[Positive health behaviours scale for adolescents: initial psychometric analysis].

Purpose: to elaborate the Positive Health Behaviours Scale (PHBS) for use with adolescents and to provide its psychometric analysis.

Methods: the instrument consists of 30 statements comprised of five subscales (nutrition, relaxation and behaviours connected with psychosocial health, safety, taking care of one's body and physical activity) which were measured on a 4-point summated rating scale from 3 (almost always) to 0 (almost never). Psychometric analysis included assessment of reliability (using Cronbach's a coefficient Jöreskog's rho coefficients), theoretical validity (factor analysis) and construct validity (using data concerning perceived health, self-assessment of care of own health, psychoactive substance use and subjective health complaints; four hypothesis were tested) and unidimensionality (principal components). The sample consisted of 307 adolescents aged 13-17.

Results: Cronbach's alpha coeffient for the whole scale amounted to 0.823 but for three out of five subscales it was lower than the accepted value of 0.6. Jöreskog's rho coefficient was at satisfactory level for all subscales. Factor structure of the scale was different from what was expected. Construct validity of the subscales was satisfactory--all four hypotheses were confirmed. All subscales were also unidimensional.

Conclusions: psychometric analysis of PHBS for adolescents did not give univocal results concerning its quality. Only reliability of the whole scale was satisfactory. Further study with application of larger, randomly selected sample of adolescents is required. The scale can be used as a tool in health education and in prophylactic health studies of adolescents.