earticle

영어

Academic debates have been heating up regarding the estimation stability of formative measurement, making it the most controversial issue. The focus of this study is to examine the issue by attempting to understand the nature of error variance of the formative construct. For this, two different analytical approaches are taken: (1) the relationship between the error variance of a formative construct and the correlation strength between dependent variables and (2) the effect of reduced concept space of a formative construct on its error variance. Using actual survey data, we have shown that there is a significant detachment between the theoretical meaning of the construct error variance and its empirical meaning. The evidence underscores that caution is necessary in interpreting studies conducted on the formative approach.