Sexual Assault at Indiana Colleges and Universities

1
Sexual Assault at Indiana Colleges and
Universities
Edgardo R. Pimentel, M.S. INCSAPP/ICAN Workshop
September 7, 2006
2
The Core Institute

• Assess the nature, scope, and consequences of
  alcohol and other drug use on college campuses.
• To date, the CORE Survey has been administered to
  1,675,391 students at about 1,567 American
  universities and colleges.

3
Purpose

• To find potential predictors of Sexual Assault
  and Rape
• Examine the relationship of those predictors to
  Sexual Assault and Rape

4
Background

• Negative consequences are what we really try to
  reduce
• Sexual assault is qualitatively different from
  other types of negative consequences
• Requires separate analysis

5
Outline

• Items
• Indiana Sexual Assault
• Analysis
• Model
• Results of Assault
• Results of Rape
• Exploration

• Model
• Results of Assault
• Results of Rape
• Relationships of the Predictors
• Discussion

6
Items

• Questions 25e and 25f
• - Forced sexual touching or fondling
• - Unwanted sexual intercourse

7
Indiana Sexual Assault
8
Indiana Sexual Assault
9
Analysis

• Logistic Regression on two conditions
• - No Assault vs. Assault
• - Touching vs. Rape

10
Analysis

• Estimates the odds of being classified as a
  victim of Assault or of Rape
• Provides an R square equivalent value

11
Model

• Classification
• Age
• Ethnic origin
• Gender
• Marital status
• Working
• Living arrangements
• GPA
• Heavy drinking
• Average drinks per week

12
Model

• All variables had significant differences on the
  dependent variables of Assault and Rape except
  for question 7 Working

13
Results of Assault

• Overall model is significant (Wald 3125.17, 1
  df, p.lt.05)
• Predictions for Step 6 of 7 are not significantly
  different from observations (Chi-square 14.45,
  8 df, p. .07)

14
Results of Assault

• Significant predictors were being single (OR
  .21), married (OR .06), divorced (OR .06),
  GPA (OR .8), Heavy Drinking (OR 1.3) and
  Average Drinks per Week (OR 1.0)
• R Square .10

15
Results of Rape

• Overall model is not significant (Wald 1.57, 1
  df, p. .21
• Predictions for Step 2 of 2 are not significantly
  different from observations (Chi-square 4.23, 8
  df, p. .84)

16
Results of Rape

• Significant predictor was Average Drinks per Week
  (OR 1.0)
• R Square .08

17
Exploration

• Ran correlations on all items on the survey
  against Assault and Rape
• Highest correlations belonged to AOD Use, Use at
  residence halls and Greek houses, other negative
  consequences
• Lowest correlations belonged to never using AOD
  at any locations

18
Model

• Heavy drinking
• Average drinks per week
• Annual AOD rates
• 30-day AOD rates
• Locations of AOD use
• Change in drug use
• Alcohol last time they had sex
• Bragged about AOD use

19
Results of Assault

• Overall model is significant (Wald 1843.39, 1
  df, p.lt.05)
• Predictions for Step 6 of 6 are not significantly
  different from observations (Chi-square 3.14, 4
  df, p. .53)

20
Results of Assault

• Significant predictors were annual cocaine use
  (OR 1.3), alcohol use at fraternities (OR
  2.1), amphetamine use at residence halls (OR
  2.6), never using other drugs at any location
  (OR 0.6), alcohol prior to sex (OR 1.8) and
  bragging about AOD use (OR 1.2)
• R Square .10

21
Results of Rape

• Overall model is not significant (Wald 3.60, 1
  df, p. .06
• Predictions for Step 3 of 3 are not significantly
  different from observations (Chi-square 4.14, 4
  df, p. .39)

22
Results of Rape

• Significant predictors were 30 day use of
  sedatives (OR 2.9) and bragging about AOD use
  (OR 1.3)
• R Square .13

23
Relationships of the Predictors

• GLM on Assault and Rape for the predictors found
  through the Logistic Regressions
• Tested for between-subjects differences and
  interactions for gender

24
Relationships of the Predictors

• All between subjects test were significant
• All but two gender differences were significant
  (Marital Status, Alcohol use at Greek house)

25
Relationships of the Predictors

• Five gender interactions were significant
• - Average Drinks per Week
• - Cocaine Use Last Year
• - Amphetamine Use in Residence Halls
• - Sedative Use Past 30 Days

26
(No Transcript)
27
(No Transcript)
28
(No Transcript)
29
(No Transcript)
30
(No Transcript)
31
(No Transcript)
32
(No Transcript)
33
(No Transcript)
34
(No Transcript)
35
(No Transcript)
36
(No Transcript)
37
Discussion

• Warning signs exist and should be used to monitor
  or educate potential victims
• Victims of assault can serve as the next best
  source of information for increasing our ability
  to predict

38
Discussion

• There are several interesting and telling
  relationships
• We did not find a good way to predict Sexual
  Assault
• Need to look at some other facets of their lives
  that may prove informative

39
Discussion

• Items such as Sedative use may be indicators of
  other issues in their lives or a result of being
  victimized
• Bragging about AOD use may be related to the peer
  relationships maintained

40
Discussion

• These items can serve as launching points to look
  at the other facets of the individuals
  environment

41
Closing

• We are a resources available to you.
• Questions are free.