Scheduled for Health Posters, Friday, April 2, 2004, 10:00 AM - 11:00 AM, Convention Center: Exhibit Hall Poster Session


Substance Use: Correlates of Intent to Use Among Elementary School Children

Cliff Knickerbocker, Jeanne Bleeker and Duston Morris, University of Arkansas-Fayetteville, Fayetteville, AR

Many school based drug education programs attempt to prevent substance use. Effective intervention programs should be targeted those variables that have been linked to substance use within the target population. Relatively few young children are actually involved in substance use, but their stated intention to use may be a good indicator of later use. The purpose of this study was to identify correlates of intent to use selected substances among students in grades 4-6. Subjects for the study were 717 students in grades 4-6 from 6 suburban elementary schools in a state located on the East Coast. Students voluntarily, and with parental consent, completed a 32 item questionnaire, in their regular classroom setting. Data were analyzed by organizing questionnaire items into seven factors, and subjecting each factor to confirmatory factor analysis. These factors were: (1) intent to use, (2) would use if friends were using, (3) could say no, (4) parents think it is ok, (5) do you think you should use, (6) most kids use, (7) kids will hurt themselves. For each factor items dealt with alcohol, tobacco, and marijuana. Results confirmed the existence of these factors, with each item, for each factor, loading at .735 or above. A multiple regression analysis was performed to identify a set of variables that accounted for variation in intent to use scores. The factor intent to use served as the dependent variable, while the other six factors were used as a set of predictor variables. Results indicated that the predictor variables did account for a significant amount of the variation in intent to use (<.001). The best two variable predictor model consisted of (1) Would use if with friends who were using and (2) Do you think you should use. The two variables together accounted for 61 percent of the variation in intent to use. Adding the other four variables to the model only increased the amount of variation for which the predictor accounted to 62 percent. These results indicate the two variables accounted for a substantial amount of the variation in intent to use. These findings should be considered as they attempt to develop effective interventions for this population.


Keyword(s): elementary education, health promotion, research

Back to the 2004 AAHPERD National Convention and Exposition