ABSTRACT
@#The paper describes the validation of the Malay Preschool Language Assessment Tool (MPLAT), a standardized normed referenced language assessment tool for Malay preschoolers within the ages of 4;0-6;11 and whose native language is Malay. The MPLAT (A Razak et al. 2010) is an assessment tool which is designed to assess the areas of receptive language, expressive language and early literacy skills. The MPLAT contains six subtests i.e. picture vocabulary, grammatical understanding, sentence repetition, referential meaning, relational meaning and early literacy skills. This paper is divided into 2 studies. Study 1 tested the psychometric properties and normative data of the diagnostic version of MPLAT on 300 participants recruited from preschools in the rural area of Gua Musang and the urban area of Kota Bahru located in the East Coast state of Kelantan. The construct validity is high with a strong correlation (r = 0.942) between test scores and age, validating the developmental nature of the test. There was also moderate to strong positive correlation between each subtest and the test total, as well as between subtests. It also has high test-retest reliability (r = 0.998 (p < 0.01) and inter-rater reliability (rho = 1.000). The second study looked at the psychometrics properties of the MPLAT short version (screening) compared to the full version diagnostic. The short version of MPLAT is about a quarter of the full (diagnostic) version. About 108 subjects in the urban area of the the Klang Valley, consisting of Kuala Lumpur and Selangor, were administered both versions of the MPLAT. The Pearson correlation revealed a strong positive correlation between the total scores and age (r = 0.718, p<.01), and strong positive correlation for the test-retest reliability (r = 0.881, n=12) for the short MPLAT version. For the full version, the Spearman correlation revealed a strong positive correlation between total scores and age (r = 0.791, p<.01) and a very strong positive Pearson correlation for test-retest reliability (r = 0.943). Cronbach's Alpha values demonstrated the internal consistency of the full version (0.972) and short version (0.929). In summary, both versions of the MPLAT were found to be valid tools to screen and diagnose language problems among Malay preschool children in Malaysia. MPLAT also has the potential to be a useful research tool to delineate language development of the preschool Malay children.