Of the 2,618 youth who participated in the survey, 2,096 had complete data on the variables of interest and reported being either White or Native American, and thus comprised the analytic sample. Of the 2,096 respondents, 17.2% (N = 361) were Native American and 82.8% (N = 1,735) were White. The mean age was 13.9 (SD = 2.04) for Native American youth and 13.0 (SD = 1.92) for White youth, t(2,094)=−8.16, p<.001. The sample consisted of slightly more females (52.1% for Native Americans, 52.8% for Whites) than males (47.9% for Native Americans, 47.2% for Whites).
Bivariate analyses of drinking behavior
Native American were significantly more likely to have consumed alcohol in their life than White youth (72.1% vs. 48.9%), χ2(1, n = 1,627) = 52.05, p < .001. Among lifetime drinkers, Native American and White youth did not differ in whether they consumed alcohol over the last 30 days (59.6% vs. 68.3%), χ2(1, n = 583) = 3.66, p < .07. However, Native American youth were more likely to have been intoxicated at least once in the past 30 days (37.3% vs. 27.0%), χ2(1, n = 864) = 8.11, p < .003. Native American and White youth did not differ in their mean age at first alcohol consumption, F(1, 760) = 1.97, p <.16, and had their first drink around 12 years of age (M=12.1, SD=2.2 for Native Americans, and M=11.8, SD=2.6 for Whites).
Multivariate analyses of drinking behavior
When controlling for gender and age, Native American youth remained more than twice as likely as White youth to have consumed alcohol in their lifetime (). Lifetime use was also higher for older youth. Gender, however, was not a significant predictor of lifetime drinking. Native Americans and Whites did not differ (β=−.02, p < .47) in 30 day alcohol use (). Age was a significant predictor of alcohol use in the last 30 days (β=.29, p <.001), as was gender (β= .07, p < .04), with males being more likely to have consumed alcohol in the last 30 days. Native Americans were more likely to have been intoxicated in the last 30 days even after controlling for age and gender (β= .40, p < .02). Age was a predictor of intoxication (β= .32, p < .001), but gender was not (β= .15, p < .33). Age at first alcohol use did not differ by race after controlling for gender and age (β = .04, p < .24). Gender and age, however, were significant with males beginning drinking at a younger age than females (β= −.12, p < .001) and older drinkers reporting beginning drinking at an older age (β=.58, p < .001). This latter finding, of course, is not substantively meaningful since age of initiation is confounded with current age.
Summary of Logistic Regressions Predicting Lifetime Drinking and Intoxication
Summary of Regression Analyses Predicting Drinking and Age of Initiation
Bivariate analyses of perceived ease of access to alcohol
The means for perceived ease of access to alcohol from each of the sources were initially compared using analysis of variance. These analyses showed that Native American and White youth differed significantly on use of all sources, except for parents and home (). Specifically, Native Americans reported easier access to alcohol in general, from a store, sibling, older person, bar and party.
Mean Perceived Ease of Obtaining Alcohola
Multivariate analyses of perceived ease of access to alcohol
Regressions controlling for age and gender indicated that Native American youth perceived it to be easier to get alcohol from stores and bars than White youth, but more difficult from parents and from home (). The ease of getting alcohol at a party, from an older person or siblings, and in general did not differ for Native Americans and Whites once age and gender were controlled. Males found it more difficult than females to obtain alcohol at parties, from older persons, and from family-related sources. Gender was not a factor in the perceived ease of obtaining alcohol from a store, bar, or in general. Not surprisingly, age was inversely related to perceived difficulty of obtaining alcohol from all sources.
Summary of Regression Analyses Predicting Perceived Ease of Obtaining Alcohol
Bivariate analyses of sources for alcohol
Native American youth who drank were less likely than comparable White youth to have obtained alcohol from their own parents, but more likely to have obtained it from someone under 21 or from another adult over 21 (). Native American and White youth did not differ in their access to alcohol through friend’s parents, own home, or friend’s home. The number of youth who obtained alcohol directly from commercial sources (bar, restaurant, grocery or liquor store) was very small (3.0% Native American, 1.5% White), and did not permit additional analyses of differences between the two groups in regards to these sources of alcohol.
Sources of Alcohol at Last Drinking Occasion
Multivariate analyses of sources for alcohol
Logistic regressions controlling for age and gender indicated that Native American youth were more likely to have obtained alcohol from adults over 21 or someone under 21 (). Specifically, Native American drinkers were almost twice as likely to have gotten alcohol from an adult and more than twice as likely as White drinkers to have obtained alcohol from someone under 21. However, White youth were four times as likely as Native American youth to have obtained alcohol from their parents. Youth did not differ in their likelihood of obtaining alcohol from a friend’s parents, their own home or a friend’s home. Males and females’ access to alcohol differed on none of the sources. Age was a significant factor for most sources. Older drinkers were more likely to report they obtained alcohol from friend’s parents, someone over 21, or someone under 21. Younger drinkers were more likely to report getting alcohol from their own parents. The primary drinking context for younger youth, however, may be within the family or family celebrations and involve only small amounts of alcohol.
Summary of Logistic Regressions Predicting Sources of Alcohol