Second, the ability to account for the effects of estimated measurement error of latent variables is a major difference between SEM and both path analysis and multiple regression analysis. This is particularly relevant to management accounting research when composite measures are often used to measure constructs. Also, the use of interaction terms in multiple regression analysis may increase this problem. Interaction terms in multiple regression may encompass significant measurement error, particularly when used with composite variables. This can lead to bias in the estimation of coefficients of interaction terms, and can undermine significance tests [Jaccard and Wan,1996]. These problems have led prominent management accounting researchers to suggest that multiple regression techniques am inappropriate in many situations [Shields and Shields, 1998; Hartmann and Moers, 1999].
Second, the ability to account for the effects of estimated measurement error of latent variables is a major difference between SEM and both path analysis and multiple regression analysis. This is particularly relevant to management accounting research when composite measures are often used to measure constructs. Also, the use of interaction terms in multiple regression analysis may increase this problem. Interaction terms in multiple regression may encompass significant measurement error, particularly when used with composite variables. This can lead to bias in the estimation of coefficients of interaction terms, and can undermine significance tests [Jaccard and Wan,1996]. These problems have led prominent management accounting researchers to suggest that multiple regression techniques am inappropriate in many situations [Shields and Shields, 1998; Hartmann and Moers, 1999].
การแปล กรุณารอสักครู่..
