Çok SınıflıLogit Model ile Öğrencilerin Fakülte Tercihlerinin Araştırılması Atatürk Üniversitesi Örneği

Fakülte tercihi, aslında bir meslek tercihi olup öğrencilerin birçok faktörden etkilendiği karmaşık bir karar verme süreci olarak değerlendirilmektedir. Yarım asrıaşkın bir geçmişe sahip olan Atatürk Üniversitesi’nde okuyan öğrencilerin fakülte tercihlerinde etkili olan faktörlerin tespitini amaçlayan bu çalışma, 900 öğrenciye uygulanan anketlerden elde edilen yatay kesit verilerine dayanmaktadır. Bu amaçla anne-babanın eğitim düzeyi ve mesleği, hane halkının geliri, hanedeki birey sayısıve ikamet edilen yerleşim birimi, mezun olunan lise türü, özel ders alma durumu ve yükseköğretimde okuyan birey sayısıgibi faktörlerin öğrencinin okuduğu fakülte tercihi üzerindeki olasıetkileri araştırılmıştır. Ayrıca, fakülte tercihinde etkili olabileceği düşünülen değişkenler dikkate alınarak aşamalı stepwise regresyon yöntemi ile çok sınıflı multinomial logit model denemesi gerçekleştirilmiştir. Model tahmin sonuçlarına göre, genel olarak babanın eğitim durumu, toplam fert sayısı, özel ders alma durumu ve yükseköğretimde okuyan birey sayısıdeğişkenleri %5 önem düzeyinde istatistikî bakımdan anlamlıolduklarıyani söz konusu fakültenin tercih edilmesinde, önemli değişkenler olduklarıtespit edilmiştir.

Universities are established with aims of producing scientific knowledge and training staff to produce this information emerged. Universities, to achieve these aims, perform with the assistance of faculties, colleges and etc. Faculty is defined as a higher education institution that engaged scientific research, education and teaching. Faculty preference is, in fact, a vocation preference and is evaluated as a complex process for making decision by students. Faculty preference is affected by many economic, demographical, social, and conditional factors and these factors. In this study, factors affecting student’s of faculty preferences are education level and vacation of the parents, income and the number of individuals in households, residential unit, the type of high school graduated, taking private lesson, the number of the individuals in higher education. These factors are examined whether they have an impact on the faculty preferences. The main purpose of this study is to determine the factors affecting the faculty preferences of the students of the Atatürk University. This study consists of four sections mainly. First section is introduction. In this section previously it is given general information about university than summarized literature related to this study. In the second section methodology of the study is discussed. In the third section characteristics of sampling and information obtained from the survey are introduced. The final section summarizes the results of this study. Ataturk University consists of 22698 undergraduate students studying at 13 faculties including the 12542 boys and 10156 girls. Depending on this population, the minimum sample size is determined approximately 647 at the level of 5 percent margin of error of 1% significance level. Although in this way to determine the minimum sample size, questionnaire in this study is applied to 900 students for representing power is high, and fills out some surveys at the thought of missing. Questionnaire the students in each faculty of applied research within the context of the overall share of students is taken into consideration. In addition, questionnaire is performed sharing in the rate of student's gender is taken into consideration. These 900 students are constituted 498 boys and 402 girl students. The data set is completed from the questionnaires applied to face to face is entered into Excel and transferred to the electronic media. Then these data, with the help of SPSS 15.0 program was analyzed in a detailed way. In this study, at first characteristics of sampling are introduced. Then some factors affecting distribution of the university student’s faculty preferences given and tested the relationship between these factors by using Pearson's chi-square statistics. The findings are given in the cross tables by taking the affecting factors student’s faculty preferences into account. In addition, given these variables, the stepwise regression method and multinomial logit model are estimated by using the data set compiled form a questionnaire applied to 900 students besides to some other statistical analyses. The some results obtained from the model estimation are as follows: • The chi-square independence test indicates that variables as gender, age, father's education level, mother's education level, father's vocational status, mother's vocational status, father's income, mother's income, total number of individuals, the number of the individuals in higher education, the place of residence, the classroom status, enrollment of private teaching institution and taking private lesson variables have statistically significant effect on faculty preference at the level of 1 percent. Whereas, the chi-square independence test indicates that variables as marital status, father’s graduated department of faculty, mother’s graduated department of faculty, father’s age and mother’s age variables have statistically insignificant effect on faculty preference at the level of 1 percent. • The father's education level primary education , mother's education level secondary education , the total number of individuals and taking private lesson yes are statistically significant at the level of 5 percent for Faculty of Dentistry. • The number of the individuals in higher education none and the total number of individuals are statistically significant at the level of 5 percent for Faculty of Pharmacy. • The classroom status, family income, mother's vocational status salaried , the number of the individuals in higher education none , the place of residence province , and taking private lesson yes are statistically significant at the level of 5 percent for Faculty of Letters-Science. • The classroom status, the total number of individuals and taking private lesson yes are statistically significant at the level of 5 percent for Faculty of Fine Arts. • The total number of individuals is statistically significant at the level of 5 percent for Faculty of Law. • The number of the individuals in higher education none is statistically significant at the level of 5 percent for Faculty of Communications. • The father's education level primary education , the father's education level secondary education and the number of the individuals in higher education none are statistically significant at the level of 5 percent for Faculty of Architecture and Design. • The classroom status is statistically significant at the level of 5 percent for Medical Faculty. • The father's education level primary education is statistically significant at the level of 5 percent for Faculty of Agriculture. • The classroom status, the father's education level primary education , mother's vocational status salaried , the number of the individuals in higher education none , the place of residence province , the total number of individuals and the total number of individuals are statistically significant at the level of 5 percent for Kazım Karabekir Faculty of Education