Adolescent psychosocial maturity and alcohol use: Quantitative and qualitative analysis of longitudinal data - Statistical Data Included

Adolescence, Spring, 2002 by Sigrun Adalbjarnardottir

Measures

Substance use. To allow comparisons of adolescent substance use between populations on either side of the Atlantic, a questionnaire regarding substance use has been developed (Hibell et al., 1997) and was used in this study. To assess drinking status, adolescents were asked how many drinks they normally consumed at one sitting. The seven response categories were: none, less than 1, 1, 2, 3-4, 5-6, and 7 or more. The focus was on the number of drinks at each sitting as opposed to frequency of drinking since the number seems to be more salient among Icelandic teenagers (Adalbjarnardottir, Davidsdottir, & Runarsdottir, 1997). A cross-national study of 24 European countries found that only Norway and Turkey have lower rates than Iceland for frequency of adolescent drinking at age 15. However, when Icelandic adolescents do drink, they tend to drink more heavily than their peers in most other European countries (Hibell et al., 1997). Also, compared to adolescents in other European nations, when "under the influence ," both boys and girls in Iceland tend to get into more trouble: scuffles, unwanted and unprotected sex, and being victimized by robbery or theft (Hibell et al., 1997). The literature also drives the choice of focus: it has been claimed that insufficient attention has been paid to heavy drinking as a risk behavior that often has serious consequences, such as violence and unprotected sex (see Barnes & Farrell, 1992; Reifman et al., 1998).

Psychosocial maturity. The GSID Relationship Questionnaire (Rel-Q) measures psychosocial maturity (PS-Maturity) for children in grades 4 to 12 (Schultz & Selman, 2000). A standard form of the measure was used with a few additional questions on drinking, reflecting the focus of the study. In this form, the Rel-Q is a 25-item multiple-choice questionnaire that assesses psychosocial development on four subscales: interpersonal understanding, hypothetical negotiation, real-life negotiation, and personal meaning. For each question, adolescents rate each of the four possible responses as poor, average, good, or excellent. An example of a question is: Someone is a good friend because he or she: (a) does what you ask, (b) lives close by, (c) shares his or her feelings with you, (d) will keep your secrets. The participant then chooses the "best" response item. Thus, the Rel-Q has two scores for each scale: an item-rating score and a best-choice score. Because of the underlying developmental competence in perspective c oordination, best-choice score and item-rating score can be averaged for each scale (Schultz & Selman, 2000). Similarly, based on the theoretical claim of a structural coherence across the four subscales, an overall psychosocial maturity score was computed for each assessment, averaging the mean scores for each of the subscales in order to provide a more reliable measure.

The Cronbach's standardized alpha reliability coefficient on the overall psychosocial maturity scale was .90 at the first assessment (age 15) and .81 at the second (age 17), indicating good internal reliability. These coefficients corroborate results in a previous study using this scale (Schultz et al., 2001). The correlation between the overall psychosocial maturity scores across the two assessments was r = .55, indicating similarities in the two related scales used (Anastasi, 1990).