TR
Avrasya Ekonometri �statistik ve Ampirik Ekonomi DergisiYl:2016 Say: 3 Alan: statistik-Saysal Yntemler-Ekonometri

ebnem Zorlutuna, Necati Alp Erilli, Birsen Ycel
Tobit Regresyon Analizi ?le Akci?er Kanseri al??mas?: Sivas ?li rne?i
 
Tobit regresyon modeli, snrl baml deikenli modellerini aklamada yaygn biimde kullanlan ve en kk kareler regresyonunun parametrik olmayan alternatifi olarak tanmlanan bir yntemdir. Model; baml deikene ait bilginin yalnzca baz gzlemler iin bulunduu sansrl rneklem modeli olarak ta bilinmektedir. Baz deney srelerinde, baml deikenlerin tm deerlerine ulalamamaktadr. Baml deikenin ald deerler snrl olduu iin bu modellere Baml Deikeni Snrl Modeller ad verilmitir. Bu almada, snrl baml deikenli modelleri aklamada yaygn biimde kullanlan Tobit modeli kullanlmtr. Sivas Cumhuriyet niversitesi Tp Fakltesi Aratrma ve Uygulama Hastanesi Onkoloji Merkezinde, 2000-2011 yllar arasnda takip edilen akcier kanseri hastalarndan elde edilen verilere tobit regresyon analizi uygulanarak akcier kanseri hastalarnn lmnde (yaam srelerinde) baz bamsz deikenlerin etkisi llmeye allmtr. Sadan ve soldan sansrlenen modellerde, kanser hastaln tetikledii dnlen hastann ya, hastaln evresi, hastaln durumu deikenlerini etkileyen deikenler belirlenerek, hastalk iin veri altyaps oluturulmutur. Tobit Regresyon sonular hastann hastalnn evresi baml deiken iken, hastann cinsiyeti, hastann durumu, hastaln patolojik sonular anlaml deikenler olarak bulunmutur. Hastann cinsiyeti, hastaln evresine pozitif etki yaparken, hastann durumu ve patolojik durumu da negatif etki yapt grlmtr.

Anahtar Kelimeler: : Tobit Regresyon, Sansrl Veri, Akcier kanseri, Snrl Baml Deikenli Model, Yapay Deiken.


Lung Cancer Study with Tobit Regression Analysis: Sivas Case
 
: Tobit regression models are the limited dependent variable models in a statement which is widely used and is defined as a non-parametric alternative for the method of least squares regression. Model is known as knowledge of the dependent variable models for some samples containing censored observations. Some experimental process, cannot be achieved for all values of the dependent variable. Dependent variables for these models are called limited dependent variable which value limited by ordered models. In this study Tobit model is used for the analysis, which widely used in the statement in limited dependent variable models. Data taken from Sivas Cumhuriyet University Faculty of Medicine Research and Application Hospital Oncology Center, consists 535 patients who have lung cancer. Tobit regression analysis was applied in the measurement of lung cancer patients (the lifetime) was tried to measure the effect of certain arguments. Using right and left censored models, cancers that are thought to trigger the patient's age, stage of disease, determining the variables affecting the disease status variables , data infrastructure has been created for the disease. Tobit regression results when the dependent variable phase of the patient's disease, the patient's gender, patient's condition , the pathological consequences of the disease was found to be statistically significant variables. The sex of the patient, while a positive effect on the stage of disease, condition and pathological condition of the patients were found to negative influences.

Keywords: Tobit Regression, Censored Data, Lung Cancer, Limited Dependent Variable Models, Artificial Variable.


Detay

ÇERK