Criterion-referenced Testing and Its Problematic Estimate of Reliability

Abstract

Two different families of test criterion-referenced and norm-referenced tests have been around for use of decision-making bases in education. One kind can be better explained by comparison with the other. The article does so here and there along the way in its attempts of presenting and discussing about criterion-referenced test in language testing, especially the contrastive practices within this test family as opposed to the other kind, which has so close a relationship to classroom teachers on a daily basis. The article describes its purpose, hence several kinds of CRT as a result of different purposes of testing, what is measured and how measuring is done, how is test results interpreted, and how reliability as one of the quality of good testing is estimated. Discussion of validity as another test quality indicator is purposely put aside here since in this area the contrastive practice is only slimly found.

Keywords: double entry diaries, strategy, reading comprehension.