Item analyses of measuring teacher competency among pre-service teaching Chinese as a foreign language teachers

Abstract

This paper emphasizes that pre-service Teaching Chinese as a Foreign Language (TCFL) teachers must be professionally trained since researchers reported a shortage of qualified TCFL teachers in different countries. This study aims to analyze the instrument for measuring pre-service TCFL teachers’ competency using the Rasch Measurement Model regarding validity and reliability. It measures five domains: pedagogical knowledge, teaching skills, self-esteem, working attitude, and job motivation. This paper applies a quantitative research design. The researcher distributed the instrument with 70 items to 80 pre-service TCFL teachers. The Rasch Measurement Model is applied to analyzing and examining the psychometric properties by Winsteps version 3.70.0. The reliability based on Cronbach Alpha is 0.92, according to the results of the psychometric characteristics analysis. Positive Point Measure Correlation (PMC) value, infit, and outfit MNSQ value between 0.5 and 1.5, and ZSTD value range from -2.0 to 2.0 are used to assess the instrument’s construct validity. The person and item dependability are, respectively, 0.90 and 1.00. The item separation index has a value of 19.62, and the person separation index has a value of 2.97, both of which are regarded as excellent separation. The conditions for unidimensionality are met by the raw variance explained by measure, 86.7%, and unexplained variance in the first construct, 2.4%. The results demonstrate that the instrument is valid and reliable for future analysis, and the instrument is understandable by the respondents without any major problems. This study could increase public understanding and attention to the value and importance of teacher competency. In addition, a psychometrically teacher competency measuring instrument can examine the level of teacher competency of pre-service TCFL teachers and provide a result of what kind of academic and psychological components they need to improve in the future.