Publication: Bireylerin otomobil tercih seçiminin çoklu tercih modelleri ile analizi
Abstract
Bu çalışmada bireylerin otomobil segmenti tercihinde etkili olan faktörler çoklu tercih modelleri ile incelenmektedir. Tez üç bölümden oluşmaktadır. İlk bölümde nitel tercih modellerinin genel özellikleri ve iki durumlu tercih modelleri yer almaktadır. İkinci bölümde çok durumlu tercih modelleri incelenmektedir. Son olarak üçüncü bölümde ise bireylerin Small, Lower Medium, Medium ve Small MPV segmentlerine ait otomobil alımlarını etkileyen faktörlerin çok durumlu logit model ile belirlenmesine yönelik uygulamaya ayrılmıştır. Nitel tercih modellerinde amaç karar vericinin sahip olduğu özellikleri ile belli bir tercihte bulunma olasılığı arasındaki ilişkiyi ortaya koymaktır. Bağımlı değişken karar vericinin göstermiş olduğu tercih eylemi olduğu için tercih kümesinde yer alan seçenekler Small, Lower Medium, Medium, Small MPV segmentleri olmuştur. Modelin bağımsız değişkenleri ise bu tercih eylemini etkilediği düşünülen faktörlerdir. Bağımlı değişkenin nominal değerler alması ve alternatif sayısının ikiden fazla olması uygulama olarak çoklu tercih modellerine yöneltmiştir Bu uygulama için soru formu oluşturulmuş 499 denek ile görüşülmüştür.
In this study, factors that are influential in the people’s preferance of automobile segment is analyzed in the form of multiple preferance models. This thesis consists of three main sections. In section one, basic features of the qualitative preference models and binary preferance models, that are special conditions of multiple conditioned qualitative preference model, can be viewed. Whereas, section two consists multiple conditioned preference models. In the multiple conditioned preference models a decision maker should make a choice between three or more alternatives. In these models there is only one dependent variable that can have more than two values. Consequently, section three is allocated for factors that influence the automobile buying preference of segments ,Small, Lower Medium, Medium, Small MPV, and the execution due to the specification of these factors with the multinominal logit model. The goal in the qualitative preference models is showing the relationship between the qualifications of the decision maker and the probability of that specific person of making a preference. Since the dependent variable is the preferable action of the decision maker, choices that exist in the preference cluster were observed as Small, Lower Medium, Medium, and Small MPV segments. Model’s independent variables are factors that can be expected to influence this preference action. The reason of choosing multiple preference model for this study is that the dependent variable takes nominal values and the number of the alternatives is more than two. A questionnaire was created for this study and 499 interviews were conducted.
In this study, factors that are influential in the people’s preferance of automobile segment is analyzed in the form of multiple preferance models. This thesis consists of three main sections. In section one, basic features of the qualitative preference models and binary preferance models, that are special conditions of multiple conditioned qualitative preference model, can be viewed. Whereas, section two consists multiple conditioned preference models. In the multiple conditioned preference models a decision maker should make a choice between three or more alternatives. In these models there is only one dependent variable that can have more than two values. Consequently, section three is allocated for factors that influence the automobile buying preference of segments ,Small, Lower Medium, Medium, Small MPV, and the execution due to the specification of these factors with the multinominal logit model. The goal in the qualitative preference models is showing the relationship between the qualifications of the decision maker and the probability of that specific person of making a preference. Since the dependent variable is the preferable action of the decision maker, choices that exist in the preference cluster were observed as Small, Lower Medium, Medium, and Small MPV segments. Model’s independent variables are factors that can be expected to influence this preference action. The reason of choosing multiple preference model for this study is that the dependent variable takes nominal values and the number of the alternatives is more than two. A questionnaire was created for this study and 499 interviews were conducted.