Development of Work Competency Indicators of School General Service Officers in the Northeast

Abstract

The objectives of this study were to develop the indicators and to examine the construct validity of the work competency indicators of general service officers in schools in the Northeast. The sample consisted of 408 general service officers from primary schools in 8 provinces in the Northeast, obtained through two–stage cluster sampling. The instrument for collecting empirical data was a 5-point rating scale close-end questionnaire containing 54 questions with the value of discrimination (r_xy) from 0.56 to 0.92, and the value of the total reliability of 0.99. The statistic employed in this study was confirmatory factor analysis.

The results revealed that:

      1. The work competency indicators of general service officers in primary schools in the Northeast consisted of 3 aspects with 10 primary indicators and 54 secondary indicators. The competency was divided into core competency, job competency and individual competency. The core competency had 5 primary indicators and 20 secondary indicators while the job competency had 2 primary indicators and 13 secondary indicators, and the individual competency had 3 primary indicators and 21 secondary indicators.

      2. The analysis for examination of consistency and harmony of the theoretical structured model and the empirical data found that the structured model was consistent and harmonized with the empirical data. The components which had the most weight of significance in indicating work competency of general service officers were the job competency, individual competency, and core competency, with the weight of 0.965, 0.831, and 0.615 respectively. The root mean square error of approximation between the model and empirical data had the value of Chi-square = 0.639, the value of probability = 0.8874 at the degree of freedom (df) = 3, the value of CFI = 1.000, the value of TLI = 1.011, the value of SRMR =0.005 and the value of RMSEA = 0.000. In conclusion, the results revealed that the model had construct validity.