Home page | USC Marshall



A Habit Strength-Based Explanation for Auditors’ Use of Simple Cognitive Processes for Complex Tasks Sarah BonnerUniversity of Southern CaliforniaKathryn KadousEmory UniversityTracie MajorsUniversity of Southern CaliforniaMay 15, 2020We thank Leo Barcellos, Lori Bhaskar, Scott Emett, Cassandra Estep, Lisa Gaynor, Emily Griffith, Ryan Guggenmos, Bright Hong, Justin Leiby, Eldar Maksymov, Chris Nolder, Mark Peecher, Truman Rowley, Adam Vitalis, Chris Wolfe, Wendy Wood, and workshop participants at Arizona State University, Texas A&M University, and the Texas Audit Symposium for helpful feedback. We thank Erik Fujinami, Bright Hong, JungKoo Kang, Andrew Kim, Sharon Kim, Suteera Pongtepupathum, Taylor Reis, Michael Rezzo, Stacey Ritter, Meredith Schultz, and Fiona Wang for helpful feedback on the instrument. We also thank Nathaniel Young for the photographs used in the instruments, and Bright Hong, Ryan Kalaf, Molly Starobin, Genna Young, and Dana Zadeh for other research assistance. We also thank the four senior auditors with whom we conducted extensive interviews for giving their time. This study was supported by a Center for Audit Quality Research Advisory Board grant. We thank Margot Cella and Lauren Tuite of the CAQ for facilitating access to participants, members of the Research Advisory Board, especially John DeMelis, for helpful input, and participants for giving their time.The views expressed in this article and its content are those of the authors alone and not those of the Center for Audit Quality.A Habit Strength-Based Explanation for Auditors’ Use of Simple Cognitive Processes for Complex TasksWe experimentally examine whether audit seniors’ use of simple cognitive processes for a complex task is affected by the strength of habits to use these processes that they developed as staff auditors. A habit is a mental association between a behavior and a specific context. We propose that, for seniors with stronger simple process habits, the typical audit room context automatically and unconsciously activates the simple processes, impeding selection of task-appropriate processes. Consistent with this, in the typical context, seniors with stronger habits identify fewer issues with a complex estimate than seniors with weaker habits. Further, seniors with stronger habits perform better in an alternative context that does not activate the simple processes, while those with weaker habits do not. Additional analyses validate the habit strength measure, including identifying determinants, which is valuable because preventing formation of strong simple process habits may be more feasible than changing established habits.JEL codes: G10, M40, M41, M42, D80, D91Keywords: habits, cognitive processing, accounting estimates, audit quality, goodwill impairment, fair value I. INTRODUCTIONComplex accounting estimates, including goodwill and valuation allowances, are susceptible to misstatement due to management bias, implying that effective auditing is critical for financial reporting quality (Bratten et al. 2013; Griffith, Hammersley, and Kadous 2015; Cannon and Bedard 2017). However, inspection reports cite continued deficiencies in audits of complex estimates, particularly in evaluations of underlying assumptions (IFIAR 2017; IFIAR 2019). These reports and a growing body of research (Kadous and Zhou 2019; Griffith 2018; Joe, Wu, and Zimmerman 2017; Griffith, Hammersley, Kadous, and Young 2015; Austin, Hammersley, and Ricci 2019) suggest that deficiencies in this area are caused by auditors’ use of simple cognitive processes, such as superficial processing, which impairs their ability to identify issues with assumptions. Yet, we know little about why this is the case. We examine whether the extent to which seniors use simple cognitive processes when auditing a complex estimate is affected by the strength of habits to use these processes that they developed as staff auditors. A habit is an association in memory between a behavior and a stable context (generally one’s surroundings, including other people) in which that behavior is enacted; the association develops as the behavior is repeated in that context with experienced rewards (Wood and Rünger 2016; Mazar and Wood 2018). For example, moviegoers can develop habits to eat popcorn by frequently attending movies in a theater and repeatedly eating and enjoying popcorn while there (Neal, Wood, Wu, and Kurlander 2011). Likewise, we posit that staff auditors can develop habits to use simple cognitive processes in the audit room context in which they typically work (a conference room at a client site housing multiple team members). This context is fairly stable over time and across clients. Many staff likely repeat the simple processes when performing simple tasks because they are sufficiently effective and particularly efficient for these tasks. Also, some staff may feel rewarded for using the processes, such as by experiencing positive reinforcement from superiors for efficiency (Nelson and Proell 2018). These characteristics of the environment staff face are thus conducive to the formation of habits to use simple processes. Habits resist extinction (Wood and Neal 2016; Verplanken, Roy, and Whitmarsh 2018), such that we expect staff, when promoted to senior, will “carry forward” their habits. However, we expect these habits will vary in strength, based on the extent that seniors, as staff, repeatedly used the simple processes and felt rewarded for doing so. Stronger habits imply stronger context-behavior associations in memory such that, when people with strong habits experience the context, it is likely they will automatically respond with the behavior even if it is undesirable for the current task (Wood and Rünger 2016). For example, people who have a strong “eat popcorn at the movies” habit eat the same amount of stale as fresh popcorn when in a theater context, despite indicating they dislike the stale popcorn (Neal et al. 2011). Similarly, we expect that seniors with strong simple process habits will use the simple processes when auditing an estimate in the typical audit room context, despite their inappropriateness for the task. Thus, we predict that seniors with stronger simple process habits will be less effective at identifying issues with the assumptions underlying an estimate than will seniors with weaker habits in this context. Altering context cues prevents the automatic activation of habitual responses, allowing people instead to act with intention (Wood, Tam, and Witt 2005). For example, moviegoers with strong popcorn-eating habits react positively to a meeting room context: they eat significantly less stale than fresh popcorn here because they can allow the tastiness of the popcorn to guide their behavior. We likewise expect that seniors with stronger simple process habits will be more likely to use the complex processes that are appropriate for the estimates task in the alternative context and, therefore, identify more issues with assumptions. In contrast, people with weak habits are less likely to have their habits activated by the context in which they were formed, so are less likely to respond positively to an alternative context. For example, moviegoers with weak popcorn-eating habits eat significantly less stale than fresh popcorn in both a theater and a meeting room (Neal et al. 2011). In our setting, seniors with weaker habits may be processing intentionally, in which case context should have no effect. However, auditors face conditions that are conducive to habit formation, such that these seniors may be processing habitually, but with complex processes instead. If so, they may perform worse in the alternative context when their habits are not activated and they must regroup (Wood 2019). Together, these expectations imply an interaction between simple process habit strength and context, which is our second prediction.We test our predictions in a 2 x 2 between-participants experiment with 128 experienced audit seniors from two large audit firms. Participants evaluate assumptions underlying a fair value estimate in a goodwill impairment case. The case contains embedded issues reflective of management bias, the identification of which is impeded by use of simple cognitive processes. Thus, our dependent measure is the number of issues auditors identify.We measure our first independent variable, simple process habit strength, using reaction times: how fast auditors, after being primed with a photograph of the typical context, complete word fragments related to the simple processes relative to control fragments. This implicit measure directly captures habit strength (Rebar, Gardner, Rhodes, and Verplanken 2018), here the strength of the typical audit room context-simple processes association in memory, and has been used in prior studies of habits (e.g., Danner, Aarts, and de Vries 2008; Adriaanse et al. 2011; Neal, Wood, Labrecque, and Lally 2012; Hargadon 2017). We use a median split to classify auditors as having stronger or weaker habits.Consistent with psychology research (e.g., Neal et al. 2012), we manipulate context to test our hypotheses about habit strength. We randomly assign auditors to imagine working in the typical or an alternative context when working on the estimates task. Participants in the typical context view a photograph of a typical audit room. This context is designed to activate, with differential probability depending on strength, auditors’ simple process habits. Those in the alternative context view the same audit room, but with key cues altered; to the extent auditors’ habits were formed only in the typical context, the alternative context should not activate them. Results