wide range of socially desirable traits, choice of partners 1) depends on the type of situation in which the interaction will occur, 2) depends on the similarity between the respondent and the partner, 3) does not depend on a stereotype match, and 4) does not depend or depends only to a small degree on the partner's high self-control. The perception of individuals with high self-control is thus variable and situationally contingent, and more than a single theory is needed to explain it.

Discussion

A broad and high-powered vignette study provided detailed evidence for a wide range of effects of a person's self-control on social perception. Contrary to the view that high self-control is always better than low self-control (positivity hypothesis of self-control), we found that the type of person that is preferred 1) depends on the type of situation one is confronted with, 2) depends on similarity, and 3) does not depend on gender stereotypes. In terms of duty situations (e.g., getting advice for a decision), people with high self-control were the preferred partners, whereas in socializing situations (e.g., going for a walk, attending a party, having a nice conversation), people with low self-control were the preferred partners. Also contradicting the positivity hypothesis of self-control, we found that how much people liked people with low or high self-control depended on their own self-control. More specifically, people with high self-control preferred other people with high self-control, and people with low self-control preferred other people with low self-control, a trend that is consistent with similarity being a major factor for determining attraction in nonromantic contexts.

There was an apparent contradiction of the main effect of self-control: When the task was to choose a partner for socializing and duty situations, participants preferred partners with high self-control overall (high vignette self-control). However, when we asked for how much participants liked the vignettes and participants' self-control was controlled for, low vignette self-control received higher liking scores overall. Note that choosing somebody in a situation differs from liking somebody and that the overall mean effect across all 10 situations depended on these situations and could easily be different for other situations. For example, the duty situations might have been more diagnostic than the sociality situations, but at the same time, they might occur less often and thus weigh less.

Although there is evidence for stereotypical male behavior being less self-controlled than female behavior, and there are theoretical grounds for predicting that a stereotype match would be positively associated with how much a person likes another person, we found no such effect: Whether men or women were presented as high or low in self-control had no differential effect on liking.

We found that the similarity effect and the situation effect occurred simultaneously and independently and dominated the main effect of vignette self-control (positivity hypothesis) by far. That is, the choice of a partner depends on the situation that the partner is needed for (situation hypothesis), the resemblance between the partner and the person who is choosing the partner (complementarity hypothesis), but not or only to a very small degree on the self-control of the partner being high (positivity hypothesis). Future research should consider these aspects of social perception in relation to self-control. We did find an advantage of people with high self-control over people with low self-control when people look for partners in duty situations. Despite this corroboration of the positivity hypothesis, other factors such as similarity and the situation seem more important. This is surprising given that the definition of self-control includes aspects of social desirability. In other words, whether one likes somebody who is socially desirable (i.e., has high self-control) also depends on the perceiver's self-control. And whether one wants to work with somebody who is socially desirable depends on the situation. Ironically, in what we termed socializing situations (e.g., going on vacation together or partying), vignette characters with low self-control (i.e., a socially undesirable trait) were preferred over vignette characters with high self-control.

Finally, and more in line with the positivity hypothesis of self-control, an exploratory approach revealed that people with high self-control are associated with a wide range of other socially desirable traits, such as being assured, unassuming, and agreeable. Some of the results are seemingly contradictory: Although some factors theoretically correlate negatively (e.g., assured-dominant and gregarious-extraverted), differences between high and low self-controlled vignettes are not in the same direction with these factors. This could be due to patterns that are different in that population, possibly due to trans situational variability, than in a standard population. That is, people very high and low in self-control or the perceptions about these people could show other patterns of correlations between the dimensions that differ from those in a standard population.

Limitations

In our outline of the social perception of people with high self-control, we considered different kinds of situations, individual differences in the participants, sex stereotypes, and a wide range of traits. However, several factors merit further investigation.

First, we used vignettes only. Our setting was thus highly artificial, and the large effect sizes were probably due to the extreme manipulations. If actual people with less extreme differences were used, we would expect smaller effects.

Second, our study may have been affected by demand characteristics. It was evident that we were interested in perceptions that were based on the vignettes and on stereotypes. Even though people are not defenseless against demand characteristics (Hofmann, Gschwendner, Friese, Wiers, & Schmitt, 2008) judgments in everyday life may differ from findings in an experiment. The ratings of the vignettes might be subject to demand characteristics whereas spontaneous responses might not. Although tasks such as the Implicit Association Test (Greenwald, McGhee, & Schwartz, 1998) have been criticized heavily (e.g., Schimmack, 2019), they allow for an assessment of such spontaneous reports (Hofmann, Gawronski, Gschwendner, Le, & Schmitt, 2005) and can thus be helpful to avoid the problem of demand characteristics.

Third, our sample and vignettes targeted 20- to 30-year-old German people. Academic and social goals may vary across the lifespan and cultures, moderate the importance of duty and socializing situations, and thus affect their weight in choosing interaction partners and evaluating others. Even what is socially desirable might depend on these factors. For example, an interesting approach could be the longitudinal perspective on romantic relationships with respect to self-control. At the beginning of a romantic relationship, partners might exercise much more self-control in order to convince their partner of sustaining the rather fragile relationship than after a few months or years.

Although we identified situation type, similarity, and self-control as three parallel effects, with the last one being the weakest in our study, we cannot generalize their significance to contexts outside the laboratory. Generalization to other cultures where different traits are considered socially desirable may not be possible either. Replications, especially in non-Western cultures, are needed. Due to the artificial setting and different manipulations and measures (i.e., manipulation of vignette self-control but no manipulation of respondent self-control), the effect sizes might be skewed.