9. Discussion and conclusion
Prior research has provided inconsistent findings relating to factors influencing the nature of product
costing systems. Weak measures have been used for both the dependent and the independent variables.
This study has sought to overcome these weaknesses by using composite scores derived from multiple
questions, and where ABC adoption has been used as the dependent variable, evidence has been provided
that firms claiming to be ABC adopters are really likely to be ABC adopters. Instead of using only
the adoption or non-adoption of ABC systems as a dependent variable this study has used four different
measures thus enabling more robust tests of the relations among the predictor variables and the dependent
variable. Multiple and logistic regression, rather than bivariate statistical tests, were mostly used to test
the hypotheses.
Evidence was presented to support the acceptance of six of the nine hypotheses presented. The importance
of cost information and size were significant variables using all of the four dependent variables and
the extent of the intensity of competition and the financial sector were significant for three of the methods.
The use of other innovative management accounting techniques and the service sector were significant
only for distinguishing between ABC and non-ABC adopters. The use of lean production techniques and
JIT also had a positive significant influence on the adoption of ABC.30
The quality of information technology, product diversity and cost structure were not significant variables
for any the four dependent variable measures. The latter two variables, however, were significant in
distinguishing between those firms that operated and those that did not operate formal costing systems.
In terms of information technology, their reduced costs in recent years have resulted in its widespread
availability and adoption by all types of companies. Thus, the quality of information technology may no
longer a barrier to implementing more sophisticated costing systems. The lack of significance of product
diversity and cost structure for any of the four dependent variable measures is surprising given that both
of these variables are often presented in the literature as being the dominant motives for implementing
sophisticated costing systems. It is possible that the questionnaire used too simplistic measures that failed
to take into account the precise ways that they influence the level of cost system sophistication.
Besides attempting to improve the methods of measuring the variables future research should consider
incorporating other important variables that have been omitted from this and other studies but which are
likely to influence cost system design. The most notable omitted variables are organizational variables
such as top management support, resistance to change by preparers and users of accounting information,