Methods to convert test scores into integer grades were discussed from both the classical and the IRT points of view. Using the classical test theory, a method to convert test score into the integer grade such that the resulting confidence interval of the true score has a prescribed width, say, 1 or 2, was presented, and a method to define the optimal number of grades was derived. From IRT point of view, it was shown that, if we calculate the average standard error of measurement, the similar method can be used to define the optimal number of grades. Also, a method to utilize the variance-stabilizing transformation was proposed.

This study analyzed the results of the National Center Test regarding the ratio to the sum scores of the five subject areas among successful public university candidates. The candidates’ success ratio to the sum scores did not increase for the upper middle achievers (the plateau effect of passing ratios), while ratios for other levels of academic abilities constantly raised with the sum scores. In order to examine this trend closer, the candidates were divided into three sub-groups by the levels of the colleges which the candidates actually applied. When the analysis was conducted within these three sub-groups, the plateau effect disappeared; the success ratio increased as the sum scores raised. It means that the candidates’ self-assessment affects their own decision-making on college application, which causes the competition ratio to be even across colleges. This phenomenon may be acting as a social system to equally distribute the candidates among universities.

Developmental aspects of academic achievement were analyzed based on longitudinal data collected during six years of elementary school. Participants were children attending a private elementary school (N=613; 275 boys and 338 girls). Their achievement test results (mathematics and Japanese) for six years were used analyzed in Study 1 by using latent growth modeling. Results indicated the poor goodness of fit of the model. Various patterns in changes of academic achievement during six years at elementary school were suggested. In Study 2, the patterns of changes were classified using the method suggested by Yamada (1990), and typical patterns were identified for both math and Japanese. Moreover, they were classified into three types. (1) Those not showing significant changes during six years. (2) Those showing drastic changes in the lower grades (1st, 2nd, and 3rd grades). (3) Those showing drastic changes in the higher grades (4th, 5th, and 6th grades). The above results support ed tendencies that have been indicated to date; high-achieving and low-achieving students are separated at an early stage and fixed, and the academic achievement of many students stagnate in 3rd and 4th grades.

This study examines the processes of thinking about and solving multiple-choice problems using the think-aloud method. We asked 18 high school students to solve two problems in the subject of Japanese Language, which were set in past tests of the National Center for University Entrance Examination. There, we found three features of the process of thinking about and solving multiple-choice problems. First, the results suggest that students did not read the entire sentences from start to finish and then solve the questions. Instead 89% of them read the parts of the sentences which they needed in order to solve the problem. Secondly, the students solved the problems mostly through a process of elimination. They solved half of the problem items based on a conceptual understanding of the sentences. Finally, we analyzed whether those problem-solving processes and strategies correlated to the percentage of correct answers. The results suggested that students who used a strategy based on a conceptual understanding of the sentences often chose the correct answers , but even those students did not connect the results of solving problems to make a coherent understanding of the whole . We discussed the possible negative effect of the problem structure and the act of choosing from multiple-choices. This seems to lead students to fragmented reading of the problem sentences.

Writing performance assessment, thought as an authentic approach to assess the ability of language learners, has been studied considerably in the field of language assessment. However, rater variation argued as a potential weakness of writing performance assessment, especially rater language background is a factor that needs to be considered. The aim of this paper is to review studies which examined the differences between native and non-native speaker teachers’ evaluation of students’ writing. These differences can be categorized into two major types: rater severity and rater assessment perspective. Prior studies found that non-native speaker teachers were more severe than native speaker teachers in rating the whole essays. Furthermore, the perspective of rhetorical patterns and content between native and non-native speaker teachers was also found to be different. With a focus on raters’ assessment perspective, it is necessary to clarify whether rating essays in Japanese as a second language differ between native and non-native speaker teachers, by a method such as Think-Aloud approach.