An examination of contextual factors and individual characteristics affecting technology implementation decisions in auditing

International Journal of Accounting Information Systems 9 (2008) 104–121

Contents lists available at ScienceDirect

International Journal of Accounting Information Systems

An examination of contextual factors and individual characteristics affecting technology implementation decisions in auditing ☆ Mary B. Curtis a,⁎, Elizabeth A. Payne b,1 a Department of Accounting, College of Business Administration, University of North Texas, P.O. Box 305219, Denton, TX 76205, United States b School of Accountancy, College of Business, University of Louisville, Louisville, KY 40292, United States

a r t i c l e

i n f o

Article history: Received 15 June 2007 Received in revised form 15 October 2007 Accepted 31 October 2007 Keywords: Audit technology Budgeting Risk UTAUT

a b s t r a c t While computer-assisted audit techniques (CAATs) have the potential to increase efficiency and effectiveness of audit engagements, research in this area suggests that such techniques are under-utilized in public accounting. We propose that this condition is due to performance evaluation pressure and the use of budgets for multiple purposes, which result in the misalignment of firm and individual employee goals. We apply technology acceptance and budgeting theories to test this contention as well as potential organizational strategies for reducing the impediments to technology acceptance in the audit profession. Results from an experiment with experienced auditors suggest that firms have the ability to influence the implementation of new technology by using longer-term budget and evaluation periods and by communicating the approval of remote superiors regarding the software. In the absence of such firm interventions, the individual characteristics of the auditor (risk-aversion and perceptions of budgetary pressure) determine implementation decisions. Specifically, risk-seeking individuals are more likely to implement technology regardless of budget pressure perception, but for riskaverse individuals the decision to implement is positively related to perceived budget pressure. © 2008 Elsevier Inc. All rights reserved.

☆ The authors would like to thank Dan Stone, Joe Brazel, workshop participants at the University of Kentucky, participants at the 2006 AAA annual meeting, participants at the 2006 Mid-South Doctoral Consortium, participants at the University of Waterloo CISA 5th Symposium on Information Systems Assurance, and anonymous reviewers for their helpful comments and suggestions. ⁎ Corresponding author. Tel.: +1 940 565 4366; fax: +1 940 565 3803. E-mail addresses: [email protected] (M.B. Curtis), [email protected] (E.A. Payne). 1 Tel.: +1 502 852 5397; fax: +1 502 852 6072. 1467-0895/$ – see front matter © 2008 Elsevier Inc. All rights reserved. doi:10.1016/j.accinf.2007.10.002

M.B. Curtis, E.A. Payne / International Journal of Accounting Information Systems 9 (2008) 104–121

105

1. Introduction Today's audit environment is one of increased responsibility and workload for audit teams, including enhanced responsibilities for detecting fraud required by SAS No. 99 and internal control attestation now required under Section 404 of the Sarbanes–Oxley Act. One approach to meeting these increased demands is through the use of audit technologies,2 which can greatly improve the efficiency (cost) and effectiveness (quality) of an audit (Banker et al., 2002). Indeed, Elliott (2002, 139) states “Present and future users of accounting and auditing services have increasing need for relevant, reliable, and timely information and IT provides the means to meet them”. For example, computer-aided audit techniques (CAATs) may automate previously-manual audit tests, resulting in reduced audit hours for the task, and the ability to easily test 100% of the population rather than a sample, greatly increasing the reliability of conclusions based on that test (AICPA, 2001). While the current audit environment intensifies the need for firms to employ techniques that can reduce workload, including those affecting technology implementation decisions, the culture of public accounting may create impediments to the adoption of new technologies by audit teams (Vendrzyk and Bagranoff, 2003). Therefore, it is important to understand the factors affecting an auditor's reluctance to recommend technology implementation that will impact both the business, and the practice, of auditing. In the current study, we examine factors affecting an in-charge auditor's decision to implement new technology on an engagement, including two contextual or firm-level factors (length of the engagement budget period and remote superior influence) and two individual characteristics (risk preference and perceived budget pressure). Technology acceptance has received much attention from information systems (IS) researchers and the technology acceptance models employed in the IS literature provide a starting point for exploring the issue in public accounting. Specifically, the Unified Theory of Acceptance and Use of Technology (UTAUT), developed and tested by Venkatesh et al. (2003) in the MIS field, informs our study of technology acceptance in the auditing field. As most technology acceptance models (TAM) are designed for multiorganizational use, we modify the UTAUT to reflect factors not previously examined in prior studies but that are present and perhaps (but not always) unique to public accounting. For example, although user resistance to new technology is common in all areas of business, the practices and pressures in public accounting, particularly evaluative pressures and the use of budgets for multiple purposes, create the possibility of even greater technology resistance in public accounting than in other business environments. Additionally, individual differences not previously considered in IS research, such as risk preference and reactions to evaluative pressure, may impact technology acceptance in the public accounting environment. The UTAUT model's performance expectancy dimension motivates the consideration of the length of the budget/evaluation period in our search for possible intervention mechanisms. Additionally, based on the model's social influence construct, we consider whether making the attitude of remote superiors (practice office managing partners) known to the auditor will impact their acceptance decision. Finally, in place of the model's typical individual differences (age, gender, etc.), we consider audit experience, risk preference and perceptions of budget pressure because of their potential importance in auditing. To test this modified model of technology acceptance, experienced auditors completed a case study in which they were given the opportunity to employ a new audit technology after the client engagement letter was signed and the budget was approved. Our statistical analyses show that both of the audit environment factors considered in the study have a significant effect on auditors' decisions to use new technology. The longer budget period (a combined three-year budget) and the communication of the attitude of a remote superior lead to increased likelihood to implement. However, in the absence of these external factors, individual differences play a role in technology implementation. Risk-seeking auditors are more likely to implement than risk-averse auditors. Unexpectedly, however, risk-averse auditors are more likely to implement when they perceive a great deal of budget pressure.

2 Audit technology is defined here as any tool or technique, manual or computerized, that can automate an aspect of the audit. Examples include check lists, expert systems, computer-aided audit techniques or automated working papers.

106

M.B. Curtis, E.A. Payne / International Journal of Accounting Information Systems 9 (2008) 104–121

