Previously, the evaluation results (four levels of academic achievement) for the paper-based National Assessment of Educational Achievement (NAEA) were available only two months after the assessment was conducted. At that time, the usefulness of the evaluation results was very low because the students interest and memory of the evaluation had already decreased. While surveying school sites for their input on the introduction of computer-based tests for the NAEA in 2022, the demand for immediate results was found to be high. Therefore, this study suggests a method to
immediately produce achievement level results using common items taught throughout the year for which an item s parameters can be secured, and employing multiple-choice and short-answer test items that can be scored automatically. A simulation was performed to explore the stability of the results. For data generation, the actual item parameters from the NAEA were used, and three conditions were considered for simulation: (a) the number of common items, (b) the selection method of common items, and (c) the number of participants. Findings show that calculation of immediate results is possible when at least six common items are selected using the item information function, and at least 1,000 participants are required for each test type. However, considering the distribution of
item difficulty on the current NAEA, the ability to discriminate between Level 1 and 2 students was found to be insufficient. Therefore, it is suggested that a large number of low-difficulty items suitable for discriminating between Level 1 and 2 should be provided for immediate and stable results.