Cover Image


View/Hide Left Panel

dents were then asked to provide data on the number of applicants and interviewees for each advertised position by gender. Finally, they were asked to identify the gender of the individual who was first offered the position and the gender of the person who was ultimately hired.3

In general, departments were much more knowledgeable about the later stages of the hiring process and thus provided more complete data on offers and hires than on interviews or applicants. The number of cases for which we had complete information on applicants, interviewees, first offers, and hires—all disaggregated by gender—varied between 534 cases (with complete hire information) and 758 cases (with complete applicant information). Thus, the number of cases considered in this chapter depends on the stage of the hiring process. Only tenured and tenure-track cases are considered in the analysis. For each stage in the hiring process (applications, interviews, offers), descriptive statistics based on the data collected from the departmental survey are first presented. Then, the appropriate statistical models are fit in order to understand the departmental characteristics associated with the percent of females at each stage of the hiring process.


A necessary precondition for hiring a female faculty member is to have women who are interested in applying for the position. The survey data clearly show that some departments are more successful than others in attracting female applicants.4 Moreover, our data show that there are still a number of positions for which no women apply.

Throughout this report, we will present summary statistics, such as the following ones, that state current values for men and women across the six disciplines surveyed. These statistics do not reflect the survey weights5 and are not treated for the different degrees of nonresponse that depended on the characteristic examined. Therefore, these statistics are NOT appropriate estimates of any national characteristics for men and women, but instead are quick impressions of the data collected, which are often the beginning of a more meaningful analysis that is conditional on the disciplinary area.6


A limitation of the survey was that it did not ask for the gender of every candidate offered a particular position.


Note that this analysis implies nothing about the quality of applicants. Some people apply for jobs for which they are not a very good fit. The committee did not assess whether male and female applicants would behave any differently in this regard.


Recall that the committee’s survey was stratified in order to collect similar numbers of respondents in each of the six disciplinary areas, and therefore respondents from different disciplines have different survey weights.


These estimates would be useful as national estimates only in situations in which the disciplines are relatively homogeneous with respect to a given characteristic and the nonresponse which occurred was such that nonrespondents did not differ in their characteristics from respondents.

The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement