An Evaluation of the Two Parameter (2-PL) IRT Models Through a Simulation Study


OLMUŞ H., Nazman E., Erbas S.

GAZI UNIVERSITY JOURNAL OF SCIENCE, cilt.30, sa.1, ss.235-249, 2017 (ESCI) identifier identifier

  • Yayın Türü: Makale / Tam Makale
  • Cilt numarası: 30 Sayı: 1
  • Basım Tarihi: 2017
  • Dergi Adı: GAZI UNIVERSITY JOURNAL OF SCIENCE
  • Derginin Tarandığı İndeksler: Emerging Sources Citation Index (ESCI), Scopus
  • Sayfa Sayıları: ss.235-249
  • Anahtar Kelimeler: Two-parameter item response theory, Joint maximum likelihood estimation, Parameter estimation, Discrimination parameter, Ability parameter
  • Sivas Cumhuriyet Üniversitesi Adresli: Hayır

Özet

The aim of the study is to evaluate parameter estimation of two-parameter item response theory (2-PL IRT model) using Joint Maximum Likelihood (JML) estimation technique. Hence, a simulation study is conducted in terms of various sample sizes, item numbers and ability parameter levels for low and high discirimination parameter levels. Hereby, the examinee ability parameter, item discrimination parameter and difficulty parameter are obtained as well as Test Information Function (TIF) and point-biserial correlation. One of the highlighted results indicates that the level of discrimination parameter plays an important role in parameter estimation for 2PL IRT models.