## Info

Note. 1 = participated; 0 = did not participate. The 15 patterns shown are those involving the largest numbers of participants. Sixteen additional patterns, each containing fewer than 20 students, are not shown.

Note. 1 = participated; 0 = did not participate. The 15 patterns shown are those involving the largest numbers of participants. Sixteen additional patterns, each containing fewer than 20 students, are not shown.

presentation of the main item sets was rotated across the three forms so that slow readers would leave different questions blank depending upon which form they received.

### Measures

Because the focus of this chapter is the methods more than the substantive analyses, we describe the measures only briefly. The main analysis to be described was multiple regression.

### Dependent Variable

For the main substantive analyses, we treated heavy alcohol use at the last wave of measurement (November 2000) as the dependent variable. This measure was made up of three questionnaire items. One question asked for the number of times in the previous two weeks the person had consumed five or more drinks in a row. A second asked how many times in the previous two weeks the person had consumed four or more drinks in a row. The third question asked how many times in the previous 30 days the person had consumed enough alcohol to get drunk. For the regression analyses, these items were summed to form a composite scale for heavy alcohol use.

### Predictors

There were six predictor variables in the multiple-regression models. Three of these were considered background variables: gender, religiosity, and the personality variable, introversion. Gender was coded 1 for women and 0 for men. The gender variable was the average of the gender question over the five waves of measurement. For 1,021 students, this average was exactly 1 or exactly 0, implying complete consistency. For two students, the average was 0.8, meaning that these students were most likely women but responded male on one of the five questionnaires. For this chapter, these were assumed to be women. For only one student was this variable missing altogether.

The religion question was a single item asking how important it was for the student to participate in religion at college. Because the answer to this question was so highly intercorrelated from wave to wave, the religion variable used in this chapter was a simple average of this variable over the five waves of measurement.

The introversion variable was a composite of 10 items from a 50-item version of the Big-5 (Saucier, 1994). The items were scored such that higher values implied greater introversion.

Also included as predictors were three other variables from the September 1999 survey: negative attitudes about authority (DefyAuthority), perceptions of alcohol consumption by students in general at this university (PeerUse), and intentions to intervene in possible harm situations (Intent-Intervene). The number of items and coefficient alpha for each scale are presented in Table 4.2.

The patterns of missing data for the variables included in the main regression analyses (described later) are shown in Table 4.3. Two of the predictor variables (gender and religiosity), which were formed by combining data from all five waves of measurement, had so few missing values (12 cases were missing for religiosity and 1 case was missing for gender) that they were omitted from Table 4.3. A third predictor variable (introversion) was measured at only one time (at whichever time was the first measurement for any given individual) but had slightly more missing data and appears in Table 4.3.

TABLE 4.2 Summary of Questionnaire Items and Coefficient Alpha

Coefficient Alpha Impute Analyze

TABLE 4.2 Summary of Questionnaire Items and Coefficient Alpha

Coefficient Alpha Impute Analyze

## Booze BasherGet All The Support And Guidance You Need To Permanently STOP The Battle With Alcohol Once And For All. This Book Is One Of The Most Valuable Resources In The World When It Comes To Transformational Tools For Battling Booze Binges And Staying Alcohol-Free. |

## Post a comment