Parent-child relation: Scale for the evaluation of the family environment of adolescents.

The study of the family environment has been very extensive, allowing the development of different evaluation instruments on different areas (communication, cohesion, affect and control, among others). However, more research is required in Mexico, because publications on the developed instruments do not explain the psychometric characteristics. Other researches have developed indicators focused on the evaluation of drug use or alcohol problems in the family. This paper presents the validity and reliability of a scale aimed to evaluate the adolescente's perception of his family environment (communication, support and cohesion). This scale is important because it may be applied to evaluate the family environment and its relation with problematic behaviors such as drug use, delinquency, suicidal attempt and so on. The sample of this study included 793 young high school students from private and public schools located in Mexico City. The average age of the subjects was 15.3 years. The final scale consisted in 42 items with four frequency options in a Likert like scale. Besides, the questionnaire had different sections related to sociodemographic variables, drug consumption, depression, suicidal attempt, social insecurity and other elements of family environment. The instrument was selfapplied in group in the classroom. The confidential aspect of the student's answers was emphasized. The reliability results were satisfactory (alpha = 0.95) for the total scale. The factor analysis results with varimax rotation and GLS method gave 5 factors. One of these factors apparently evaluated two diferent conceptual dimensions (parents support and significative child support), instead of one. This aspect was tested through a Confirmatory Factor Analysis (CFA) that supported the two factor model. The reliability of each resulting area was higher than 0.68 in all cases (hostility and rejection = 0.79; parent's communication = 0.84; child communication = 0.88; parent's support = 0.83; significative child support = 0.68; daily child support = 0.71). The results were congruent in general with the areas that we pretended to measure with the instrument (communication and support). Cohesion did not form a factor as we expected. Instead, we obtained a more specific dimension: hostility and rejection, that grouped these items. This is congruent with the findings of other authors. These results, as well as the psychometric characteristics reported, support its applicability to evaluate adolescent's perception of family environment, as well as its relation with different problematic behaviors. Finally, we believe it is necessary to further investigate aspects related with the criterion validity of the instrument, specially in families with specific dysfunctions.