Objective

This study examines whether single-item attitude measures commonly used in research adequately capture adolescent attitudes towards specific delinquent behaviors. It also explores whether delinquent attitudes form a unidimensional or multidimensional construct. Finally, this study tests Ajzen and Fishbein’s principle of compatibility to determine if delinquent attitudes have a behavior-specific and a situation-specific effect on behavior.

Methods

Data from a sample of 223 high school students in the Pacific Northwest (PNW) and from the Gang Resistance Education and Training (GREAT) evaluation are used. The means and correlations between single-item global attitude items are compared to situation-specific measures. Exploratory factor analysis is conducted to examine dimensionality, and logit regressions are used to predict the willingness to offend in the PNW and self-reported offending in the GREAT.

Results

The findings indicate that: (1) global attitude measures underestimate the degree of approval of delinquency and are not strongly correlated with situation-specific attitudes; (2) delinquent attitudes form a multidimensional construct that coalesce around crime-types; (3) there is some evidence that attitudes towards specific crimes have crime-specific effects on behavior, and; (4) situation-specific attitudes predict the willingness to offend in a specific situation above and beyond one’s overall approval of that behavior.

Conclusions

Attitudes towards fighting and theft are too complex to be captured by single-item measures. The findings also support the principle of compatibility which suggests that attitudes may be well-suited to explain why individual engage in specific types of crimes in specific situations. Other implications and limitations are discussed.