การพัฒนาแบบวัดทักษะชีวิตและการทำงานในศตวรรษที่ 21 ของนักศึกษาวิทยาลัยชุมชนในพื้นที่สามจังหวัดชายแดนภาคใต้โดยประยุกต์ใช้ทฤษฎีการตอบสนองข้อสอบแบบพหุวิภาค

Abstract

The purposes of this research were 1) to develop of measurement scale on Life and Career Skill in the 21st Century of Community College Students in the Three Southern Border Provinces by applying polytomous Item response theory. and 2) to examine the quality of measurement scale on Life and Career Skill in the 21st Century of Community College Students in the Three Southern Border Provinces by applying polytomous Item response theory. The samples were 600 associated degree 3 Students at Community College in academic year 2020 the Three Southern Border Provinces selected by stratified random sampling technique was calculated by means of interpolation. The instrument was characterized as a multiple-choice scenario model with 4 options. And has more than two scoring characteristics, with rubric ratings ranging. Test Theory (CTT) by construct validity, discrimination and reliability with Cronbach's alpha coefficient, and the quality of the measurement scale by parameter testing was accomplished by confirmatory factor analysis (CFA) and parameter testing was accomplished by Polytomous (IRT), with discriminant and difficulty, and test information (TIF) was conducted via Grade Response Model (GRM), respectively. The research findings were as follows: 1. The measurement scale on life and career skills in the 21st century consisted of five factors: flexibility and adaptability, initiative and self-direction, social and cross-cultural skills, productivity and accountability and leadership and responsibility. The construct validity approved by experts was between 0.80 to 1.00. All the items discrimination with has 49 item has 0.923. The results were consistent of the empirical data 2= 9.06, df = 4, P-value = 0.05962, RMSEA = 0.113, RMR = 0.43, CFI = 0.98, GFI = 0.96 The weight of Standard factor was between 0.38 to 0.87 Internal consistency reliability (R2) were between 0.15 and 0.76. 2. When examining the quality of the measurement based on the Polytomous IRT. The slope parameter (α) is between 0.52 to 5.82 and the difficulty value the threshold value of each item (β) which all of the values in ascending items and the reliability of the measurement scale using Marginal Maximum-Likelihood (MML) showed that reliability of the test was 0.9737. Moreover, the value of test information indicated that test was appropriate to assess students competency whose competency θ was between -1.0 to 1.0 or the students of medium competency levels.