We have already considered laboratory space as an institutional resource, because it is so crucial to the ability of faculty to get their research done. In our survey, male faculty reported having significantly more lab space than female faculty. This holds true both when we consider all faculty taken together and when we consider only those faculty who do experimental research. Here we look at what factors might contribute to the amount of lab space that a faculty member has and to the gender disparity in lab space.
Figure 4-10 shows the distribution of lab space in the entire sample (except mathematics) in the log scale. When space was reported as “0,” a negligible amount (10 square feet) was added to allow for the transformation. Note that after the log transformation, lab space has a distribution that is approximately symmetric. Thus, a linear model was fitted to the log of lab space.
Explanatory variables in the model were gender, discipline, faculty rank, type of institution (public or private), prestige of the institution, grant funding, publi-
|
|
 |
|
 |
 |
loglab MIDPOINT
(Natural) Logarithm of Lab Space in Square Feet
FIGURE 4-10 Histogram of the logarithm of lab space.
cations (refereed journal articles and conference proceedings), type of research (experimental, theoretical, both, educational, other), academic age (defined as time elapsed between receipt of Ph. D. and December 2004), and all two-way interactions with gender. Gender, discipline, rank, institution type, and prestige were classification variables; other variables were included as continuous. A random effect for institution was included, but the institution variance component was negligibly small. Because none of the interactions between gender and the other covariates were significantly different from zero, the model was re-fitted, but included only main effects. The model fitted the data reasonably well (R2 = .32).
Significant associations with lab space were found for discipline (p < 0.0001), rank (p < 0.0001), type of institution (p < 0.05), prestige (p < 0.01), grant funding (p < 0.0001), research type (p < 0.0001), and publications (p = 0.012). Importantly, gender was not associated with lab space. Since, overall, when other variables were not taken into account, male faculty had significantly more lab space than female faculty, the absence of a significant gender difference in this analysis suggests that the overall difference is a function of gender differences in one or more of the other variables in the analysis. The most likely candidates were discipline and rank. We know that the percentage of female faculty varies between disciplines and between ranks. Therefore, if the disciplines and ranks with more male faculty were also the disciplines and ranks with more lab space, a simple comparison of the lab space of male and female faculty would show an overall advantage for men.
There are several interesting effects of variables on lab space. First, there was a positive association between grants and lab space. Everything else being equal, a faculty member who doubles his or her funding in a year can expect an 11 percent increase in lab space. Therefore, the effect of increased funding on space depends on the level of funding. A faculty member who has $10,000 in sponsored funding would only need to raise about $20,000 to increase his or her space by 11 percent. Yet someone who already has $ 100,000 would need to reach $200,000 in funding to have the same effect on his or her space. Second, not surprisingly, experimental researchers (most faculty call themselves experimental, and therefore research type was dichotomized to experimental or nonexperimental—and most women declared themselves to be experimental) reported having more lab space. Third, faculty at public institutions received more lab space than faculty at private institutions. Fourth, faculty at the most prestigious institutions reported having more lab space than faculty at institutions of medium prestige, who in turn report having more lab space than faculty at the least prestigious institutions. Fifth, our study indicated that the more senior faculty (those who have moved up in rank) had more lab space. This, however, is not a conclusion well supported by our data, which by their cross-sectional nature do not permit drawing longitudinal inferences. A snapshot impression can be misleading if, for example, senior faculty with large labs at a given point in time also had large labs when they were junior faculty. The effect of increase in rank and the effect of time itself are confounded when we can only explore faculty with a range of rank but during a single period of time. Finally, every additional publication, all else being equal, was associated with an increase in lab space of about 1 percent.