Measuring phonological awareness - blend phoneme test in the beginning of schooling: an comparison of classical test theory and item response theory
Abstract
The main purpose of this paper is to show dis/advantages of two approaches classical test theory (CTT) and item response theory (IRT) to assess the level of diagnostic tools for evaluating phonemic awareness among first-grade students using longitudinal survey of children during the first grade of primary school (25 schools in Prague and Central Bohemia, 439 pupils). Phonemic awareness (FA) was measured by the task blend phonemes into syllables and words.
The advantage of IRT compared to CTT is primarily a concept of independence of the test items. The probability of a response is dependent on the level of latent ability and characteristics of items.
So if we know the characteristic of items parameter (difficulty) it could be estimated from the pupil's responses the level of language skill for diagnostics. The results showed the need for blend phoneme test modifications (in particular for the end of the first grade) so as to adequately measure the high growth of FA ability levels during the first grade.