Development of an adolescent risk-taking instrument
The purpose of the research study was to examine the psychometric properties of the Risk-Taking Instrument: the internal consistency and construct validity of the instrument. The problem for the study was to identify the adolescent's perception of risk-taking and the variables which influence adolescent risk-taking.
The conceptual framework for the research study was based on psychometric and developmental theory. The work on adolescent identity formation, cognitive development, and social cognition helped identify the concepts which formed the conceptual framework of adolescent risk-taking. The interaction of the variables identified in the conceptual framework guided the development of the subscales in the Risk-Taking Instrument (RTI).
A methodological design, and descriptive-correlational methods of research were used to standardize data collection, and to estimate reliability and validity of the RTI. Content validity and initial estimates of internal consistency were obtained in a pilot study.
The study sample consisted of 270 adolescents between the ages of 16-22 years who were enrolled in a vocational training program. Participation in the study was voluntary.
Cronbach's alpha correlation coefficient was used to estimate reliability, and factor analysis procedures were used to estimate factorial validity. An alpha correlation coefficient of 0.3681 was obtained on the RTI. The subscales ranged from a high alpha correlation coefficient of 0.7237 (thrill-seeking) to a low of 0.4951 (social adaptation). Because of the low alpha correlation coefficient of the RTI, and the number of negative item to total correlations, the subscales were grouped at each end of the risk-taking continuum to estimate internal consistency. An alpha correlation coefficient was maximized at 0.7232 for the grouped subscales on self-esteem, autonomy, and social adaptation and 0.7609 for the grouped subscales on impulsivity, vulnerability, and thrill-seeking.
Factor analysis procedures produced eight factors with factor loadings of 0.4000 or greater. The subscale on social adaptation produced two factors: sense of helplessness and sense of self-determination. The subscale on autonomy also produced two factors: social conformity and independence. All subscales loaded three items or greater, and the factor loadings for the items were correlated.
Statistical analysis of the RTI produced two distinct scales: adolescent social adaptation and adolescent risk-taking. Both scales reached acceptable alpha correlation coefficients, and were supported for initial estimates of validity by factor analysis.