會計情報의 負荷가 意思決定의 質에 미치는 影響-企業倒産의 경우를 중심으로-
The Effect of Accounting Information Load on Decision Quality in a Bankruptcy Decision Task
- 인하대학교 산업경제연구소
- 경상논집
- 연구논문집 제7집
-
1993.09123 - 157 (35 pages)
- 25
Schroder et al.(1967) presented a model describing human information utilization patterns when the available information exceeds the processing capabilities of decision makers. The capability of a decision maker to integrate information inputs into decisions is limited and different between decision makers. The amount of integrated infomation by a decision maker is to increase initially up to a certain amount of information provided, but it is to decrease thereafter when the decision maker experienced information overload by the increase in the information provided. Therefore, the amount of integrated information is to follow a bell-shaped curve. Others such as Revsine(1970), Miller(1972) and Miller and Gordon(1975) discussed the implications of this model for examining the effects of accounting information load on decision maker's behavior. This model was examined by several experimental studies such as Casey(1980), Snowball(1980), Shields(1980), Iselin(1988) and Chewning, Jr. and Harrell(1990). Most of these studies suffered from ineffective load manipulaitons and did not examine both decision inputs(number of information cues) and decision outputs. They also did not consider the effects of professional experience on the integration of information. This study is to examine the effects of information load on Korean professionals' cue utilization and decision quality in a bankruptcy prediction task. The specific purpose of this study is to examine 1) the difference in the effects of information load on cue utilization between Korean CPAs and securities analysts and 2) the effects of cue utilization on the decision quality. In order to select relevant financial ratios used as decision cues, 17 defferent variables were found to possess predictive ability from previous researches on bankruptcy predictions. The industries in which 19 listed companies were bankrupt during the period from 1990 to June 1992 were identified and the financial ratios for the period from 1989 to 1991 were collected from the listed companies of these industries. These industries included from 446 to 472 listed companies in each of the three years. Factor analysis was used to reduce the redundancy contained in the financial ratios. As previous researches on human information processing such as Ashton(1974), McGhee et al.(1978), and Danos and Imhoff(1982) indicated that average cues utilized by decision makers was in the range from five to eight cues, eight factors were identified and for each factor the financial ratio with the highest loading was assigned. Pearson product-moment correlations for each pair of the ratios selected were relatively weak, thereby indicating that the ratios used in bankruptcy prediction were relatively uncorrelated and were reasonably representative of the decision environment. This study employed four, six and eight cues to represent information load levels: four cues inluded the ratios from factor 5 to factor 8 and six cues included the ratios from factor 3 to factor 8. Each cue was manipulated at two levels: top one-third and bottom one-third of the firm's industry. Two professional groups participated in this study: 24 CPAs and 28 securities analysts. Each participant was requested to provide the bankruptcy likelihood in the following three years for each case in the package which contained 16 four cue cases, 32 six cue cases and 32 eight cue cases. In order to identify the relationship between decisions and the cues provided, a regression model for each cue sets of cases for each subject was computed using bankruptcy predictions as the dependent variable and financial ratios as independent variables. The number of cues utilized at each load level was detemined by counting the number of ratios with statistically significant bata coefficients. The number of cue utilization was significantly different at six cue level between groups of CPAs and securities analysts. Average number of utitilized cues increased in th
Ⅰ. 문제의 제기
Ⅱ. 기존연구의 검토
Ⅲ. 자료의 수집과 분석방법
Ⅳ. 수집자료의 분석
Ⅴ. 결론
참고문헌
별첨
Abstract
(0)
(0)