2. Background and theoretical development 2.1. Technology characteristics and culture of public accounting In audit situations where use of technology is optional, the implementation decision is typically made by joint discussion between the audit manager and the in-charge auditor (Houston,1999). If this decision is made prior to the planning process and there is flexibility in total audit revenue, the auditor may be able to factor the cost of technology implementation into the budget with the permission of the partner and client. However, in situations where there is little budgetary flexibility or the implementation decision is made after the planning phase is complete, the audit team must accommodate the additional costs within the existing budgetary limits. Thus, technology implementation may potentially have a major budgetary impact on audit engagements. When viewed from a one-client, one-engagement-year perspective, technologies typically cost more the first time they are implemented than they save in that first period and also are likely to cost more than the overall improvement in audit quality gained in the first period. For example, the audit software itself is expensive and engagements are typically charged a fee for its use. Additionally, auditors who employ new tools and techniques in the field will need training and tend to require more hours the first year to accomplish the task due to learning curves, further increasing the cost of implementation. Because implementation costs can be substantial, net gains in efficiency typically occur in subsequent periods. Thus, the implementation of technology is associated with long-term paybacks (Lovata and Linda, 1988). In addition to delayed payback, there are other risks associated with implementing new technology, including excessive costs due to difficulties in implementation and training, lack of technical support when needed, and failure to meet desired improvements in efficiency and effectiveness (Fischer, 1996). Additionally, in hierarchical organizations such as public accounting, the need to meet the expectations of superiors creates a degree of uncertainty. For example, a decline in productivity due to implementation demands may be viewed negatively, just as might a failure to implement software that the superior supports. These factors can result in user resistance to technology (Nelson and Cheney, 1987), which is a widespread problem across professional fields (Venkatesh, 1999). Finally, individual differences in risk preference and perceived pressure to meet budget, to increase audit quality, and to use new technology may also play a part in such decisions. Because achieving budget is so important, yet tenuous, those who are risk-averse by nature or perceive great pressure to meet budget may be less inclined to partake of any activity that may increase the uncertainty in their environment. One primary difference between the typical technology acceptance research context and the traditional audit context is that, in MIS research studies, the software already has been installed successfully and the question is whether people will use it — thus, the primary organizational cost has been incurred, with only personal cost (training, lost work time) remaining. In an audit context, the audit engagement often must absorb the cost and risk of installation. Thus, in an audit, the question is whether to install and use, while in the MIS context, the decision is whether to use. Additionally, the implication of repeat engagements in an audit is completely outside of the MIS context. Moreover, the legal implications of poor performance on an audit (e.g. Lowe et al., 2002) are more significant than any MIS research has considered to date. Finally, when audit software is optional (as much of the substantive testing software is) rather than required by the firm, one's decision to implement the technology can lead to second-guessing by superiors and increased evaluative pressure. Add to this the pressure of the up-or-out personnel promotion policies employed by most large public accounting firms, based heavily on budget attainment (Brazel et al., 2004), and we see an audit environment much more heavily impacted by individuals' risk preferences and perceptions of pressure than any considered by existing MIS research. 2.2. Acceptance and use of audit technology Auditing technology studies have primarily examined how the use of technology (typically decision aids) affects cognitive processing and the resulting decisions auditors make.3 The focus of the current study

3 For example, recent studies examine the source reliability of decision aids for explanations of unusual fluctuations (Anderson et al., 2003), the orientation (negative or positive focus) of the decision aid on identifying risk factors (Bedard and Graham, 2002), and how decision aids affect learning (Rose and Wolfe, 2000).

M.B. Curtis, E.A. Payne / International Journal of Accounting Information Systems 9 (2008) 104–121

107

is somewhat different. We are interested in identifying factors that affect an auditor's decision4 of whether to implement new technology on an audit engagement. In the only prior auditing study in this area, Bedard et al. (2003) employed the Technology Acceptance Model to examine the effects of training on user acceptance of electronic work papers. Their results show that training enhances auditors' technology acceptance through its effect on users' perceptions of their task and computer self-efficacy. Similar to Bedard et al. (2003), we employ the technology acceptance literature as a theoretical foundation to examine technology implementation (CAATs) in an auditing setting. However, we explore quite different factors in our search for both the cause of, and solution to, technology reluctance. In contrast to the auditing field, technology acceptance is one of the most widely researched topics in the information systems area. Although there are several competing models that predict user acceptance of information technology, we will rely on one of the most recent models as a theoretical starting point. The Unified Theory of Acceptance and Use of Technology (UTAUT — Venkatesh et al., 2003)5 incorporates three predictors of intention to use technology — performance expectancy, effort expectancy and social influence. The predictors are defined as follows (Venkatesh et al., 2003, 447–453): Performance expectancy — “…the degree to which an individual believes that using the system will help him or her to attain gains in job performance.” Effort expectancy — “…the degree of ease associated with use of the system.” Social influence — “…the degree to which an individual perceives that important others believe he or she should use the new system.” In the following sections we elaborate on both performance expectancy and social influence as predictors of technology acceptance in an audit context. These two predictors are examined because of the extent to which they may differ in an audit setting vs. other organizational settings. 2.3. Performance expectancy The performance expectancy variable in the UTAUT predicts a positive relationship between intention to use technology and gains in job performance. Measuring an auditor's performance is a complex process that involves many factors including budget attainment and audit quality (Hunt, 1995). Audit technology has the potential to improve the efficiency and effectiveness of audit tasks in the long run. However, the budgetary impact of new technology can be substantial in the first period such that ‘profits’ from the investment will not be seen until future periods. Therefore, the auditor's current-period budget attainment could be significantly impacted by new technology implementation, while the benefits from technology use may only appear in future budgetary periods — and possibly to auditors other than those who will be impacted by the current-year budgetary short-fall. Budgets typically serve multiple purposes in organizations, including long-term planning, short-term cost control, project monitoring and periodic performance evaluation (Hopwood, 1972; Shields and Shields, 1998). Such is also the case in auditing (Pierce and Sweeney, 2004) where budgets communicate the importance and planned extent of testing for each audit component, help to monitor progress of the audit, and serve to control costs by making cost overruns salient to the observer. Most importantly to this study, budgets are used to evaluate the performance of each member of the engagement team, with those at each hierarchical level responsible for budget attainment of those below them (Shapeero et al., 2003). Prior research supports the notion that auditors engage in dysfunctional behavior due to budget-based performance evaluation (Pierce and Sweeney, 2004). These dysfunctional effects can include myopic decision making (Hansen et al., 2003; Levinthal and March, 1993) and barriers to change (Neely et al., 2001), creating conflict between organizational and individual objectives. For example, Merchant's (1990) study of profit center managers demonstrated that the use of budgets discourages long-term investments 4 Our study focuses on individual decisions to implement technology. While many technology decisions are mandated at the firm level, others are left to the discretion of the engagement team depending upon such factors as the type of client, size of client, etc. For example, the use of computer-aided audit techniques for substantive testing is oftentimes optional. 5 Venkatesh et al. (2003) reviewed the eight most prominent models/theories that predict behavioral intentions and/or usage, developed a unified model that incorporates elements of the previous eight models, and empirically validated the new model. The UTAUT explains up to 70% of variance in intention to use technology, outperforming previous models, such as TAM.

108

M.B. Curtis, E.A. Payne / International Journal of Accounting Information Systems 9 (2008) 104–121

