Drinking, Alcohol Policy, and Attitudes Toward a Campus Riot

Our finding that drinking was most common among those who lived in Greek houses was consistent with other studies (Lo & Globetti, 1993; Wechsler, Lee, Kuo, & Lee, 2000). Despite this, the total effect of residence on riot attitudes was minimal for several reasons. First, although the correlations between residence and alcohol consumption were statistically significant, they were not very large (perhaps because residence is a weak measure of Greek membership, seeing that many Greeks live in off-campus apartments). second, the results reported arc based on a model that assumes that although residence and drinking are correlated, we do not know the causal direction between them. But even if we assume that the causal direction is from residence to drinking, the fact that riot attitudes are further removed from residence than is alcohol consumption would result in only weak effects of residence on riot attitudes.

Validity Issues: Relationship of these Attitudes to Actual Behavior

How well do the responses to our survey reflect the true attitudes of the respondents? As discussed previously, the conditions under which the survey was administered should have made respondents confident in the anonymity of their responses, thereby encouraging honest responses.

How closely are the survey responses related to actual behavior? Even though some (e.g., Wicker, 1969) have questioned whether attitudes predict overt behavior, Kim and Hunter (1993) found that when one corrects for methodological artifacts, the average correlation between attitudes and behavior is quite high (.79).

The indicators of Enjoying the Riot appear to be good proxies for actual participation. First, one of the indicators of Enjoying (being close enough to see it) involved memory of an overt behavior that took place not very long before the survey and was likely to have been quite salient. Thus, this behavior should be remembered quite accurately. second, anyone who was both close enough to see the riot and thought it fun, was probably a member of the crowd. In contrast, anyone not close enough to see it was clearly a nonparticipant.

The assumption that enjoying is a good proxy for participation leads to an estimate of participation that corresponds quite well to other estimates. Just over 26% of respondents said that they were both were close enough to see it and that it seemed like fun. This percentage suggests that about 8800 students, out of approximately 33,000 undergraduates, were in the crowd. Although this is slightly below the usual estimate of 10,000 participants, nonstudents can account for the remainder.

Moreover, the relationships we found between Enjoyment and various student characteristics are consistent with several findings regarding participation in this riot and others, (a) That males were most likely to enjoy the riot is consistent with the finding that over 80% of those arrested in this riot were male (June, 1999). (b) That younger students were most likely to enjoy the riot is consistent with data showing that 46 of the 64 MSU students arrested were in their first or second years, and with a similar finding about participants in a University of Northern Colorado riot (Janofsky, 2001). (c) That African Americans were less likely than Whites to condone or enjoy the riot is consistent with the fact that no African Americans were arraigned (June, 1999). (d) The absence of an effect of GPA on riot attitudes is consistent with the data showing that those arrested had similar GPA's to those of other MSU students (June, 1999).