TABLE 5.
The percentages of students and gender differences in who selected each factor that would make them feel uncomfortable answering questions in front of the whole class to instructors of large-enrollment college science coursesa
| What makes students feel uncomfortable answering questions | % (n) | Odds ratio of gender difference | P value for gender in respective model |
|---|---|---|---|
| If I do not feel confident about my answer | 91.4 (243) | Women are 3.3× more likely to select this factor* | 0.027 |
| If the material is not clear to me | 82.3 (219) | Women are 1.9× more likely to select this factor | 0.130 |
| If I am afraid of speaking in front of the whole class | 68.0 (181) | Women are 1.8× more likely to select this factor | 0.079 |
| If I feel that other students will judge me | 63.2 (168) | Women are 1.8× more likely to select this factor | 0.082 |
| If I feel like there are too many people in the classroom | 54.5 (145) | Women are 1.2× more likely to select this factor | 0.607 |
| If the instructor responds negatively when other students answer questions | 48.1 (128) | Men are 1.0× more likely to select this factor | 0.926 |
| If I am unable to discuss my answer with other students before answering in front of the class | 40.2 (107) | Men are 1.2× more likely to select this factor | 0.575 |
| If other students are not answering questions | 37.2 (99) | Men are 1.0× more likely to select this factor | 0.899 |
| If it is hard to answer a question from where I sit in class | 33.8 (90) | Men are 1.1× more likely to select this factor | 0.835 |
| If it is obvious to others in class that I am the one answering the question | 27.1 (72) | Women are 1.5× more likely to select this factor | 0.314 |
| None of these apply to me | 1.5 (4) | NA | NA |
For this analysis, we only included the total n = 266 students who reported that they were uncomfortable answering questions. For each factor, we used binomial logistic regression to test whether a woman or man is more likely select it, controlling for race/ethnicity, college generation status, year in school, and GPA. The results of each regression can be found in Text S1. In this table, we include the odds ratios that a woman or man is more likely to select a particular factor, and we indicate with an asterisk which odds ratio is significant at the significance level of 0.05. We also include the P values for gender in the regression model in the last column. NA, we did not test whether there were differences with regard to who selected “none of these apply to me,” since so few students selected this category.