The test result of language ability for infants of second year of life was examined statistically. In descriptive statistics, average score, standard deviation, and proportion of right answers for all test items for each age group subjects were included. Then factor analysis was carried out to determine the structural characteristics of the test. The analysis revealed that test pertains of the following five stages. First two stage of five were overlapped with the previously reported result for pre-speech language test for infant within one year of life. Then, the stage was determined as follows;
(4) Initial language acquirement and differentiation of language functions
(5) Speech increasing stage (increase of spontaneously uttered words and understood words related to persons and objects)
(6) Differentiation of expressed words (expression of words differentiated to suit objects)
(7) Understanding of words representing actions or conditions
(8) Understanding of words representing uses or functions, and verbalization of words representing actions or movements
In spite of the fact that six and seven stage represent the same developmental level in term of chronological age, they were distinctive into those which achieve speech understanding earlier (stage 7), and others whose development centers on speech expression (stage 6).
Then reliability and validity was performed. Reliability:(1) As to homogeneity, the reliability coefficients were calculated on the 60 test items by Spearman-Brown's odd-even method.
(2) As to inter-rater reliability, Pearson's product movement correlation coefficient was calculated based on the results produced by two separate experienced testers.
(3) As to test stability, reliability was determined through repetitive testing. Correlation coefficient was determined for each 60-day interval between age 0 and 840 days. These results suggested that the test was clinically available well.
Varidity:(1) As to predictive varidity, it was determined by the same way of repetitive testing. The correlation was higher on the age groups of 300 to 720 days. (2) As to content varidity, test content was explained, then a two dimension array was made from test items, composed of age groups and aspects of expressive and comprehensive language. It proved that the test items have an appropriate distribution in it.
抄録全体を表示