Uncomplicated descriptive statistics of Table .When compared with Table , at years postBSE, the addition of controls erased the gender distinction for the population as a entire (Neither table finds a gender differenceof BSE engineers are comparable for men and females.TABLE Average probability of remaining in engineering (working or studying) or out on the labor force all cohorts combined.of all BSE grads engaged in engineering of BSE grads working FT in engineering Out in the Labor Force Male Female # ObservationsMale Female Femalemale Male Female Femalemale Male Female Femalemale difference years postBSE years PubMed ID:http://www.ncbi.nlm.nih.gov/pubmed/21550118 postBSE years postBSE ………difference ………distinction ……… Gender distinction ttest p p .averages can’t be offered mainly because the #observations in some circumstances are too modest to report.TABLE Coefficient on female from linear probability models of remaining in engineering all cohorts combined.Probability of remaining in engineering Population all years postBSE . years postBSE.Probability of leaving the labor force Population all . . . .Population working FT . .. years postBSE. Years postBSE if still in Eng at years. .Coefficient significance p p p .Standard errors in parentheses.Controls consist of dummies for engineering subfield, survey year, BSE year, if parent had BABS, immigrant status, race.#obs All population years ,; years ,; years ,; years .#obs FT only years ,; years ,; years ,; years .Frontiers in Psychology www.frontiersin.orgAugust Volume ArticleKahn and GintherDo recent females engineers stayretention disadvantage for fulltime workers at this stage).At years, for the whole population, what was an .ppt.gender difference in Table becomes .ppt.with controls (Table); in contrast, among those operating complete time, there is no longer a considerable gender distinction.Lastly, with controls, gender differences in being out on the labor force (Table) are somewhat smaller sized than devoid of controls (Table) and no longer important at years.All round, then, the manage variables do explain several of the gender differences observed within the descriptive statistics.In function not shown, we investigated which with the controls variables have been the important mediating components.We found that subfield was 1 essential factor but that raceethnicity was one of the most significant control variable accountable for several of the average gender gap .Women in engineering are much less probably than males to be white (nonHispanics)the race together with the highest retention ratesand much more likely to become Asian or black, both groups with decrease retention prices.This outcome suggests that racial retention rates are critical to study in future investigation.The final row models retention at an even later career stages by asking, “Of those who remain functioning in engineering right after their Naringoside Biological Activity degree, what is the gender difference within the likelihoodof remaining in engineering around years later” This makes it possible for us to incorporate BSEs as early as , although the earliest BSEs we can observe at their careers’ beginning are from .This row indicates that there was no important gender retention distinction throughout years among these individuals who had been nonetheless in engineering at the starting of this stage.When we look only at those who are still fulltime employed at year postBSE, on average females are additional probably than males to stay in engineering.Variations across CohortsTables , present gender differences for cohorts defined by narrow ranges of BSE years.Table gives averages per cohortg.