when short-term payoffs are not expected. Both economic theory (agency theory) and psychology theories (impression management and self-presentation) predict dysfunctional behaviors associated with budgetbased performance evaluations since individuals typically act in their own self-interest and seek to present themselves in the best light in the short run, even if such behavior results in less than optimal performance for the firm in the long run. This impact may actually be greater in public accounting where auditors receive annual evaluations as well as interim engagement evaluations from multiple sources and budgetary attainment plays such a significant role in those results (Hunt, 1995). There is always the possibility that auditors will not display the same myopia in their technology investment decisions due to the unique professional responsibilities of auditors compared to a typical organizational manager. In fact, the “Principles of the Code of Professional Conduct of the American Institute of Certified Public Accountants express the profession's recognition of its responsibilities to the public, to clients, and to colleagues. They guide members in the performance of their professional responsibilities and express the basic tenets of ethical and professional conduct. The Principles call for an unswerving commitment to honorable behavior, even at the sacrifice of personal advantage.” (AICPA, 2006, ET 51.02) Since audit technology has many potential long-term benefits to the firm, client, investors and others, it is possible, particularly in a post Sarbanes–Oxley audit environment, that an auditor's enhanced feelings of public responsibility may result in making a technology investment for the greater good even at the cost of personal sacrifice in the current period via budget overruns and lower performance evaluations. If this is true, the budget/performance evaluation period will have no effect on the technology implementation decision. However, the extant research supports the need to align organizational and individual goals, even in public accounting. Due to the likelihood of dysfunctional effects associated with multi-purpose budgets, many managers advocate complete separation of planning and budget-based performance evaluation within an organization, so that planners can be free to adapt to changing circumstances without fear of performance evaluation consequences (Hope and Fraser, 2003). Alternately, dysfunctional effects of using the same budget for both planning and performance evaluation may be mitigated by innovations in the way budgets are designed, such as employing longer budgetary time periods (Fisher et al., 2003). This allows for evaluations of long-term investment decisions that ignore interim information, concentrate on the overall outcome, and reduce the extent to which evaluations divert decision makers from a firm's overall profit maximization or quality improvement goals (Kite et al., 1996–97). Applying these notions to the audit context, a multi-year budget6 would allow the cost of technology to be spread across several years, reducing the impact on the first year. To the extent that the audit technology decision reflects a self-serving interest shown in prior studies of other decision makers (e.g. Merchant, 1990), we expect a longer budget/performance evaluation period to result in greater intention to use technology than the usual one-year engagement/evaluation period. This prediction is further supported by prior research showing that dysfunctional behavior often results from audit budget pressure (Pierce and Sweeney, 2004). H1. For technology investments with long-term payoffs, auditors with shorter-term (longer-term) budget and evaluation periods will be less (more) likely to implement audit technology. 2.4. Social influence The social influence predictor in the UTAUT includes consideration of the person's perception of the opinion of others, his or her reference group's subjective culture and specific interpersonal agreements with others, as well as the degree to which use of an innovation is perceived to enhance one's image or status in one's social system (Venkatesh et al., 2003, 452). This construct suggests that an auditor would be sensitive to the opinions of others not on their audit team, resulting in decisions consistent with the social norms around them. The notion of social influence is well-grounded in psychology literature, where findings of accountability studies have shown that individual decisions are often influenced by the known views of

6 Evidence suggests that although firms do not currently use long-term budgets with their clients, they do many times have longterm commitments from their clients, thereby making the use of long-term budgets feasible. A survey of 15 managers of the firm indicates that approximately 40% of firm clients commit to multi-year engagements lasting an average of 3 years.

M.B. Curtis, E.A. Payne / International Journal of Accounting Information Systems 9 (2008) 104–121

109

superiors (Tetlock, 1992; Lerner and Tetlock, 1999). In highly hierarchical settings, such as public accounting, it would be expected that the attitudes of those who have evaluative authority over an individual would be taken into consideration in that individual's decision making. Recent accounting research has confirmed the role of social influence in occupational conduct and satisfaction (Lord and DeZoort, 2001; DeZoort and Lord, 1997; Wilks, 2002; Turner, 2001). For example, Tan et al. (1997) found that the decisions of auditors who were aware of their superiors' preferences were significantly different from those who were not aware of such preferences. In the software use context, Loraas and Wolfe (2005) found that staff auditors were more likely to voluntarily use commonly available software (such as Excel) rather than manual approaches to an audit task, when their audit supervisor supported the decision. Since prior research has addressed the influence of superiors on the audit team's technology acceptance, this study moves away from this direct influence to explore the impact of more remote individuals on the acceptance decision. How far removed from the source of social pressure can the subject be and retain influence? The answer is highly dependent upon context and culture. In an audit firm, it is doubtful that the managing partner of an office would be informed of such minor details as an audit in-charge's technology implementation recommendation. It is also not expected that this individual would be involved in the in-charge's performance reviews. Therefore, the impact of the office managing partner's influence can be asserted to be independent of any direct influence through evaluation or supervision. However, due to the culture of public accounting, increasing the use of technology on audits without imposing the requirement of their use7 may be as easy as communicating the remote superior's interest. While this may seem simplistic, with the strict hierarchical structure and the up-or-out personnel management policies of public accounting firms, auditors likely react to any information regarding the performance preferences of those in authority. The critical determinant is whether the auditor would accede to this pressure, despite self-interest derived from the highly budget-oriented environment of public accounting. To fully test the impact of social influence on such decisions, separate from direct supervisory pressure, this study evaluates the impact of office managing partner attitudes on the acceptance decision. H2. Auditors are more likely to implement technology when a remote superior favors implementation than when they have no knowledge of the superior's preference. The current study also offers the unique opportunity to further explore the construct of social influence in this specific audit setting by considering this construct along with the budget period. The social influence variable of the UTAUT suggests that an auditor's intention to use technology is based purely on affect (e.g. enhancing one's image) without regard to the other predictors, such as performance. In other words, the model predicts a main effect for social influence. The interesting question is whether the partner's influence is purely social. It may be that the partner's influence serves to reduce the decision uncertainty of the auditor due to the risks of implementation, blowing the budget, etc. The auditor may feel more comfortable making the acceptance decision when a remote superior's encouragement can be used as an excuse for taking the risk. If this is the case, we expect an interaction between the two predictors such that in a 1-year budget/evaluation period intention to use technology is higher when the remote superior recommends (thereby removing the risk), but there should be no difference in intention in the 3-year budget/evaluation period based on the superior's recommendation because the longer budget period has already removed the risk. Based on the UTAUT we predict a main effect for influence and will explore the potential for an interaction in additional analyses. 2.5. Individual differences The UTAUT recognizes several individual difference moderators, such as experience and gender. In an audit context, there are likely numerous potential individual characteristics that may act alone or in combination with other characteristics and events (DeZoort and Lord, 1997; Sweeney and Pierce, 2004) to

7 While planning software may be generally applied to all clients (to provide for risk assessment, for example), not all substantivetesting software is appropriate for all situations. Therefore, the most effective and efficient policy should encourage the use of software where appropriate and avoidance of software where its use would be ineffective or inefficient.

110

M.B. Curtis, E.A. Payne / International Journal of Accounting Information Systems 9 (2008) 104–121

