Person Explanatory Item Response Theory Analysis: Latent Regression Two Parameter Logistic Model


ATAR B., Aktan D. C.

EGITIM VE BILIM-EDUCATION AND SCIENCE, cilt.38, sa.168, ss.59-68, 2013 (SSCI) identifier identifier

  • Yayın Türü: Makale / Tam Makale
  • Cilt numarası: 38 Sayı: 168
  • Basım Tarihi: 2013
  • Dergi Adı: EGITIM VE BILIM-EDUCATION AND SCIENCE
  • Derginin Tarandığı İndeksler: Social Sciences Citation Index (SSCI), Scopus, TR DİZİN (ULAKBİM)
  • Sayfa Sayıları: ss.59-68
  • Hacettepe Üniversitesi Adresli: Evet

Özet

In this paper, an application of latent regression two-parameter logistic (2-PL) model as an explanatory item response model (EIRM) was illustrated using a part of TIMSS 2007 Science data for Turkey. For this purpose, initially, item parameters were calculated via 2-PL IRT and latent regression 2-PL models. Then, in the latent regression 2-PL model, the effects of gender, positive affect toward science, valuing science, self-confidence in learning science, and education level of parents as person properties on the student achievement were examined. It was seen that among those properties, only self confidence in learning science and education level of parents had a statistically significant effects on explaining the differences in students' achievements. Based on fit indices, it was found that latent regression 2-PL model had a better model fit to data than 2-PL IRT model.