where CPA firms may (or may not) want to increase accountability. The low-complexity task involves listing compliance atid substantive tests in an internal control evaluatioti. There is only one relevatit information cue to be processed {control deviation for liabilities) and there is no need to integrate multiple cues. The number of required informationprocessing acts (retrieval of the relevant tests from memory) is limited because only a small and finite set of alternatives (compliance/substantive procedures) needs to be considered.Coordinative complexity is also low because there is little need to coorditiate the sequence or timing of acts or to perform backward/forward reasoning to retrieve the required output (unlike the ratio analysis task, descrihed later).
For this task, we expect the association between the required output (compliance and suhstantive procedures) and the input (test for liabilities) to be fairly strong in the auditors' memories, given that they have extended exposure to compliance/substantive procedures (Abdolmohammadi [1991] and Bonner and Pennington [1991]). Generation of the required output .should therefore not require significant effort or capacity usage in memory. Furthermore, the number of cognitive acts required to perform the task is low. As a result, we anticipate that accountability is unlikely to affect the performance of this task; our prediction is stated formally in HI:
HI: For the low-complexity task, accountability will not improve performance.
Our second task, of medium complexity, embodies a knowledge-accountability interaction where accountability will improve performance for auditors with high knowledge but not for those with low knowledge.The task is part of an internal control evaluation which requires the listing of financial statement errors that can occur because of an internal control deviation. In terms of component complexity, the number of separate acts required by this task (retrieval of both the related financial statement errors and the double entries from memory, plus the determination of whether the accounts are overstated or understated) is higher than that required for the low-complexity task (retrieval from memory of compliance and substantive procedures).
Component complexity is also higher because the number of alternatives to be considered (potential errors in a wide variety of accounts) is higher than the number for the low-complexity task (a more restricted set of compliance/substantive tests). Information processing for this task is therefore likely to be more detailed and complex than for the low-complexity task. Like the lowcomplexity task, but unlike the high-complexity task (discussed later),coordinative complexity is low and the task does not require much forward/ backward reasoning; hence problem-solving ability is not expected to influence performance (Bonner and Lewis [1990] and Libby and Tan [1994]).
journal of Accounting Research
Vol. 37 No. I Spring 1999
Printed in U.S.A.
By : HUN-TONG TAN AND ALISON KAO*
Nama : FADILLAH KR SUKAYAT (24209983)
Kelas : 3EB09