affect the decision to implement technology. In particular, the unique characteristics of the public accounting culture create a potential source of pressure and stress for the decision maker, and possible individual difference considerations not present in other contexts. Technology implementation carries numerous risks including unexpected, additional costs due to difficulties in implementation and training, lack of hands-on technical support in the field, and failure to meet desired improvements in efficiency and effectiveness. Research suggests that decision making is affected by how individuals react to risk (Wang, 1996; Zickar and Highhouse, 1998; March and Shapira, 1987). Individuals with high risk propensity tend to be more comfortable dealing with situations of risk and in fact perceive the objectively same situation as less risky than do others (Sitkin and Weingart, 1995). They are therefore likely to experience less “debilitating anxiety about an entrepreneurial career, perceive a greater sense of control over outcomes, judge the likelihood of receiving positive rewards more highly, and thus possess higher self-efficacy” (Zhao, Seibert and Hills 2005, 1267). Although many economic and decision theoretic models overlook risk-aversion as an individual difference characteristic, a wide range of psychological studies has shown that people differ in how they perceive risk and make choices under uncertainty (c.f. Trimpop et al., 1999), and the applied sciences widely recognize that risk influences decision making. For example, Rose et al. (2004) found that personal risk preference impacts the evaluation of information technology investment decisions, particularly in ambiguous contexts. In evaluating the adoption decisions of others, risk-seeking evaluators gave higher evaluations to risky IT investment decisions than risk-avoiding evaluators. The marketing literature proposes that consumer risk has a significant influence on product purchase decisions (Chen and He, 2003). The notion of consumer risk seems appropriate to the adoption of audit technology since the auditor who chooses to adopt an audit technology is essentially purchasing that technology for his or her audit and therefore may perceive many of the same consumer risks. The audit environment itself creates a unique context for risk to influence judgment. Many issues with regard to audit budget performance are fairly unstructured and uncontrollable. For example, “making budget” for an auditor depends on the budgetary performance of all staff auditors, client responsiveness to time commitments, as well as the amount of slack present in the budget (Pierce and Sweeney, 2004). Because the auditor's performance evaluation will be at least partially based on the budgetary performance of his or her audits, it seems reasonable that those who are risk-averse by nature (i.e. general risk-aversion) may be less inclined to partake of any activity which may increase the uncertainty in their environment. The effect of individual differences in risk preference on the decision to implement new technology may be a strong enough force such that the effect is present regardless of the external factors such as budget period and superior influence. In this case, we would expect no interactions between risk preference and the external variables. On the other hand, the external factors may be the stronger force. A multi-year budget period may relieve a significant amount of the risk so that individual differences in risk would not matter as much as they would in a single-year budget period. Additionally, influence from the superior may be so strong a force that individual differences in risk only come into play when there is no influence from the superior. In these situations, we would expect to find interactions between risk preference and the external variables. While we expect risk preference to play a role in the decision to implement technology, it is not clear whether risk preference will act alone or moderate the effects of the external factors. We will, therefore, examine both the main effect and the interactive effects of this variable. H3a. Auditor risk preference is positively associated with the decision to implement technology. H3b. Auditor risk preference is positively associated with the decision to implement technology, only in the absence of contextual factors. Auditors face significant pressure during the conduct of an audit engagement, including pressure to achieve budget, to improve audit quality and to use technology. DeZoort and Lord (1997, 31) define pressure as “an objective stimulus construct referring to individual characteristics or combinations of characteristics and events that impinge on the perceptual and cognitive processes of individuals”. Prior research has established the significance of time pressure in explaining dysfunctional behavior among auditors (Kelley and Margheim, 1990; Otley and Pierce, 1996). Just as there are individual differences in auditors' ability to meet budget expectations (Otley and Pierce, 1996), organizational behavior researchers and psychologists recognize that there are individual differences in perceptions of pressure (French and Caplan, 1972; Lazarus,

M.B. Curtis, E.A. Payne / International Journal of Accounting Information Systems 9 (2008) 104–121

111

1995). Indeed, in their study of cost–quality conflicts in audit firms, Pierce and Sweeney (2004) found that only a fraction (36%) of their audit junior and senior respondents perceives significant budgetary pressure and that these individuals were more likely to engage in dysfunctional behavior in their audit engagements. In regard to technology acceptance, perceptions of budget pressure may induce individuals to forgo technology implementation because of concerns that the implementation would place greater strains on their current-year budget. Since individual auditors are not guaranteed assignment to the same engagements in following years, there is no notion that current-year sacrifices will yield subsequent gains for the individual making the acceptance decision. Greater perceptions of pressure may overarch all other factors, discouraging implementation under any circumstances, or it may be moderated or relieved with a longer-term budget, or when a greater pressure such as influence from a superior is present. Because pressure effects may be due to individual characteristics alone or in combination with other characteristics and events (DeZoort and Lord, 1997), we will examine both the main and interactive effects of these pressure variables. H4a. Auditors who perceive greater (lesser) levels of budget pressure are less (more) likely to implement technology. H4b. Auditor who perceive greater (lesser) levels of budget pressure are less (more) likely to implement technology, only in the absence of contextual factors. 3. Research method 3.1. Participants In-charge auditors from one Big 4 accounting firm participated in the study. Because in-charge auditors run the audit day-to-day, they are most likely to identify instances where short-term sacrifices could result in long-term gains of efficiency or effectiveness. Indeed, Sweeney and Pierce (2004) state that the greatest conflict between the business of auditing and the profession of auditing is manifested at the audit in-charge level where quality pressure conflicts with time pressure. In audit situations where use of technology is optional, the implementation decision is typically made by joint discussion between the audit manager and the in-charge auditor (Houston, 1999).8 Therefore, the participants in our study have the appropriate level of experience and knowledge to make such a decision. We further have no reason to expect the results to be any different for more experienced participants (Peecher and Solomon, 2001). From a total of 181 participants, twenty-nine instruments were discarded due to lack of completion, and 13 were discarded for incorrect responses to manipulation check questions, resulting in a sample of 139. On average, the participants were 26.5 years in age, had 3.0 years of public accounting experience, and had used 1.4 electronic audit tools. The sample is 51% male. No significant differences in demographic variables between treatments were found, indicating that random assignment of treatments was effective. 3.2. Materials and procedures The experiment was administered during a firm training session in the presence of one of the researchers. Participants were randomly assigned one of four versions of the research instrument,9 which contained a case study and questionnaire. The case began by describing the budget period (one or three years) and specific audit fees and expenses for a hypothetical engagement. Following this, the case described an opportunity to implement a new technology (including estimated hours) after the budget had been established. The new technology demands a modest outlay of additional hours in the first year, but will potentially provide time

8 This was confirmed through discussions with managers and partners in the firm. Additionally, in a survey at the in-charge training class where this data was collected, 15 managers report varying opinions regarding the involvement of in-charges in the technology acceptance decision. On average, these mangers report that the in-charge would make the decision to implement 13% of the time (range of 0–50%), while the manager and partner would make the decision 56% (range 0–100) and 31% (range 0–100), respectively. In most cases, the in-charge would have input into the decision. 9 The case study was initially pilot tested with a separate group of 35 auditors during an in-house training session. Based on the results obtained, minor changes were made to the instrument primarily to improve comprehension of questions.

112

M.B. Curtis, E.A. Payne / International Journal of Accounting Information Systems 9 (2008) 104–121

savings of the same amount in future years. A modest time savings was used in the study so as not to induce a demand effect. Finally, information was provided as to the office managing partner's opinion of the software (unknown or encouraged use). Following the narrative portion of the case, three questions were posed to measure the participant's intention to use the new technology described in the case. After the case study, questions were presented to measure the effectiveness of manipulations; risk preference; individual perceptions of budget pressure, technology pressure, quality pressure; and several demographic variables. 3.3. Design and measurement of variables The research design is a 2 × 2 between-participants design. The budget period was described as either a (1) one-year engagement-by-engagement basis (with a three-year commitment from the client) or a (2) three-year basis. The total revenues and costs for the client were the same under either basis, and were such that the new technology would increase costs the first year, but result in cost savings in years 2 and 3. Influence from superiors was manipulated by one of the following statements: (1) “There is no information regarding whether the engagement partner or the practice office managing partner supports use of the software” or (2) “The engagement partner told you that the practice office managing partner is not going to require use of the new software, but is encouraging implementation.” While the goal of this study is to understand technology usage, the role of intention to use as a predictor of actual usage is well-established in information systems and related research (e.g. Venkatesh et al., 2003; Davis, 1989). Our dependent variable assesses intention to use audit technology, relying on previously established linkages suggesting a strong likelihood that behavior will follow intention (Davis, 1989). We measure this construct with three items adapted from Davis (1989) which have been tested extensively in this literature, including Venkatesh et al. (2003). Each of the following items was measured on a 10-point scale ranging from very unlikely to very likely: (1) I will recommend implementing the software on this year's engagement. (2) If I made the acceptance decision, I would use the software for this engagement. (3) If placed in a situation very similar to this audit scenario, I plan to recommend the implementation of the software provided. To measure individual differences in risk preference, we modified Zaleskiewicz's (2001) instrumental risk preference scale to the audit context. Our scale employs 4 of his 7 scale items with minor wording modification and then added 3 new questions, in order to reflect risk preference specific to the audit setting (preference for budgetary certainty and for first-time audits). Two of the items were reverse coded (see Appendix B10). Responses for each question were obtained using a 7-item scale (1 = disagree completely and 7 = agree completely). To measure individual perceptions of budget pressure, participants were asked to respond on the same 7-item scale: On past audit engagements, I have felt significant pressure to control budget hours (1 = not at all and 7 = always). There are certainly other factors that may affect the technology implementation decision, for example, the ability to use the technology on other engagements and the ability to employ the same audit staff in the future years. Information regarding these issues was provided in the case to control for differing participant assumptions. Thus, the case includes the following two sentences: It is not clear whether this software will be used on other clients in your office. It is probable that the current engagement team will be reassigned to this engagement next year. 4. Results 4.1. Effectiveness of manipulations Two questions were included in the exit questionnaire to assess the effectiveness of the independent variable manipulations. Participants were provided with two responses to each question; each response

