It is often assumed that men spend a greater percentage of their time doing research than women. The percentage of time spent on research or scholarship was combined with percentage of time spent seeking funding in our survey data. Overall, men reported spending a slightly greater percentage of their time on research activities than women: 42.1 compared to 40.0 percent. This difference, while approaching significance, is quite small in absolute terms. Drawing on similar faculty from the NSOPF:04, there was no significant difference between men and women in the time spent on research activities: 43.2 compared to 39.7 percent.[59]
|
Engineering Engineering |
□ Men □ Women
FIGURE 4-1 Mean percentage of time faculty spent on research (self-reported) by gender. SOURCE: Faculty survey carried outbytheCommittee on Gender Differences in the Careers of Science, Engineering, and Mathematics Faculty.
It is worth noting that the overall percentage of time faculty report spending on research activities is remarkably similar in the two surveys.[60]
Figure 4-1 shows the reported percentage time spent by faculty in research activities (including preparation of grant and contract proposals) disaggregated by gender and by discipline. Averages were computed over faculty who provided this information on the survey. To investigate whether there are differences in the percentage of time spent in research across disciplines or across genders, we fitted a simple linear model with percent time as the response variable and with discipline, gender, and the interaction between discipline and gender as the effects. We found no significant differences in percentage time spent in research, either across disciplines or between genders within discipline. Because comparing genders within discipline involved carrying out six comparisons, we used the Tukey-Kramer approach[61] to adjust the individual p-values. The smallest of the six p-values was obtained when comparing men and women faculty in chemistry (p- value = 0.217). All other p-values were above 0.35. Please note that discipline and gender accounted for a very small (about 2 percent) proportion of the variability observed in self-reported time spent in research activities. Thus, these p-values are to be interpreted cautiously. A model in which other potential confounders are also included is presented later in this chapter.
In the NSOPF:04 data, there were no significant gender differences in any of the aggregated disciplinary groups reported (biology, physical sciences, mathematics, and computer science).