The response variable of interest was binary: a woman was first offered the position or the position was offered to a man. We considered all the institutional and position-level variables described earlier, with the following modifications. Instead of the percentage of female applicants, we now included the percentage of women in the interview pool, and instead of an indicator of whether the candidate pool is reviewed by a dean or an external committee, we included an indicator of whether a dean approves the hiring recommendation made by the committee. Since the probability that a woman will be offered the position when none was interviewed is clearly zero, we restricted these analyses to those positions for which interview pools included at least one woman. Similarly, we also deleted from these analyses those positions for which all interviewees were women. Thus, results presented here are conditional on having at least one woman and at least one man in the interview pool.
The only two factors that appear to be associated with the probability that a woman will be offered the position first are the percentage of women in the interview pool (p < 0.001) and whether the dean approved an offer (weak association with p = 0.06). When the dean reviews offers, the probability that a woman will be offered a position is 0.38, with a confidence interval of 0.26 to 0.50. This value is significantly larger than the 0.06 (95 percent confidence interval of 0.00 to 0.51) obtained in cases in which the dean has no role in reviewing offers. (The uncertainty around this latter value is high because of a very small sample size.
|
Distribution of Percentage of Females in Interview Pool
|
TABLE 3-6 Distribution of Percentage of Interviews with and Offers, to Women by Discipline
|
NOTES: Numbers in parentheses correspond to the total number of positions in each group defined by discipline, type, and percentage of women in the interview pool. The next-to-last column shows the total number of positions by discipline and type offered to a woman. The last column shows the total number of positions by discipline and type for which we have complete information on the gender of the candidate receiving the first offer.
SOURCE: Survey of departments carried out by the Committee on Gender Differences in Careers of Science, Engineering, and Mathematics Faculty.
In almost all cases, deans play a role at the time of offering a tenure-track or tenured position to an applicant.) The size of the “dean effect” must therefore be interpreted cautiously. For every 1 percent increase in the percentage of females in the interview pool, the probability that a woman would be offered the position increased by about 5 percent. Finally, the probability that a woman would be offered the position was lowest at the top 20 research-intensive institutions compared with non-top 20 research-intensive institutions surveyed. At the highest prestige institutions (top 10), the probability that a woman would get an offer approached significance (p = 0.08). No other factors were associated with the probability that a woman would get an offer.