10

Zaleskiewicz (2001) reports reliabilities (Cronbach's alpha) of .67 for his 7-item scale for two different samples.

M.B. Curtis, E.A. Payne / International Journal of Accounting Information Systems 9 (2008) 104–121

113

Table 1 Means (standard deviations) in total and by group Total (n = 139) Intention Budget pressure Risk preference

21.7 (6.1) 5.2 (1.5) 19.4 (3.6)

Budget period

Influence

1-year (n = 65)

3-year (n = 74)

p-value a

None (n = 71)

Yes (n = 68)

p-value a

20.2 (6.4) 5.1 (1.5) 19.0 (3.8)

23.0 (5.6) 5.3 (1.5) 19.8 (3.4)

.006 .592 .196

21.1 (6.5) 5.2 (1.5) 19.6 (3.5)

22.3 (5.8) 5.1 (1.5) 19.2 (3.8)

.276 .671 .549

Intention = Sum of 3 items measuring intention to use new technology on the engagement (range 3–30). Budget pressure = Response to “On past audit engagements, I have felt significant pressure to control budget hours” (range 1–7). Risk preference = Sum of 4 items measuring general risk preference (range 4–28). a p-values are 2-tailed.

was representative of one level of the variable. For example, one question stated “The budget period in the audit scenario was…”, and the possible answers were 1 year or 3 years. Responses not correctly matching the intended level of the independent variable were considered incorrect. Correct responses were 145/152 or 95.4% for the budget period and 146/152 or 96.0% for the partner influence, indicating participants understood the manipulations. 4.2. Reliability Cronbach's alpha reliability for the 3 questions designed to assess the dependent variable ‘intention to use the technology’ is .94. Since the initial Cronbach's alpha reliability measure for the risk preference questions was very low (.32), a factor analysis was performed. After removing question #3 due to low communality, the factor matrix resulting from Varimax rotation reveals two risk factors. The first of these factors is comprised of questions 1, 2, 4, and 5 and is used in further analyses. This risk preference factor appears to capture general risk and has a Cronbach's alpha reliability of .62. Because the second factor is comprised of only two questions (two of the three questions added to Zaleskiewicz's original instrument to address audit-specific issues) and has a much lower reliability (.40), it is not used. 4.3. Descriptive statistics Table 1 shows means and standard deviations for the dependent variable (intention to use technology), budget pressure, and risk preference. The descriptive statistics are shown in total and by treatment group for the two independent variables, budget period and partner influence. Intention to use technology differs significantly between budget periods (p b .01, 2-tailed), with a greater intention indicated in the 3-year budget period. These results also suggest that these auditors feel a good deal of budget pressure, with the mean of 5.2, well above the scale midpoint of 4.

Table 2 Pearson correlations

PER INF BP RP

USE

PER

INF

BP

.236* .093 .105 .139

.023 .046 .110

−.036 −.051

.078

*p b .01, 2-tailed. USE = Sum of 3 items measuring intention to use new technology on the engagement (range 3–30). PER = Budget period (0 = 1-year, 1 = 3-year). INF = Influence (0 = none from partner, 1 = partner encourages use). BP = Response to “On past audit engagements, I have felt significant pressure to control budget hours” (range 1–7). RP = Sum of 4 items measuring general risk preference (range 4–28).

114

M.B. Curtis, E.A. Payne / International Journal of Accounting Information Systems 9 (2008) 104–121

Table 3 ANOVA results for intention to use technology

PER INF RP BP PER × INF PER × RP PER × BP INF × RP INF × BP RP × BP PER × INF × RP PER × INF × BP PER × RP × BP INF × RP × BP Error Total

Numerator df

F statistic

p-value a

1 1 1 1 1 1 1 1 1 1 1 1 1 1 124 139

2.266 6.987 8.462 6.262 .004 2.191 3.067 8.052 4.120 5.732 .559 .720 3.445 5.187

.068 .005 .002 .007 .952 .141 .082 .005 .045 .018 .456 .398 .066 .024

PER = Budget period (0 = 1-year, 1 = 3-year). INF = Influence (0 = none from partner, 1 = partner encourages use). RP = Sum of 4 items measuring general risk preference (range 4–28). BP = Response to “On past audit engagements, I have felt significant pressure to control budget hours” (range 1–7). a p-values are presented as 1-tailed for hypothesized main effects, as 2-tailed for interactions.

Pearson Correlations between variables are shown in Table 2. The significant correlation between budget period and intention lends support for H1. Although not significant, the unexpected positive association between budget pressure and intention portends subsequent analytical results. 4.4. Statistical analyses To test our hypotheses, we used ANOVA with budget period (1 or 3 years), partner influence (none or yes), risk preference and budget pressure, as presented in Table 3. H1 predicted budget period to affect intention. Although the correlation between budget period and intention was significant, in the presence of other factors the main effect is only marginally significant (F = 2.266, p b .07). H2 predicted partner influence to affect intention. This hypothesis is supported since the main effect is significant (F = 6.987, p b .01). The interaction between these two variables was not significant. Individual characteristics play a role in how people react to risks and pressure, such as the decision to implement new technology, and these characteristics may act alone or in combination with other characteristics or events (DeZoort and Lord, 1997). Indeed, both risk preference (F = 8.462, p b .01) and budget pressure (F = 6.262, p b .01) were individually significant in predicting the technology implementation decision. The significant, positive relationship between risk preference and technology acceptance supports H3a. The significant, positive relationship between budget pressure and technology acceptance is the opposite of the hypothesized relationship; therefore, H4a is not supported. One possible explanation for this is that auditors who perceive greater levels of budget pressure also see the new technology as a way to relieve budget pressure in the long run. To examine the moderating effects of risk preference and perceived budget pressure on the manipulated variables, all two- and three-way interactions were included in the ANOVA.11 Two-way interactions indicate that budget pressure (F = 4.120, p b .05) and risk preference (F = 8.052, p b .01) both moderate the impact of managing partner influence on the technology acceptance decision, but budget pressure (F = 3.067, p N .08) only marginally alters the impact of budget period on this decision and the risk preference-budget period interaction was not significant (F = 2.191, p N .14). Those who perceive greater budget pressure are more likely to support adoption when they know the managing partner supports use of the 11 It is appropriate to include all interaction terms in the model since there is theoretical support (DeZoort and Lord, 1997) for these interactions (Hartmann and Moers, 1999).

