Assessing Validity and Reliability of Hookah Dependence's Questionnaire
Zahra pourmohammad shandiz
This current study aims to assess validity and reliability of questionnaire of Hookah dependence. The current sturdy has descriptive-analytical method of cross-sectional type in order to determine validity and reliability of hookah dependence's questionnaire for people ranged from 18 to 60 years of age in Mashhad city. The population of this study is all hookah consumers. The sample is 120 hookah consumers,ranged from low- to high-smoking,chosen by purposefulsampling method while attending recreationand tourist places (given that intended society has specific properties). The tool of study consists of researcher-made questionnaire, having 12 items, isbased on five-point Likert Scale ranging from always to never. Using statistical software spss with exploratory factor analysis and principal component analysis in analysis, research questionsare examined. To determine validity, Cronbach's alpha and split-half method are used. Formal Validity of the questionnaire is examined and approved by four relevant professors. In all questions, achieved average was higher than 4. Totally, assessing the professors' views suggest formal validity of given tool. Exploratory factor analysis is used to assess the construct validity as well. The result of this study is that exploratory factor analysis is used to determine latent factors in tendency tohookah. In first step performing exploratory factor analysis shows that KMO index in the current study is 0.79 being acceptable rate and suggesting that selected sample size (n=120) is enough to perform analysis. Barthel Index in reviewing adequacy matrix is 303.7 being significant in P≤0.01 level. That is, achieved matrix has necessary adequacy and the data of this research has the ability of factorization. The correlation matrix suggests that the value of KMO for every element in questionnaire is more than 0.7 indicating adequacy of this index for every element of questionnaire. In addition to this review, Reproduced and residual correlations show that the figures related to residual correlation is small for all questions and this means that factor analysis gives a good explanation of data. Three factors have ability to explain about 61.67% of variance. According to varimax rotation, reviewing the content of each factor shows factor 1, interest in hookah smoking, factor 2, spending time and money, and factor 3, confrontation with opposition. Also Cronbach's alpha coefficient is used to assess reliability of tool. Calculated alpha coefficient for questionnaire is 0.79 that is acceptable rate. Achieved alpha coefficient in factor 1 (interest in hookah smoking) is 0.83, in factor 2 (spending time and money) 0.74 and in factor 3 (confrontation with opposition) 0.73. In addition to Cronbach's alpha coefficient, split-half method is used. Reported correlated coefficient for odd and even questions of test is 0.83 suggesting strong correlation between two parts, so the reliability of and validity of tool are approved.