Reader Comments

Post a new comment on this article

Recency of incidents

Posted by mrcc on 17 Jul 2014 at 17:04 GMT

You mention that some respondents were tenured or retired/emerita faculty, but that harassment most likely occurred during training years. It doesn't seem that you asked how recently the events occurred. Did you find a difference between reporting rates of tenured or retired women (who were trained 25 or 40 years ago) and more recent trainees?

No competing interests declared.

RE: Recency of incidents

katiehinde replied to mrcc on 24 Jul 2014 at 01:03 GMT


Thank you for your interest in our study. You are correct that we did not ask how recently the experiences occurred that the respondents provided in the survey. A majority of the respondents, however, were trainees at the time that the survey was conducted (N = 386/666, 58%). We considered conducting analyses that may have assessed whether there was evidence of changing patterns across time, but did not include them due to constraints on analyzing and interpreting such data. These constraints were the consideration of notable experiences, the possible confound between duration of time since training and duration of time spent in the field, and limited sophistication of model building due to underpopulated cells. These are described in greater detail below.

1. Notable Experiences and Time Depth: In the survey we asked about experiences of being the target for inappropriate comments and unwanted sexual contact. In this question, we indicated that if there had been multiple such experiences, to provide information about the most notable. If trainees are more targeted because of power differentials, then the most notable experiences may be more likely to occur during the trainee stages, while less notable experiences that occur at more recent time points may be less likely to be captured after the “notable” prompt. If experiences are, in part, a function of amount of time spent in the field- field work person years, if you will- then faculty respondents will have a “completed trainee” period of experiences to consider (as well as after that) that may reflect cultural shifts across time or the length of time they have been conducting field work.

2. Under-populated Cells for Statistical Analysis: Considering status and age at time of survey, status at time of experience, gender identity, and the position of the perpetrator relative to the respondent at the field site, under-populates some cells for more sophisticated statistical analyses.

Instead, in response to your query we divided the dataset into two age groups: <40 years of age (N=466) and 40+ years of age (N=199) among the respondents at the time the survey was conducted. This grouping resulted in 94% of “senior” faculty respondents to be included in the 40+ age group. We then re-ran the analyses for experiences of inappropriate comments and unwanted sexual contact as a function of gender and included age group as a co-variate.

The significant main effect for gender remained (as before p<0.0001). Age group was not associated with experiences of unwanted physical sexual contact (p>0.1). However, the older age group was more likely to experience inappropriate comments (chi-squared=49.37, df=2, N=650, model p<0.001, age group p=0.004, OR= 1.8). This may be in part because of a substantially higher percentage of inappropriate comments experienced by 40+ men (though note, an interaction between gender and age group was not significant p>0.1, and did not produce an appreciably better or worse AICc than gender and age group as the only predictors w/o an interaction).


Women Experiencing Inappropriate Comments
Age <40= 69% (259/378)
Age 40+= 76% (102/134)

Men Experiencing Inappropriate Comments
Age <40= 31% (25/81)
Age 40+= 54% (31/57)

We are hesitant to interpret these results, however, in light of the considerations outlined in #1 above.


No competing interests declared.