M.B. Curtis, E.A. Payne / International Journal of Accounting Information Systems 9 (2008) 104–121

115

software. Similarly, those who are risk-averse are more likely to support its adoption when they know the partner supports its use. In regard to budget period, those who are risk-seeking are even more likely to support adoption when the period is longer. Two very interesting 3-way interactions also emerge from these analyses. Mean splits of these variables allowed us to perform post-hoc analyses, in order to interpret the interactions. First, the ANOVA indicates a significant interaction between influence, risk preference and budget pressure (F = 5.187, p b .024). The nature of this interaction (illustrated in Fig. 1, panels A and B) is that in the presence of partner influence, perceived budget pressure does not affect intention to use technology for either risk-seekers (t = 1.404, p N .17, 2-tailed) or risk-averse individuals (t = 1.693, p N .10, 2-tailed), yet in the absence of partner influence, perceived budget pressure has no effect on intention to use technology for risk-seekers (t = −1.443, p N .15, 2-tailed), but appears to be the catalyst for risk-averse individuals to use technology (t = 2.081, p b .05, 2-tailed). The other marginally

Fig. 1. Interaction between partner influence, budget pressure and risk preference on intention to use technology. Panel A: no partner influence. Panel B: partner influence.

116

M.B. Curtis, E.A. Payne / International Journal of Accounting Information Systems 9 (2008) 104–121

significant interaction is between budget period, risk preference and budget pressure (F = 3.445, p b .07). The nature of this interaction (illustrated in Fig. 2, panels A and B) is that in a 3-year budget period, perceived budget pressure has no effect on intention to use technology for either risk-seekers (t = −.083, p N .93, 2-tailed) or risk-averse individuals (t = 1.116, p N .27, 2-tailed). However, in a 1-year budget period, again, risk-seekers are not affected by perceived budget pressure (t = −.609, p N .54, 2-tailed), but risk-averse individuals have higher intentions to use technology when they feel greater budget pressure (t = 2.333, p b .03, 2-tailed). As mentioned earlier when discussing the main effect of perceived budget pressure, the results in the three-way interactions are counter to expectations regarding budget pressure. That is, we expected those who feel higher levels of budget pressure to be less likely to use new technology. However, the results suggest that those under high levels of budget pressure are forward-thinking; they see the new technology as a way to relieve budget pressure in the future. Overall, these results suggest that external variables such as partner influence and longer budget periods affect the decision to use new technology, as do individual differences. In the absence of these external variables, individual differences will determine behavior. Risk-seekers are more likely to use

Fig. 2. Interaction between budget period, budget pressure and risk preference on intention to use technology. Panel A: 1-year budget period. Panel B: 3-year budget period.

M.B. Curtis, E.A. Payne / International Journal of Accounting Information Systems 9 (2008) 104–121

117

technology regardless of perceived budget pressure, but for risk-averse individuals intention to use technology is positively related to levels of perceived budget pressure. Thus, budget pressure influences individuals toward the acceptance of technology, despite the fact that technology is only beneficial to budget attainment in the long run and is detrimental to budget attainment in the short run. These findings provide support for H3a, H3b and H4b, but do not support H4a. Demographic characteristics were considered as moderators in our results, including age, gender, and a wide range of experience indicators (firm, position, fraud, large-client, small-client, and electronic tools). None were significant when entered individually into the model and the model results did not change. Finally, two items in the questionnaire asked whether the auditors thought their firm would be willing to use long-term budgeting for (1) project management and cost control, and (2) audit team performance evaluation. On a scale from 1–7 (not at all likely–definitely), the mean responses to these questions were 3.6 (SD 1.5) and 3.1 (SD 1.6), respectively. These responses suggest a somewhat pessimistic response since the means are less than the scale midpoint of 4. 5. Discussion Although audit technology may offer opportunities for increased efficiency and effectiveness, there are numerous risks and costs associated with implementing new technology in an audit engagement. These decisions are particularly problematic when the gains from technology use would be strictly internal, where it is unlikely that the audit client would agree to an increased budget to cover the cost of software. While much has been learned about technology implementation in the MIS area, very little attention has been given to this topic as it relates to an audit engagement. Thus, the overall purpose of this study was to gain a richer understanding of the unique factors affecting the technology implementation decision in an auditing context. The study considers two aspects of public accounting culture: the primary one is time-related performance pressure and the role budgets play in this; the secondary aspect is the influence (power) that partners possess in influencing the behavior of individuals within firms. These, in turn, play on a pressure fairly unique to public accounting — that of the up-or-out promotion processes and the evaluation pressure individuals feel in this regard. To develop our study, we started with a well-known systems model of the decision to implement new technology, the Unified Theory of Acceptance and Use of Technology (Venkatesh et al., 2003) and modified this model to incorporate factors unique to the audit setting. We examined two external/contextual factors including the length of the budget and performance evaluation period (1-year vs. 3-year) and influence from a remote superior (none vs. managing partner encourages use), as well as individual differences in risk preference and perceptions of budget pressure. In addition, we gathered information to ascertain whether audit firms would be amenable to using longer-term budget/evaluation periods. The use of fixed budgets in auditing and the risks associated with new technology complicate the implementation decision. External factors and individual characteristics may act alone or in combination to encourage or discourage individuals faced with the decision to implement new technology. Our analyses show that auditors are more likely to implement a new audit technology when in a 3-year budget/ evaluation period. A longer budgeting and evaluation period will smooth start-up costs from the first year so they can be matched to cost savings in later years. In this situation, auditors will have less concern about the negative effects of budget overruns in the first year on their engagement and annual evaluations. These results are consistent with findings in managerial accounting studies and support the idea that the use of longer-term budgets can overcome myopic decision making. Our analyses also show that auditors are more likely to implement new technology when they are aware that the managing partner is encouraging implementation within the firm. In this case, the practice office managing partner's influence was remote to the in-charge auditor (separated from direct influence by their engagement manager and partner) and there was no suggestion that this individual would know of the auditor's acceptance decision or participate in their evaluation. Yet, this remote influence was so pervasive that it had a direct influence on the technology acceptance decision. These results support previous accountability studies showing that auditors' decision making is influenced by the known views of their superiors and suggest that partners have extraordinary, perhaps unexpected, influence on individuals at all levels of their firms. It is interesting to note that these two external or contextual factors influence whether in-charge auditors will use new technology and this influence appears to be strong enough to overcome individual

