The research hypotheses were tested via the following eight multivariate regression models: (Equation 1)(Equation 2)(Equation 3)(Equation 4)(Equation 5)(Equation 6)(Equation 7)(Equation 8) where DET_1 is the extent to which the system analyses costs by cost centre, product and activity; DET_2 is the extent to which the system allows the preparation of customized reports according to users' specification; DISAGG is the extent to which the system classifies costs according to behaviour; VAR is the extent to which the system calculates variances; FREQ_1 is the extent to which the system provides frequent reports on a systematic basis; FREQ_2 is the extent to which the system provides information upon request; RELEV is the extent to which cost information is relevant for decision making; ACC is the extent to which cost information is accurate; TIME is the extent to which cost information is provided in time; DATE is the extent to which cost information is up‐to‐date; NEEDS is the extent to which cost information meets users' needs; APPR is the extent to which cost information has the appropriate level of analysis; REL is the extent to which cost information is reliable; and USE is the extent to which cost information is used to make decisions.
The correlation matrix of all variables is presented in Table IV. Pearson's correlation coefficients for combinations of all variables demonstrate significant associations between cost systems structure dimensions and information quality properties in the expected direction. A similar picture is presented in relation to Spearman's correlation coefficients. The correlations among the structure features of the cost accounting systems (darker‐shaded area) are all positive and strong in terms of statistical significance, which implies that these features even though they correspond to different characteristics they share positive relationships within a cost accounting system. Additionally, the correlations among the information quality characteristics (lighter‐shaded area) are also positive and strong in statistical terms. These statistically significant correlation coefficients have values that indicate the existence of close bonds among quality characteristics.
The results of the ordinary least squares (OLS) regressions are displayed in Table V. All models are significant (Fsig=0.000) and the adjusted R2 range from 25.1 per cent to 50.3 per cent. The results of regression Equation (1) reported in Table V indicate that RELEV is positively and significantly associated with the degree to which variances are calculated (VAR) and the extent to which reports are provided on a systematic basis (FREQ_1), while, contrary to expectations, a negative and significant association was found between the extent to which costs are analysed by cost centre, product and activity (DET_1) and cost accounting information relevance (RELEV).
Based on this unexpected negative association, it can be inferred that detailed cost information may be satisfactory in terms of accuracy, reliability and meeting managers' needs as discussed below but at the same time the analysis of cost data according to general criteria (i.e. cost centre, product and activity) does not necessarily imply a high level of relevance. Relevance measures the system's ability to provide the information managers consider as important to perform a number of tasks. Thus, it could be concluded that a cost system that analyses information according to criteria that correspond to aspects customarily encountered in cost accounting systems settings such as cost centres, products and activities does not necessarily provide the appropriate information for multipurpose decision making as reflected by the variety of managerial tasks examined in the survey.
With respect to regressions 2 and 5, the statistical analysis indicates that the accuracy of cost accounting information (ACC) as well as the system's ability to meet users' needs (NEEDS) are positively and significantly associated with the extent to which costs are analysed by cost centre, product and activity (DET_1), the degree to which variance analysis is conducted (VAR) and the extent to which information is provided upon request (FREQ_2). The abovementioned associations are in the expected direction.
The results of regressions 3 and 4 also support our hypotheses. They provide supporting evidence that the more frequent the cost information (FREQ_1 and FREQ_2) and the more variances are calculated (VAR) the greater the extent to which cost information is provided in time (TIME) and is up‐to‐date (DATE).
Regression 6 statistical results present a positive association between the degree to which variances are calculated (VAR) and the extent to which cost information has the appropriate level of analysis (APPR). Moreover, the results show that the dependent variable is positively and significantly associated with the frequency of cost information dissemination (FREQ_1 and FREQ_2).
The results of regression 7 provide supporting evidence that the greater the extent to which a cost accounting system analyses costs by cost centre, product and activity (DET_1) and permits variance calculation (VAR) the more reliable the information that it provides to users (REL).
A positive and statistically significant association between the extent to which cost information is used for decision making (USE) and frequency of information provision (FREQ_1 and FREQ_2) is revealed by the results of regression 8. The signs of the regression coefficients are in the expected direction.
Finally, it should be noted that neither the extent to which the cost accounting system allows the preparation of customized reports according to users specifications (DET_2) nor the degree to which costs are classified according to behaviour (DISAGG) are found to be significant predictors of any of the dependent variables.
