Primary Scholar'Chinese Reading Ability Scales: Development, Validity, And Reliability

Goals: The present research is aimed at: (1) to construct the Chinese reading ability (2) to make up a complete system of Chinese reading ability measurement scales applicable to all the three stages of primary schools;(3) to exam the three Primary School Reading Ability.Method: 5 public primary schools participated in the current research randomly. The data of 795 two graders, 684 four graders and 624 six graders are applicable. With the item analysis, a coefficient, half-split reliability were used to exam the reliability of the evaluation scales; In addition, the multi-generalizability theory was used to test the whole reliability of the scales, the whole reliability of the scales as a standard reference test as well as the determination reliability of the ranking cut-lines. In the meanwhile, the validity tests include the content-validity, criteria-validity and the construct validity. Finally, the factor-analysis method was used to exam the Chinese reading ability module initiated by the present research.Results: (1) with item-analysis, the average item difficulties of the three formal tests respectively are: 0.62 for junior graders;0. 62 for mid-graders and 0.55 for senior graders; and the average item discriminations are respectively 0.34, 0.31 and 0.30; most of the item discriminations are more than 0.3 and the item discriminations of all the items are statistically significant. (2) with the classical measurement theory, the α coefficient for junior graders is 0.80; that for mid-graders is 0.81 and that for senior-graders is 0.82; the half-split reliability of junior graders is 0.80, that for mid-graders is 0.82 and that for senior graders is 0.86. (3)with multi-generalizability theory, the scales for junior graders' reading ability suggested that Composite AbsoluteError Standard Deviation is 0.0058; and Composite Error Standard Deviation for Mean is 0.0063; the Generalizability coefficient is 0.81; the index of dependability is 0.80; and the reliabilities of the decision of ranking cut lines are above 0.99. (4) with agreement of all the professional researchers, the design of all the sub-scales is considered to cover all the Chinese reading areas and the items of all the sub-scales are considered to be typical. It suggested that the contents of the current scales were relatively good. (5)the relationship between pupils' test scores and teachers' grading are aboveO.7, P<0.01, suggesting high agreement and good validity. (6) The construct validity of the scales is relatively good: the relationships of the whole scale and sub-scales are between 0.48-0.66, most of the relationships of the sub-scales are low(<0.3), very few relationships are in mid-level (above0.4 but less than the relationship of the whole scale and the sub-scales) . This suggested relatively good construct validity. (7) The results of Confirmatory Factor Analysis suggest that the fitness of the theoretical module and practical testing data is good, which supports the theoretical construction of the scales.Conclusion: (1) three evaluation scales for primary school students' reading ability have been made up. They relatively completely cover all the Chinese reading areas, with good item difficulty, item discriminations, validities and reliabilities. (2) The Chinese reading ability can be divided into four parts: microstructure reading comprehension ability, macrostructure Reading comprehension ability, the ability to give comments on reading materials and emanative reading ability.