مطالعه اکتشافی در حسابداری برای مدیریت کیفیت در چین

Accounting, managerial accounting in particular, has developed some financial and nonfinancial measures to satisfy the needs for improving quality management in the last two decades. Quality-oriented accounting systems centered by the cost of quality (COQ) reporting and other nonfinancial measures of quality performance have been proven to be an effective tool for quality management in North America and other industrialized countries. This paper presents the results of a survey study on the applicability of the quality-oriented accounting measures under the business environment in China. We found that Chinese business managers and accountants have generally recognized the importance of quality management to firms' survival and growth. There is a general support for applying quality-oriented accounting information to assist quality control and management. However, the traditional nonfinancial measures of quality performance are currently preferred by a majority of the respondents, while the COQ reporting has not yet received sufficient support in practice in the country.

The recent trade normalization between the People's Republic of China (PRC) and the United States paved the way for the PRC's entry into the WTO. Dismantling of PRC trade barriers pressures domestic manufacturers to improve product quality for competing in a global market place. It is estimated that at least 10% of manufactured goods had to be reworked because of quality defects (Green, 1990). Improving productivity standards is difficult for an economy that has a serious shortage of trained TQM managers and uneducated blue-collar workforce (Frazer, 1999). We designed 5 survey scales to measure accountants' and managers' TQM knowledge. For PRC firm managers, these scales can identify the current level and measure future changes of TQM knowledge. Machowski & Dale's (1998) UK study indicated low level of TQM knowledge among UK manufacturing and business managers. By designing new scales, future PRC studies may replicate and extend this UK study. Secondly, PRC quality managers may profile their employees' knowledge in these 5 quality domains (Saraph, Benson & Schroeder, 1989, p. 817). Kwok & Sharp (1998) emphasized the need for scale development in accounting research. Only factor and principal component analysis can identify a scale's construct validity, dimensionality and reliability. We can substantially improve a study's correlation and regression results by refining a construct's measurement attributes (Hunter & Schmidt, 1990, p. 94).

The absence of sound factor and component results will continue to hinder the development of new PAF and disclosure dimensions (Kwok & Sharp, 1998). Recently several promising TQM scales (Saraph, Benson & Schroeder, 1989 and Daniel & Reitsperger, 1991) have been developed, but not adequately tested and refined (see Table 1). For instance, SBS (1989) results did not disclose which EFA extraction method was used. SBS (1989) also extracted factors sequentially one at a time and failed to test for convergent and divergent validityies between 8 constructs. SBS (1989) reported extracting 9 EFA factors from 66 TQM items. Using half of these items, Ittner & Larcker's (1995) study reported 12 PCA components. Conflicting EFA and PCA results between studies call into question the correct dimensionality of SBS (1989) multi-dimensional instrument. In contrast our study's EFA and NPC results showed consistent item loading patterns and found the same number of PAF, PQSI and accounting disclosure dimensions. EFA and NPC results present strong evidence for 3 of the 5 PAF and financial and nonfinancial disclosure measures. A major advantage of each scale is parsimony, i.e. less than 7 loading items. Future studies in diverse countries and industries may refine and improve these scales' psychometric properties. Scale refinement and replication will allow testing of the same hypothesis in diverse settings using a standardised set of PAF, PQSI and disclosure measures. The use of standard scales will detect true differences in a study's empirical results that are due to between-country differences. Without standard instruments, it is not possible to determine whether conflicting empirical results between studies are a function of true between-country differences or merely a statistical artifact produced by using different instruments (Daily, Johnson & Dalton, 1999).