The research hypotheses were tested via the following eight multivariate regression models: (Equation 1)(Equation 2)(Equation 3)(Equation 4)(Equation 5)(Equation 6)(Equation 7)(Equation 8) where DET_1 is the extent to which the system analyses costs by cost centre, product and activity; DET_2 is the extent to which the system allows the preparation of customized reports according to users' specification; DISAGG is the extent to which the system classifies costs according to behaviour; VAR is the extent to which the system calculates variances; FREQ_1 is the extent to which the system provides frequent reports on a systematic basis; FREQ_2 is the extent to which the system provides information upon request; RELEV is the extent to which cost information is relevant for decision making; ACC is the extent to which cost information is accurate; TIME is the extent to which cost information is provided in time; DATE is the extent to which cost information is up‐to‐date; NEEDS is the extent to which cost information meets users' needs; APPR is the extent to which cost information has the appropriate level of analysis; REL is the extent to which cost information is reliable; and USE is the extent to which cost information is used to make decisions.
The correlation matrix of all variables is presented in Table IV. Pearson's correlation coefficients for combinations of all variables demonstrate significant associations between cost systems structure dimensions and information quality properties in the expected direction. A similar picture is presented in relation to Spearman's correlation coefficients. The correlations among the structure features of the cost accounting systems (darker‐shaded area) are all positive and strong in terms of statistical significance, which implies that these features even though they correspond to different characteristics they share positive relationships within a cost accounting system. Additionally, the correlations among the information quality characteristics (lighter‐shaded area) are also positive and strong in statistical terms. These statistically significant correlation coefficients have values that indicate the existence of close bonds among quality characteristics.
The results of the ordinary least squares (OLS) regressions are displayed in Table V. All models are significant (Fsig=0.000) and the adjusted R2 range from 25.1 per cent to 50.3 per cent. The results of regression Equation (1) reported in Table V indicate that RELEV is positively and significantly associated with the degree to which variances are calculated (VAR) and the extent to which reports are provided on a systematic basis (FREQ_1), while, contrary to expectations, a negative and significant association was found between the extent to which costs are analysed by cost centre, product and activity (DET_1) and cost accounting information relevance (RELEV).
Based on this unexpected negative association, it can be inferred that detailed cost information may be satisfactory in terms of accuracy, reliability and meeting managers' needs as discussed below but at the same time the analysis of cost data according to general criteria (i.e. cost centre, product and activity) does not necessarily imply a high level of relevance. Relevance measures the system's ability to provide the information managers consider as important to perform a number of tasks. Thus, it could be concluded that a cost system that analyses information according to criteria that correspond to aspects customarily encountered in cost accounting systems settings such as cost centres, products and activities does not necessarily provide the appropriate information for multipurpose decision making as reflected by the variety of managerial tasks examined in the survey.
With respect to regressions 2 and 5, the statistical analysis indicates that the accuracy of cost accounting information (ACC) as well as the system's ability to meet users' needs (NEEDS) are positively and significantly associated with the extent to which costs are analysed by cost centre, product and activity (DET_1), the degree to which variance analysis is conducted (VAR) and the extent to which information is provided upon request (FREQ_2). The abovementioned associations are in the expected direction.
The results of regressions 3 and 4 also support our hypotheses. They provide supporting evidence that the more frequent the cost information (FREQ_1 and FREQ_2) and the more variances are calculated (VAR) the greater the extent to which cost information is provided in time (TIME) and is up‐to‐date (DATE).
Regression 6 statistical results present a positive association between the degree to which variances are calculated (VAR) and the extent to which cost information has the appropriate level of analysis (APPR). Moreover, the results show that the dependent variable is positively and significantly associated with the frequency of cost information dissemination (FREQ_1 and FREQ_2).
The results of regression 7 provide supporting evidence that the greater the extent to which a cost accounting system analyses costs by cost centre, product and activity (DET_1) and permits variance calculation (VAR) the more reliable the information that it provides to users (REL).
A positive and statistically significant association between the extent to which cost information is used for decision making (USE) and frequency of information provision (FREQ_1 and FREQ_2) is revealed by the results of regression 8. The signs of the regression coefficients are in the expected direction.
Finally, it should be noted that neither the extent to which the cost accounting system allows the preparation of customized reports according to users specifications (DET_2) nor the degree to which costs are classified according to behaviour (DISAGG) are found to be significant predictors of any of the dependent variables.
การแปล กรุณารอสักครู่..