E was a significant predictor even in a model that included systemizing vs. empathizing (which was omitted from the main analysis above because it seemed uniquely MLN9708 web relevant to the male vs. female contrast). doi:10.1371/journal.pone.0150194.tPLOS ONE | DOI:10.1371/journal.pone.0150194 March 3,10 /”Brilliant” “Genius” on RateMyProfessors Predict a Field’s Diversityof effort put in by their instructors; e.g., [44]). If so, the relationship between this languagebased measure and women’s PhD attainment would simply amount to predicting fewer women at the PhD level based on observing fewer women in college. However, this alternative cannot explain why the frequency of “brilliant” and “genius” also predicts the representation of African Americans at the PhD level; no empirically documented differences in valuing brilliance vs. effort distinguish African Americans from other groups. Thus, the most parsimonious explanation for this set of findings is that our word-count measure indeed taps into a field’s shared beliefs about success. When these beliefs emphasize the need for brilliance, members of groups stereotypically viewed as lacking such a quality are less likely to obtain PhDs. Consistent with this interpretation, prior studies found that adjusting for the gender composition of the respondents from each discipline did not affect the predictive relationship between disciplines’ ability beliefs and their PhD diversity [1, 5]. Although such an adjustment is not possible here (since the gender of the students filling out evaluations on RateMyProfessors.com is not recorded), there is no reason to suppose that it would have any more of an effect on these relationships. With our current data, however, we cannot completely rule out this alternative. To explore the divergent validity of SART.S23506 our language-based measure of field climate, we tested whether the brilliance language score was a significant predictor of Asian Americans’ PhD attainment. We expected it might not be: The career aspirations of groups who are not fpsyg.2017.00209 targeted by negative stereotypes about intelligence shouldn’t be strongly affected by a field’s emphasis on brilliance. The results suggested that, although the relationship between the brilliance language score and the representation of Asian Americans at the PhD level was in the same direction as those for women and African Americans, it was of smaller magnitude and not significant, r(16) = ?25 [?64, .24], p = .315. Brilliance language did not significantly predict Asian Americans’ Ph.D. representation beyond our controls either, = -.22 [-.64, .20], p = .275 (see Table 3). This null result, combined with the significant results for women’s and African Americans’ PhD representation, supports the claim that groups who are the targets of negative stereotypes about their intelligence are particularly likely to be underrepresented in fields that order P144 Peptide cherish brilliance and genius. The Gendered Language Tool allows word searches to be performed separately for positive vs. negative reviews (i.e., reviews that scored higher vs. lower than the midpoint of the “overall quality” rating on RateMyProfessors.com, respectively). In a separate set of analyses, we explored whether brilliance language scores computed separately over the positive and negative reviews predicted women’s and African Americans’ PhD representation. A priori, there is little reason to expect an asymmetry between these two language scores, since frequent use of “brilliant” and “geni.E was a significant predictor even in a model that included systemizing vs. empathizing (which was omitted from the main analysis above because it seemed uniquely relevant to the male vs. female contrast). doi:10.1371/journal.pone.0150194.tPLOS ONE | DOI:10.1371/journal.pone.0150194 March 3,10 /”Brilliant” “Genius” on RateMyProfessors Predict a Field’s Diversityof effort put in by their instructors; e.g., [44]). If so, the relationship between this languagebased measure and women’s PhD attainment would simply amount to predicting fewer women at the PhD level based on observing fewer women in college. However, this alternative cannot explain why the frequency of “brilliant” and “genius” also predicts the representation of African Americans at the PhD level; no empirically documented differences in valuing brilliance vs. effort distinguish African Americans from other groups. Thus, the most parsimonious explanation for this set of findings is that our word-count measure indeed taps into a field’s shared beliefs about success. When these beliefs emphasize the need for brilliance, members of groups stereotypically viewed as lacking such a quality are less likely to obtain PhDs. Consistent with this interpretation, prior studies found that adjusting for the gender composition of the respondents from each discipline did not affect the predictive relationship between disciplines’ ability beliefs and their PhD diversity [1, 5]. Although such an adjustment is not possible here (since the gender of the students filling out evaluations on RateMyProfessors.com is not recorded), there is no reason to suppose that it would have any more of an effect on these relationships. With our current data, however, we cannot completely rule out this alternative. To explore the divergent validity of SART.S23506 our language-based measure of field climate, we tested whether the brilliance language score was a significant predictor of Asian Americans’ PhD attainment. We expected it might not be: The career aspirations of groups who are not fpsyg.2017.00209 targeted by negative stereotypes about intelligence shouldn’t be strongly affected by a field’s emphasis on brilliance. The results suggested that, although the relationship between the brilliance language score and the representation of Asian Americans at the PhD level was in the same direction as those for women and African Americans, it was of smaller magnitude and not significant, r(16) = ?25 [?64, .24], p = .315. Brilliance language did not significantly predict Asian Americans’ Ph.D. representation beyond our controls either, = -.22 [-.64, .20], p = .275 (see Table 3). This null result, combined with the significant results for women’s and African Americans’ PhD representation, supports the claim that groups who are the targets of negative stereotypes about their intelligence are particularly likely to be underrepresented in fields that cherish brilliance and genius. The Gendered Language Tool allows word searches to be performed separately for positive vs. negative reviews (i.e., reviews that scored higher vs. lower than the midpoint of the “overall quality” rating on RateMyProfessors.com, respectively). In a separate set of analyses, we explored whether brilliance language scores computed separately over the positive and negative reviews predicted women’s and African Americans’ PhD representation. A priori, there is little reason to expect an asymmetry between these two language scores, since frequent use of “brilliant” and “geni.