Abstract:
The purposes of this research were 1) to evaluate and compare research qualities on construction of aptitude tests, having different characteristic. 2) to synthesize research about the quality of aptitude tests, comprising of validity, reliability, difficulty and discrimination that have different research characteristics by meta-analysis method. The researches to be synthesized were 61 research reports published in 1966-2008. The data for this research consisted of 957 standard indices which were divided into 280 standard indices of reliability, 201 standard indices of validity, 238 standard indices of difficulty, and 238 standard indices of discrimination. Descriptive statistic and ANOVA were used to analyze the data. The results of research synthesis were: 1. The quality evaluation results of the research reports revealed that the total quality of the research reports were moderate level. From the results of analysis of variance, it was found that the research characteristics variables cause the differences in the research quality means were statistically significant at .001 level such as year of publication, the field that produce the research, type of research, instrument construction, sampling method, data collection procedures, theory and method which used in construction, and table of specification. The variable that was statistically significant at .01 level was number of item tryout. In addition, the samples the domicile, and aptitude test which used in construction were statistically significant at .05 level. 2. The research characteristics variables cause the differences of reliability, validity, difficulty, and discrimination such as the institute where produce the research, the theory, the samples education level, the samples the domicile, and time for testing