Pengujian Model Pengukuran Congeneric, Tau-Equivalent dan Parallel pada Satisfaction With Life Scale (SWLS)
Abstract
Many researchers make an error in data analysis, where researchers analyzing data using the raw score on the instrument with an ordinal scale. Error in the use of raw score for an instrument with an ordinal scale can be overcome by using measurement model testing, namely tau-equivalent and parallel. The purpose of this study is to examine the best measurement model of the Satisfaction with Life Scale (SWLS). The research method is Secondary Data Analysis approach (SDA). The secondary data was combined from two previous studies. The quantitative research analysis technique used to test the three measurement models in SWLS was confirmatory factor analysis. The unidimensional model of confirmatory factor analysis indicates that tau-equivalent is the best measurement model in SWLS testing (χ2(9)=13.759, p > .05 and RMSEA < .05). Based on the result, an implication measuring instruments using raw score can be used while measurement model testing of an instrument is tau-equivalent.
Moraref