A Study Of Reliability And Validity Of Adolescent Deviant Behaviors Questionnaire In Macao

The economy of Macau is led by the gambling industry. The opening up of the gambling industry has been successfully supporting the economic development, but it also creates many social problems. Nowadays, the juvenile crimes become more serious-escalating from damage of others' property, such as theft and burning cars, to violent crimes harming others' life, such as such as holding guns, fighting, and cut killing. In 1994, Macao scholars Li Xiao Peng, Tsoi Zhang and other scholars devised "Adolescent Deviant Behaviors Questionnaire", which has been put into use to now. Until now, there is not any study to check the reliability and validity of the questionnaire. Therefore, this research aims to study the reliability and validity of this questionnaire.The target samples of the research are students of 2 secondary schools and juveniles in a special juvenile education institute (that takes care of juveniles who committed crimes). There were 1,641 samples in the survey. The survey was based on the latest version of the questionnaire revised in 2002. The main statistical analysis covers assessment index of items, individual items' relatedness to the total scores, the consistency among items, the test-retest reliability and interjudge reliability, explore factor analysis and discriminant analysis.According to the research results, only about half of the 50 items in the questionnaire are with assessment index lower than 0.20. These items are more related to crimes and the frequency is low, thus they should be revised or deleted. Regarding reliability, the internal consistency is relatively high, which reaches 0.90 or above; however, the test-rest reliability (0.79) is not yet satisfactory. Regarding construct validity, there are 9 factors with Eigenvalues over 1, which does not comply with the classification of 5 groups in the questionnaire; moreover, there exist relatively large number of items with loading and double loading lower than 0.40, which is the most important part of the questionnaire that requires improvement. Empirical validity study involves the study of various aspects, including the differences of genders, repeating classes or not, families of different martial status of parents, and normal juveniles vs. those committing crimes; the results support that the questionnaire has certain level of validity. In summary, this research has shown that the questionnaire has certain levels of reliability and empirical validity, but it also has quite many problems. Therefore, the questionnaire should be further improved. The research results and the recommendations may be used as a base for the questionnaire in psychological test and provide supporting information for future revision.