On 05/04/2016 09:24 AM, Neotarf wrote: > although you would think the Gruniad would not report on something > that was obviously flawed.
It's hard to know Gruniad's intention, given the research articles are also relatively old (not just published or forthcoming): 1. Bingham & Scherer (2001) 2. Tinkler, Li, & Mollborn (2007) Also TL&M (2007) don't cite B&S (2001)... > It goes without saying that a meaningful study should have a random > selection process, although it happens all the time that researchers > can't always get ideal populations so they study the populations > they have. Unfortunately the study is behind a paywall, so you can't > see how it was designed, Here's the appropriate bit for B&S (2001). ! We randomly assigned departments to the experimental (program participation) ! and control (program nonparticipation) conditions. Staff and faculty ! departments were chosen from separate lists, with staff departments organized ! by type of labor (administrative vs. nonadministrative) and faculty ! departments organized by college; this stratification procedure assured that ! different categories of staff and faculty departments in each of the colleges ! would be represented in the two conditions. Random selection of departments ! rather than individuals resulted in a quasi-experimental design in which each ! employee and combination of employees did not have an equal chance of being ! assigned to one of the two conditions of the study. Because our unit of ! analysis was the individual employee, we explored possible sources of bias due ! to differences between employees who were assigned to the experimental and ! control conditions. Specifically, we compared the questionnaires returned by ! experimental and control groups on both demographic and experiential ! variables. Pearson chi-square analyses (p ≤ .05) were performed to test the ! associations between participation condition and each variable. The ! composition of the two respondent groups was not significantly different in ! terms of gender, race/ethnicity, position at the university, prior experience ! as a perpetrator of sexual harassment, experience being accused of sexual ! harassment, or experience as a sexual harassment victim. The phi coefficients ! for this set of variables ranged from .03 to .06. These results suggest that ! employees in the two groups were comparable in important respects prior to the ! intervention. So it doesn't look like it's suffering from selection bias (wherein the abusers are sent to the training and have worse attitudes from the start). > All behind paywalls. If someone did want to write a WP article, I'm willing to help on this point. _______________________________________________ Gendergap mailing list [email protected] To manage your subscription preferences, including unsubscribing, please visit: https://lists.wikimedia.org/mailman/listinfo/gendergap
