ขนาดกลุ่มตัวอย่างขั้นต่ำสำหรับการปรับเทียบคะแนนในแนวระดับ ตามแนวทฤษฎีการตอบสนองรายข้อ ระหว่างแบบที่มีแบบแผนการปรับเทียบ และความยาวแบบสอบแตกต่างกัน / พิชัย ละแมนชัย = Minimum sample size for horizontal test equating based on the item response theory among the tests with different equating designs and various test lengths / Pichai Lamanchai
The purpose of this study was to determine minimum sample size for horizontal test equating based on the Item Response Theory among the tests with different equating designs: equivalent group design and anchor test equating design, and various test lengths: 30, 50, 70, 90, 110, 130,and 150 items. The IRTDATA software was used to simulate data for this study. Discrepancy index in the range of very satisfy was the acceptable criterion. The results were as follows: 1. Minimum sample size of equivalent group design, 1-Parameter model with the test length of 30, 50, 70,90,110, 130, and 150 items were 1,000, 850, 700, 600, 550,450, and 350, respectively. 2. Minimum sample size of equivalent group design, 3-Parameter model with the test length of 30, 50, 70, 90,110, 130, and 150 items were 2,550, 2,350, 2,150, 1,750,1,550, 1,400, and 1,150, respectively. 3. Minimum sample size of anchor test equating design, 1-Parameter model with the test length of 30, 50, 70, 90,110, 130, and 150 items were 750, 650, 500, 400, 350, 300,and 200, respectively. 4. Minimum sample size of anchor test equating design, 3-Parameter model with the test length of 30, 50, 70, 90,110, 130, and 150 items were 2,000, 1,800, 1,600, 1,350,1,200, 1,000 and 800, respectively. 5. The 1-Parameter and 3-Parameter model of each test lengths required different minimum sample size. The 1-Parameter model required smaller sample size than the 3-Parameter model. 6. Among equating designs of each test lengths required different minimum sample size. The anchor test equating design required smaller sample size than the equivalent group design.
