The reliability of the proposed short form of the Revised Illinois Test of Psycholinguistic Abilities (Newcomer & Hammill, 1974) was computed on data derived from 6 year old ITPA standardization test booklets. Measures of internal consistency and between forms reliability were obtained for both full length and short form scorings of single ITPA test administrations. Tests of significance yielded significantly lower measures on 10 of 12 subtests for the short form test when compared to the internal consistency of the full length test. The correlation between the two forms was significantly reduced in all cases when the part-whole correspondence was accounted for. The results were discussed in terms of their impact on the future use of the proposed short form as a psychometric instrument of questionable reliability. Suggestions were proposed for those concerned with reducing administration time.
A Note on Combining Dependent Tests of Significance
