| Abstract
|
:
|
Problem. The problem of the study was to analyze the National Adolescent School Health Survey (NASHS) data to determine relationships among knowledge, attitude (using a three component model), and behavior related to suicide and assaultive violence. The roles of the quantity of health education instruction and specific categorical health instruction as related to violent behaviors were also studied using selected phases of Green's PRECEDE Model (Green & Kreuter, 1991). Procedures. The NASHS used a national, multi-stage probability sample of eighth and tenth grade students. Two different groups of students provided data for analysis: responses from 3998 students were analyzed to examine correlates of suicide behavior, while responses from 3789 students were analyzed to examine correlates of assaultive violence behavior. Further evidence of content validity for the NASHS items related to knowledge, attitudes (beliefs, feelings, intention to act), and behaviors on violence, was provided through a panel of experts. Reliability measures were determined using Cronbach's alpha, split-half, and Kuder-Richardson methods. For the purposes of testing the related hypotheses, the data were analyzed by utilizing Pearson product-moment correlation technique and stepwise multiple regression method. Findings. The results of the study demonstrate acceptable evidence of reliability and validity for the set of instruments. There were statistically significant relationships for suicide behavior with knowledge, belief, feeling, and intention to act, along with significant relationships for assaultive violence behavior with knowledge and belief (p <.05). Furthermore, correlation coefficients were higher between categorical health instruction on violence and the antecedents under study, than among general health education instruction classes. The utilization of selected components of the PRECEDE model accounted for 13% of the variance in suicide behavior and 25% of the variance in assaultive violence behavior. Conclusions. These results suggest that knowledge, belief, feeling, and intent to act variables may help in the further understanding of violent behavior in adolescents. Consequently, instruction concerned with preventing adolescent high-risk suicide and assaultive violence behavior should take into account the variables knowledge, belief, feeling, and intention to act, when developing and implementing health education programs. In addition, despite low coefficient of determination (R2) values, the application of the PRECEDE model may be useful in explaining adolescent high-risk violent behaviors and therefore assist in the systematic selection of specific intervention strategies.
|