![İstatistik ve Uygulamalı Bilimler Dergisi](https://asosindex.com.tr/dergi-img/jsas.png)
İstatistik ve Uygulamalı Bilimler Dergisi
Yazarlar: Serpil AKTAŞ ALTUNAY
Konular:İstatistik ve Olasılık
DOI:10.52693/jsas.940764
Anahtar Kelimeler:BANOVA,ANOVA,Bayes Factor,Air Pollution
Özet: Bayesian approach is a posterior prediction method via a priori distribution knowledge on the contrary to classical methods. In the methods like BANOVA, Bayesian criteria are employed for the null hypothesis instead of the p-value. The question of whether only the overall mean term will represent the ANOVA model or whether the treatment effect will be added is answered with the Bayesian (BF). According to the BF, the null hypothesis is compared to the alternative hypothesis, and which model has stronger evidence is given accordingly. This decision step with the BF, which reveals how strong evidence we have about whether the treatment effect is significant or not, is a more detailed inference than classical ANOVA. In this study, the BANOVA method is applied to PM10, PM2,5, and SO2 data from eight stations of the province of Ankara for the period January-December 2018, taken from the National Air Quality Monitoring System of the Ministry of Environment and Urbanization. Whether there is any difference between stations is analyzed and conclusions are made.