118

M.B. Curtis, E.A. Payne / International Journal of Accounting Information Systems 9 (2008) 104–121

characteristics. These results should be interesting to firms in their general effort to increase the use of technology and to understand factors affecting implementation decisions by audit teams. Unfortunately, the auditors in this study were not very optimistic when asked whether they thought their firms would be willing to use longer-term budgets. This skepticism regarding the willingness of firms to consider longerterm budgets may explain why the relationship between budget-term and the implementation decision was only marginally significant. While these results only reflect the views of in-charge auditors, future research can examine whether those in the decision-making capacity relative to budget period would be amenable to using longer-term budgets. In order to consider the impact of budget length on technology acceptance decisions, this study is posed within the context of a fixed budget. It is possible that decisions could differ in situations where the budget can be revised for increased costs. Further research on technology acceptance decisions between fixed and flexible budgets is necessary to identify how these decisions differ. Additionally, future studies may explore other variables impacting the implementation decision, including the ability to use the technology on future engagements and the affects of rotating audit staff. The results also show that, in the absence of these external factors, the implementation decision depends upon individual characteristics, including risk-aversion and individual perceptions of budget pressure. Risk-seekers are more likely to implement than risk-averse individuals, and their decision to implement is positively related to perceived budget pressure. While the positive association of risk-seeking individuals and technology acceptance was expected, the positive association between budget pressure and technology acceptance for risk-averse individuals was not. These findings certainly suggest the need for further research into the interactive nature of risk preference and perceived pressure. It is possible that Prospect Theory can be explored for its relevance to this question (Rose et al., 2004). This also contributes to our understanding of this issue in that, if firms choose to address firm culture issues that create impediments to technology adoption, they must address not just the real impediments (such as 1-year budgets used for performance evaluation) but also the perceptual impediments possessed by their employees. The later may be the most difficult to modify. Although pressure was held constant in this study, in order to facilitate the measurement of individual differences in perceptions of pressure, pressure can also be manipulated as a firm-level factor, employed for management control purposes. Prior studies (Kelley and Margheim, 1990; Otley and Pierce, 1996) regarding the impact of budget pressure employed for control purposes suggest that pressure may be a medicine best applied with caution. Further research regarding the source and influence of individual perceptions of budget pressure is encouraged. In addition, the experiment included a situation whereby the managing partner encourages use of the software, but it is not certain that the software would be used for other clients. We acknowledge the fact that partner encouragement in light of the possible inability to use the software on other engagements may not be realistic in many situations. Because pilot study discussions with practitioners indicated the ability to use on other engagements was an important factor in their decision making, we employed this approach to control for varying assumptions between participants. In regard to risk preference, our measure resulted in a reliability score that is somewhat lower than typical norms. While the scale questions appear to reflect an individual's general risk propensity, the lower reliability score suggests careful interpretation of this variable and comparability with other measures of risk preference. While this model was tested on in-charge auditors, due to the suggestions that conflicts between cost and quality occur most strongly at this level, it seems reasonable that similar decision processes would be employed by those at higher levels in the firm. Future research should identify how the determinants of technology acceptance change for those at higher levels in the firm. If projections of the advancing demand for continuous reporting (Elliott, 2002) hold true, firms will have no choice but to automate their substantive testing to provide the continuous assurance such reporting dictates. Therefore, it is important to determine whether those at the higher levels of the firm are prepared to accept this development, and thus maintain their competitive advantage. Preliminary surveys of managers at this firm suggest widely disparate opinions regarding willingness to employ CAATs on their engagements (see footnote 7). Furthermore, future research can determine if the results of this study generalize to all accounting firms. In our search for a theoretical model to study this issue, we found no precedents. Thus, there was a risk that a model designed to address technology use (TAM) might not inform technology implementation

M.B. Curtis, E.A. Payne / International Journal of Accounting Information Systems 9 (2008) 104–121

119

decisions. A contribution of this research is that we were able to demonstrate this broader application of the TAM theory than has previously been demonstrated. The public accounting environment provided a unique context in which to study technology selection and implementation decisions, because these decisions are made on every audit, every year. In contrast, technology selection and implementation is much less common in the corporate environment, making its study much more difficult. However, we encourage the continued study of technology selection and implementation, beyond the mere acceptance of technology already implemented. Data in this study supports anecdotal evidence that auditors actually use very little audit software today. If CAATs do increase audit quality and efficiency, then some mechanisms must be at work to discourage their use. In the absence of explicit requirements for the use of particular audit tools and techniques, audit teams tend to be relatively entrepreneurial. Therefore, firm resources and rewards must be aligned with the firm's long-term interests in order to influence the use of optional software tools. Appendix A. Case study You are the returning in-charge auditor for an annual financial statement audit. Your firm is on a 1-year budgeting basis with this client — both in terms of budgeting cost and performance evaluation, although you do have a commitment of 3 years for the financial statement audit. Based on this 1-year budget, you will be evaluated for budget attainment at the end of each year. (Your firm has a 3-year commitment with this client and has therefore decided to use a combined 3-year budgeting basis for the engagement — both in terms of budgeting cost and performance evaluation. Performance evaluation will be based on budget attainment for the three years combined, therefore one-time expenditures will be spread over the three-year budget period.) The annual budget for the financial statement audit (excluding Sarbanes–Oxley internal control review) is 1500 total hours, with a $200 average hourly billing rate. This will provide total annual revenue from this engagement of $300,000. After the current-year engagement letter is signed (based on this budget), you learn the firm has introduced new software for testing payroll and disbursements which promises future audit efficiency. If the new audit technology is employed on the audit, both the staff person assigned to audit payroll and the incharge must learn to use the software. You estimate that the budgetary impact of the new software is a current-year additional cost of 50 h (not included in the previous budget). There will be no improvements in efficiency this year, but the software should provide a savings in future years of 50 h a year. This will replace existing tests, but is not expected to improve overall audit quality. This audit engagement has met budget in prior years. (This audit engagement has been approximately 3% over budget in the past two years.) There is no information regarding whether the engagement partner or the practice office managing partner support use of the software. (The engagement partner told you that the practice office managing partner is not going to require use of the new software, but is encouraging implementation.) It is not clear whether this software will be used on other clients in your office. It is probable that the current engagement team will be reassigned to this engagement next year. Assume that your recommendation will be instrumental in the final software adoption decision. Appendix B Origin of Questions Employed in Risk-Preference Measure Questions from Zaleskiewicz (2001) The skill of reasonable risk-taking is one of the most important managerial skills. To achieve something in life one has to take risks. If there is a big change of profit I take even very high risks. I willingly take responsibility in my work-place.

Modified for audit context 1. The skill of reasonable risk-taking is one of the most important managerial skills. 2. To achieve something in life, one must take risks. 3. If there is a big chance of profit I am willing to take big risks. 4. I willingly take new responsibilities and new projects in my firm and on my audits.

Questions derived for audit-specific context 5. I like working on first-time audits at my firm. 6. In my audit engagements, I prefer budgetary certainty to budgetary uncertainty. (reverse coded) 7. I like “playing it safe” with regard to engagement budgets. (reverse coded)

120

M.B. Curtis, E.A. Payne / International Journal of Accounting Information Systems 9 (2008) 104–121

