[Evaluation of theoretical accuracy, reliability, discriminative power and difficulty for physical, mental and social well-being scales in adolescents].

The purpose of this study was to develop and validate the questionnaire used to measure adolescents' wellbeing and its physical, mental and social dimensions. The questionnaire was composed of two parts: the first part contained wellbeing indicators and the second part contained criterion indicators (health disorders and harmful behaviours). Physical wellbeing scale contained the most common complaints due to psychosocial and life style factors, and was measured by three indicators: headache, abdominal pain and backache. Mental wellbeing scale contained emotional and moral dimensions that were recognised as closely connected with the physical disorders from one hands and the process of socialisation from the other hands. There were eight indicators: fatigue, stress, fear, depression, loneliness, helplessness, feeling of guilt, and low self-complacency. Five indicators: perceived social support, relationship with mother, father, friends and teacher measured social wellbeing scale. Each indicator of wellbeing is scored in three scales: dichotomise scale, five-point Likert scale and visual analogue scale. The cluster sample of 445 schoolchildren aged 14-15 years, randomly selected from the last grade of elementary schools of Warsaw was surveyed in October-November 1999. The physical, mental, social and total wellbeing scales were found to be reliable, but differing in internal consistency. The total and mental scales of wellbeing demonstrated high reliability, while the physical and social scales demonstrated moderate reliability. Analysis of correlation between criterion and tested variables showed acceptable discriminative power of the physical, mental, social and total wellbeing scales. Respondents assessed the five-point Likert scale as easier in comparison to dichotomise and visual analogue scales.