Examining Group Differences in Mathematics Achievement: Explanatory Item Response Model Application

Erdem Boduroğlu; Duygu Anıl

Makale detayları

Examining Group Differences in Mathematics Achievement: Explanatory Item Response Model Application

Yazarlar:

Erdem Boduroğlu | Duygu Anıl

Yayıncı:

İdeal Kent Yayınları

Dergi:

OPUS Toplum Araştırmaları Dergisi Vol.20 Issue 53, pp.385-395

Belge Tipi:

Makale

Yayın Tarihi:

5/2023

Belge Dili:

İngilizce

Document ID (Org No.):

2001290611

pISSN:

2791-9862

Anahtar kelimeler:

Özet:

 Students take many different exams throughout their educational lives. In these exams, various individual and item characteristics can affect the responses of individuals to the items. In this study, it was aimed to examine the effects of person and item predictors on the mathematics common exam results of 365 9th grade students with explanatory item response models. Gender and school type as person variables and cognitive domain, content domain and booklet type as item variables were added to the models due to their widespread inclusion in the literature. When the predicted item parameters were examined, it was seen that the smallest parameter values were obtained for all items with the Rasch model. When the model data fit values of four different models were examined, it was concluded that the latent regression and latent regression linear logistic test models showed better fit than the Rasch model. By adding person and item predictors to the model, the parameters obtained for each variable group were compared, and differences were observed between the groups for school type, cognitive domain, and content domain variables. It was concluded that the item parameters did not differ for the variables of gender and booklet type. It is thought that it would be beneficial to use these models more widely in studies to be conducted in the field of education and psychology since they provide more detailed information about the reasons for the differences in the estimated parameters.

 Students take many different exams throughout their educational lives. In these exams, various individual and item characteristics can affect the responses of individuals to the items. In this study, it was aimed to examine the effects of person and item predictors on the mathematics common exam results of 365 9th grade students with explanatory item response models. Gender and school type as person variables and cognitive domain, content domain and booklet type as item variables were added to the models due to their widespread inclusion in the literature. When the predicted item parameters were examined, it was seen that the smallest parameter values were obtained for all items with the Rasch model. When the model data fit values of four different models were examined, it was concluded that the latent regression and latent regression linear logistic test models showed better fit than the Rasch model. By adding person and item predictors to the model, the parameters obtained for each variable group were compared, and differences were observed between the groups for school type, cognitive domain, and content domain variables. It was concluded that the item parameters did not differ for the variables of gender and booklet type. It is thought that it would be beneficial to use these models more widely in studies to be conducted in the field of education and psychology since they provide more detailed information about the reasons for the differences in the estimated parameters.

 Students take many different exams throughout their educational lives. In these exams, various individual and item characteristics can affect the responses of individuals to the items. In this study, it was aimed to examine the effects of person and item predictors on the mathematics common exam results of 365 9th grade students with explanatory item response models. Gender and school type as person variables and cognitive domain, content domain and booklet type as item variables were added to the models due to their widespread inclusion in the literature. When the predicted item parameters were examined, it was seen that the smallest parameter values were obtained for all items with the Rasch model. When the model data fit values of four different models were examined, it was concluded that the latent regression and latent regression linear logistic test models showed better fit than the Rasch model. By adding person and item predictors to the model, the parameters obtained for each variable group were compared, and differences were observed between the groups for school type, cognitive domain, and content domain variables. It was concluded that the item parameters did not differ for the variables of gender and booklet type. It is thought that it would be beneficial to use these models more widely in studies to be conducted in the field of education and psychology since they provide more detailed information about the reasons for the differences in the estimated parameters.

 Öğrenciler eğitim hayatları boyunca birçok farklı sınava katılmaktadır. Bu sınavlarda, çeşitli birey ve madde özellikleri öğrencilerin maddelere verdikleri yanıtları etkileyebilmektedir. Bu çalışmada 9.sınıflarda öğrenim gören 365 öğrencinin matematik dersi ortak sınav sonuçları üzerinde birey ve madde yordayıcılarının etkisinin açıklayıcı madde tepki modelleri ile incelenmesi amaçlanmıştır. Alanyazında araştırmalara yaygın olarak dahil edilmesi sebebiyle; birey değişkeni olarak cinsiyet ve okul türü, madde değişkenleri olarak ise bilişsel alan, içerik alanı ve kitapçık türü değişkenleri modellere eklenmiştir. Kestirilen madde parametreleri incelendiğinde, Rasch modeli ile tüm maddeler için en küçük parametre değerlerinin elde edildiği görülmüştür. Dört farklı modelin model veri uyumu değerleri incelendiğinde ise örtük regresyon ve örtük regresyon doğrusal lojistik test modellerinin Rasch modeline göre daha iyi uyum gösterdiği sonucuna ulaşılmıştır. Birey ve madde yordayıcıları modele eklenerek her bir değişken grubu için elde edilen parametreler karşılaştırılmış ve okul türü, bilişsel alan, içerik alanı değişkenleri için gruplar arasında farklılıklar gözlenmiştir. Cinsiyet ve kitapçık türü değişkenleri için ise madde parametrelerinin farklılaşmadığı sonucuna ulaşılmıştır. Bu modellerin, kestirilen parametrelerdeki farklılıkların nedenlerine ilişkin daha detaylı bilgiler sunması sebebiyle eğitim ve psikoloji alanında yapılacak çalışmalarda daha yaygın kullanılmasının faydalı olacağı düşünülmektedir.

Makale detayları

Examining Group Differences in Mathematics Achievement: Explanatory Item Response Model Application

Yazarlar:

Erdem Boduroğlu | Duygu Anıl

Yayıncı:

İdeal Kent Yayınları

Dergi:

OPUS Toplum Araştırmaları Dergisi Vol.20 Issue 53, pp.385-395

Belge Tipi:

Makale

Yayın Tarihi:

5/2023

Belge Dili:

İngilizce

Document ID (Org No.):

2001290611

pISSN:

2791-9862

Explanatory item response models

rasch model

math successExplanatory item response models

rasch model

math success

Sayı

Makale detayları

Examining Group Differences in Mathematics Achievement: Explanatory Item Response Model Application

Yazarlar: Erdem Boduroğlu | Duygu Anıl

Yayıncı: İdeal Kent Yayınları

Dergi: OPUS Toplum Araştırmaları Dergisi Vol.20 Issue 53, pp.385-395

Belge Tipi: Makale

Yayın Tarihi: 5/2023

Belge Dili: İngilizce

Document ID (Org No.): 2001290611

pISSN: 2791-9862

Explanatory item response models

rasch model

math successExplanatory item response models

rasch model

math success

Sayı

Yazarlar:

Erdem Boduroğlu | Duygu Anıl

Yayıncı:

İdeal Kent Yayınları

Dergi:

OPUS Toplum Araştırmaları Dergisi Vol.20 Issue 53, pp.385-395

Belge Tipi:

Makale

Yayın Tarihi:

5/2023

Belge Dili:

İngilizce

Document ID (Org No.):

2001290611

pISSN:

2791-9862