Scrutinization of Risk Analysis Related To Success of the Private Skill Exam by Using Mixture Logistic Regression Model

Abstract

In this study, the successes of students who participated in private skill exam were modelled by using logistic regression. The sample was consisted of 642 (150 girls + 492 boys) individuals who had participated in private skill examination in a university which is located in the East of Turkey. According to success circumstances, dependent variable which was dichotomic was coded as "1" or "0". After dichotomic design, success and failure circumstances were examined in terms of sub-population and covariance's changing. At the same time, it would have been like to search the risk ratio for each gender. Therefore, effecting of independent variables on dependent variable was modelled with ln(P(y(i) = 1)/P(y(i) = 0)) = logit(pi(i)) = ln(pi(i)/1-pi(i)) = beta(0) + beta(1)x(i1) + ... + beta(k)x(ip) logistic regression model. Logistic and risk analysis' were performed by SAS software. As to risk analysis findings, successful male students have more 15 times advantage than unsuccessful students. Besides, successful female students have more 8 times than unsuccessful female students. Wald statistics had been confirmed the risk analysis findings in appropriate border.