In this study we followed the extension of Tversky’s research about features of similarity with its application to open sets. Unlike the original closed-set model in which a feature was shifted between a common and a distinctive set, we investigated how addition of new features and deletion of existing features affected similarity judgments. The model was tested empirically in a political context and we analyzed how positive and negative changes in a candidate’s profile affect the similarity of the politician to his or her ideal and opposite counterpart. The results showed a positive–negative asymmetry in comparison judgments where enhancing negative features (distinctive for an ideal political candidate) had a greater effect on judgments than operations on positive (common) features. However, the effect was not observed for comparisons to a bad politician. Further analyses showed that in the case of a negative reference point, the relationship between similarity judgments and voting intention was mediated by the affective evaluation of the candidate.

The current work shows empirical verification of the similarity theory extended to open sets of features, which states that an increase in similarity between 2 objects results from the deletion of distinctive features or the addition of common features, albeit with different effects. To date, theoretical simulations and empirical demonstrations using real features have shown how the model works. However, to test the validity of the model we conducted an experiment where direct judgments were used. A total of 188 participants were evaluated at 3 different levels related to similarity, affect, and willingness to live, in reference to cities that differed in terms of their amounts of positive and negative features. The results revealed that when their initial similarity to the ideal was greater than 0.5, deletion of distinctive, negative features had a stronger effect on the object evaluation and was more effective than adding positive features of the same value. When the initial similarity to the ideal was less than 0.5, opposite results were obtained; that is, the addition of common, positive features had a stronger effect on the object evaluation than deleting negative features of the same value. The results have shown a positive-negative asymmetry in the evaluation process, which supports the similarity theory extended to open sets of features. In addition, the findings revealed that the positive-negative asymmetry derived from the theory can be explained by the ratio difference principle.

No abstract

A stereotype threat arises when a negative stereotype about group to which an individual belongs is activated. It affects the achievement and interest of students in a particular academic domain, e.g., girls at math or boys at language arts. Hence, it is important to assess the level of stereotype threat at school (STaS) in order to identify the vulnerability of students to its negative consequences. This study devised and validated two parallel versions of the STaS scale: girls in mathematics and boys in language arts in a nationally representative sample of Polish secondary school students (N = 1,241; 13–16 years). The results of a confirmatory factor analysis (CFA) in a complex sample approach showed one general factor. Furthermore, a multiple-group CFA confirmed metric invariance and partial scalar invariance. The variances for boys and girls were equal. This suggests that the construct of stereotype threat is similarly conceptualized by both genders despite being in different domains. Finally, the comparison of means of latent variables revealed a higher level of stereotype threat among boys in the language domain than girls in mathematics. Possible theoretical and practical implications are discussed.