It should be noted that the DOJ (NCVS) study also reported the summary results of two other studies (NISVS and CSA) that yielded much higher incidents of sexual assault. The DOJ study focused on assaults of a criminal nature and had a heavy interviewer bias (observer effect) while the other studies focused on public health aspects of sexual assault. Another way of viewing the results is that one survey may count only those cases that would result in prosecution while the other surveys suffer from representing only the victim's of sexual assault. The DOJ statistics were part of a large overall study of nonfatal crimes reported and not reported to police against persons age 12 or older; this study required the participant to determine if the activity was criminal. The other studies asked questions about sexual behavior and then the study made the determination about possible criminal behavior. The methodology and objectives differ substantially so a difference in results would not be unexpected.
Comparing the various study data from the NCVS report, requires it to be normalized to reflect the same time period, i.e. the period of college enrollment. For comparative purposes I will assume an average of five years of total college enrollment as working students or mothers of young children frequently take longer than four years to complete their degrees. This also reflects that these assault statistics are solely based on assaults on women ages 12 to 25. The studies were conducted between 200713 for the NCVS report, 2007 for the CSA findings, 2011 NISVS for the NISVS data. The CSA 2007 study reported
140 assaults per 1000 during college enrollment; the 2011 NISVS study reported a 2% annual assault rate during the previous 12 months, or
100 assaults per 1000 during a five year college enrollment; and in 2010 NCVS reported an annual assault rate of 1% (which was more than double the rate of the entire 7 year study period, an unexplained discrepancy), which would be
50 assaults per 1000. Note all the studies appear to support assault rates during college much higher than 5-9 per 1000 (140,100, 50) and closer to the order of magnitude of 200 per 1000 (20%).
Also important is the population surveyed; only the CSA study limited the sample population to college students. The other two studies had population of non-institutionalized persons; including far more than just college students but also excluding the homeless, the mentally ill, and the incarcerated groups which may include parts of the student population. In the recent Iowa caucuses Trump was predicted the winner among likely Iowa GOP voters while Cruz was the predicted
winner among likely GOP caucus attendees. Obviously predictions based on the appropriate sample population can be more accurate. The DOJ study used a very elaborate interview process and a large sample population, but did not reveal how they selected participants (the sample population). From their report it is hard to determine how representative of the general population their sample is or how representative of the student sub-set. The DOJ study involved interviews and telephone followups with households over 42 months. Households have to be stable, that is non-transient, over that 42 months to be included in the sample population. Students who moved at least once during their college career would be excluded from the sample population. The requirement that the data (NCVS) is gathered over a 42 month period would seem to exclude many if not most college students. I am not surprised that the DOJ study might under report the incidence of sexual assaults on campus as their methodology could exclude many if not most college students from their sample population.
GB54;842649729 said:
Our government has studied this.
The rate of rape and assault against non students was 1.2 times greater than for students.
The rate has been between 5-9 per 1000 females in the last 15 years, no way 20%
http://www.bjs.gov/content/pub/pdf/rsavcaf9513.pdf