References American Institute of Certified Public Accountants (AICPA). The effect of information technology on the auditor's consideration of internal control in a financial statement audit. . Statement of Auditing Standards No. 94. New York NY: AICPA; 2001. American Institute of Certified Public Accountants (AICPA). Code of Professional Conduct. New York NY: AICPA; 2006. Anderson JC, Moreno KK, Mueller JM. The effect of client vs. decision aid as a source of explanations upon auditors' sufficiency judgments: a research note. Behavioral Research in Accounting 2003;15:1–11. Banker RD, Chang H, Kao Y. Impact of information technology on public accounting firm productivity. Journal of Information Systems 2002;16(2):209–22. Bedard JC, Graham LE. The effects of decision aid orientation on risk factor identification and audit test planning. Auditing: A Journal of Practice & Theory 2002;21(2):39–56. Bedard JC, Jackson C, Ettredge ML, Johnstone KM. The effect of training on auditors' acceptance of an electronic work system. International Journal of Accounting Information Systems 2003;4:227–50. Brazel JF, Agoglia CP, Hatfield RC. Electronic versus face-to-face review: the effects of alternative forms of review on auditors' performance. The Accounting Review 2004;79(4):949–66. Chen R, He F. Examination of brand knowledge, perceived risk and consumers' intention to adopt an online retailer. TQM and Business Excellence 2003;14(6):677–93. Davis FA. Perceived usefulness, ease of use and user acceptance of technology. MIS Quarterly 1989;13(3):319–40. DeZoort FT, Lord AT. A review and synthesis of pressure effects research in accounting. Journal of Accounting Literature 1997;16:28–85. Elliott RK. Twenty-first century assurance. Auditing: A Journal of Practice & Theory 2002;21(1):139–46. Fischer MJ. “Realizing” the benefits of new technologies as a source of audit evidence: An interpretive field study. Accounting, Organizations and Society 1996;21(2/3):219–42. Fisher J, Peffer SA, Sprinkle GB. Budget-based contracts, budget levels, and group productivity. Journal of Management Accounting Research 2003;15:51–74. French JR, Caplan RD. Organizational stress and individual strain. In: Marrow AJ, editor. The failure of success. New York: AMACOM; 1972. p. 30–66. Hansen SC, Otley DT, Van der Stede WA. Practice developments in budgeting: an overview and research perspective. Journal of Management Accounting Research 2003;15:95–116. Hartmann F, Moers F. Testing contingency hypotheses in budgetary research: an evaluation of the moderated regression analysis. Accounting, Organizations and Society 1999;24(4):291–315. Hope J, Fraser R. Beyond budgeting: how managers can break free from the annual performance trap. Boston, MA: Harvard Business School Press; 2003. Hopwood AG. An empirical study of the role of accounting data in performance evaluation. Journal of Accounting Research 1972;10:156–82 (Supplement). Houston R. The effects of fee pressure and client risk on audit seniors' time budget decisions. Auditing: A Journal of Practice and Theory 1999;18(2):70–86. Hunt SC. A review and synthesis of research in performance evaluation in public accounting. Journal of Accounting Literature 1995;14:107–39. Kelley T, Margheim M. The impact of time budget pressure, personality, and leadership variables on dysfunctional auditor behavior. Auditing: A Journal of Practice and Theory 1990;9:21–42 (Fall). Kite DM, Katz JP, Zarzeski MT. Can managers appraise performance too often? Journal of Applied Business Research 1996–97;13:41–52 (1-winter). Lazarus RS. Psychological stress in the workplace. In: Crandall R, Perrewe PL, editors. Occupational stress: a handbook. Washington, DC: Taylor & Framcos; 1995. p. 3–14. Lerner J, Tetlock P. Accounting for the effects of accountability. Psychological Bulletin 1999;125(2):255–75. Levinthal DA, March JG. The myopia of learning. Strategic Management Journal 1993;14:95–112 (Winter). Loraas T, Wolfe C. Why wait? Modeling the factors that influence the decision of when to learn a new use of technology. Working paper. Texas A&M University; 2005. Lord AT, DeZoort FT. The impact of commitment and moral reasoning on auditors' responses to social influence pressure. Accounting, Organizations and Society 2001;26:215–35. Lovata, Linda M. The utilization of generalized audit software. Auditing: A Journal of Practice and Theory 1988:72–86 (Fall). Lowe DJ, Reckers PMJ, Whitecotton SM. The effects of decision-aid use and reliability on jurors' evaluations of auditor liability. The Accounting Review 2002;77(1):185–202. March JG, Shapira Z. Managerial perspectives on risk and risk taking. Management Science 1987;33(11):1404–18. Merchant KA. The effects of financial controls on data manipulation and management myopia. Accounting, Organizations and Society 1990;15(4):297–313. Neely A, Sutcliff MR, Heyns JR. Driving value through strategic planning and budgeting. New York, NY: Accenture; 2001. Nelson R, Cheney P. Training end users: an exploratory study. MIS Quarterly 1987:547–59 (Dec). Otley DT, Pierce BJ. Auditor time budget pressure: consequences and antecedents. Accounting, Auditing and Accountability 1996;9:31–58 (Spring). Peecher ME, Solomon I. Theory and experimentation in studies of audit judgments and decisions: avoiding common research traps. International Journal of Auditing 2001;5:193–203. Pierce B, Sweeney B. Cost–quality conflict in audit firms: an empirical investigation. European Accounting Review 2004;13(3):415–41. Rose JM, Wolfe CJ. The effects of system design alternatives on the acquisition of tax knowledge from a computerized tax decision aid. Accounting, Organizations and Society 2000;25:285–306. Rose JM, Rose AM, Norman CS. The evaluation of risky information technology investment decisions. Journal of Information Systems 2004;18(1):53–66. Shapeero MHC, Koh, Killough LN. Underreporting and premature sign-off in public accounting. Managerial Auditing Journal 2003;18 (6/7):478–89.

M.B. Curtis, E.A. Payne / International Journal of Accounting Information Systems 9 (2008) 104–121

121

Shields JF, Shields MD. Antecedents of participative budgeting. Accounting, Organizations and Society 1998;23(1):49–76. Sitkin SB, Weingart LR. Determinants of risky decision-making behavior: a test of the mediating role of risk perceptions and risk propensity. Academy of Management Journal 1995;38:1573–92. Sweeney B, Pierce B. Management control in audit firms: a qualitative examination. Accounting, Auditing and Accountability Journal 2004;17(5):779–812. Tan CEL, Jubb CA, Houghton KA. Auditor judgments: the effects of the partner's views on decision outcomes and cognitive effort. Behavioral Research in Accounting 1997;9:157–75 (supplement). Tetlock P. The impact of accountability on judgment and choice: toward a social contingency model. In: Kerkowitz L, editor. Advances in Experimental Social Psychology, 25. New York, NY: Academic Press; 1992. p. 331–76. Trimpop RM, Kerr JH, Kirkcaldy B. Comparing personality constructs of risk-taking behavior. Personality and Individual Differences 1999;26:237–54. Turner CW. Accountability demands and the auditor's evidence search strategy: the influence of reviewer preferences and the nature of the response (belief vs. action). Journal of Accounting Research 2001;39(3):683–706. Vendrzyk VP, Bagranoff NA. The evolving role of IS audit: a field study comparing the perceptions of IS and financial auditors. Advances in Accounting 2003;20:141–63. Venkatesh V. Creation of favorable user perceptions: exploring the role of intrinsic motivation. MIS Quarterly 1999;23(2):239–60. Venkatesh V, Morris MG, Davis GB, Davis FD. User acceptance of information technology: toward a unified view. MIS Quarterly 2003;27(3):425–78. Wang XT. Framing effects: dynamics and task domain. Organizational Behavior and Human Decision Processes 1996;68:145–57. Wilks TJ. Predecisional distortion of evidence as a consequence of real-time audit review. Accounting Review 2002;77(1):51–71. Zaleskiewicz T. Beyond risk seeking and risk aversion: personality and the dual nature of economic risk taking. European Journal of Personality 2001;15:105–22. Zickar MJ, Highhouse S. Looking closer at the effect of framing on risky choice: an item response theory analysis. Organizational Behavior and Human Decision Processes 1998;75:75–91. Zhao H, Seibert SE, Hills GE. The mediating role of self-efficacy in the development of entrepreneurial intentions. Journal of Applied Psychology 2005;90(6):1265–72.