পদ্ধতির তুলনা করুন

	বহুপদী পার্থক্যের আইটেম কার্যকারিতা (Polytomous DIF)	কনফার্মেটরি ফ্যাক্টর অ্যানালাইসিস (CFA)
ক্ষেত্র	মনোমিতি	মনোমিতি
পরিবার	Latent structure	Latent structure
উদ্ভবের বছর	1990s–2000s	1969
প্রবর্তক	Bruno D. Zumbo and colleagues (ordinal logistic regression framework); Robert D. Ankenmann, Hariharan Swaminathan and others (IRT-based extensions)	Karl Gustav Jöreskog
ধরন	Measurement fairness / item bias detection	Hypothesis-testing latent variable model
মৌলিক উৎস	Zumbo, B. D. (1999). A handbook on the theory and methods of differential item functioning (DIF): Logistic regression modeling as a unitary framework for binary and Likert-type (ordinal) item scores. Directorate of Human Resources Research and Evaluation, Department of National Defense.	Jöreskog, K. G. (1969). A general approach to confirmatory maximum likelihood factor analysis. Psychometrika, 34(2), 183–202.
অপর নাম	Polytomous DIF, DIF for polytomous items, ordinal DIF analysis, graded-response DIF	CFA, confirmatory FA, measurement model, restricted factor analysis
সম্পর্কিত	4	4
সারসংক্ষেপ	Polytomous differential item functioning detects whether a test or survey item with more than two ordered response categories (e.g., Likert-type scales, partial-credit items) functions differently across groups such as gender, ethnicity, or language background, after controlling for the latent trait being measured. It extends classical binary DIF methods to ordinal response formats.	Confirmatory factor analysis tests a researcher-specified factor structure against observed data. Unlike exploratory approaches, the researcher decides in advance which indicators load on which latent factor, and the model is evaluated by how closely the implied covariance matrix reproduces the sample covariance matrix. CFA is central to scale validation, construct validity assessment, and measurement invariance testing.