support our predictions. In the typical context, seniors who have stronger simple process habits identify fewer issues than weaker habits seniors. This effect is not reduced as auditors have more experience. We also find an interaction between habit strength and context. Seniors with stronger habits identify more issues in the alternative versus typical context. Seniors with weaker habits identify marginally fewer issues in the alternative context. Additional analyses suggest that at least some of these auditors have complex process habits. The observed interaction between context and habit strength is consistent with theory and supports that habit strength drives our results. However, we strengthen this inference. First, we analyze theoretical determinants of habit strength. We find that the extent to which seniors, as staff, repeated behaviors reflecting the simple processes and felt rewarded for doing so predict habit strength. Further, completion of simple (versus complex) tasks, as staff, led to repetition of simple processes, but less so for auditors with high effectiveness preferences. Second, we show that stronger simple process habits auditors who likely also developed their habits in the alternative context are not influenced by context because their strong simple process habits are activated here as well. Third, we show that stronger simple process habits auditors who inhibit the habitual response by exerting self-discipline are not influenced by the alternative context. Our findings contribute to the auditing literature and practice. Habits are a form of Type 1 processing (Evans and Stanovich 2013) that, while garnering significant interest in psychology (Verplanken 2018), have not yet been examined in auditing. Type 1 processes likely are generally influential for auditor judgments because auditors face a number of depleting factors, such as stress, fatigue, and cognitive load, that lead to reliance on these processes (e.g., Potthoff et al. 2018). Moreover, the audit environment, with its stable context and repetition of tasks, is conducive to habit formation. Thus, habits may be a particularly important Type 1 process in auditing. While researchers have speculated that features of the environment such as the piecemeal, checklist-like nature of audit programs could encourage simple processing (Griffith et al. 2015a; Griffith et al. 2015b), our study provides evidence that seniors’ strong simple process habits that develop as a result of factors that vary when they are staff, including exposure to simple tasks, experienced rewards for using simple processes, and effectiveness preferences, may be a fundamental cause of this behavior. Importantly, while these habits likely were adaptive for these seniors as staff, they become problematic after promotion because seniors continue to work in the same context as they did when staff, but their tasks become more complex. Indeed, firms are realizing the importance of habits, but they generally focus on personal health habits (PwC 2020). We believe the habit construct can be extended to audit tasks. Conceiving of habits as an underlying cause of behavior illuminates why deficiencies in audits of estimates and other complex accounts may persist. Habits are “hardwired” in memory and, when strong, resist interventions such as informational appeals, education, and instructions (Verplanken and Wood 2016; Lally and Gardner 2013). This implies that interventions such as revisions to audit standards or increased training are unlikely to change the behavior of seniors with strong simple process habits. Psychology theory suggests that getting these auditors to use complex processes requires changing context, or helping them inhibit habitual processing, both of which present challenges. Successful context change requires identifying which auditors would benefit and which would not. Inhibiting effects of strong habits requires self-discipline and is especially challenging in depleting environments (Itzchakov, Uziel, and Wood 2018). A third possibility involves helping these seniors develop new, complex process habits that would replace the old ones. However, strong (old) habits are difficult to extinguish; in support, we find that the negative effect of strong simple process habits is not diminished by experience. In light of these challenges, firms likely would benefit most from preventing staff from developing simple process habits in the first place. Our analysis of determinants offers concrete ideas. Firms could reduce staff’s time spent on simple tasks by encouraging seniors to involve staff in complex tasks. Firms could prime professional identity, as we show that auditors with high identity are less apt to use simple processes. Finally, firms could reduce staff’s repetition of simple processes by leading them to instead use complex processes. This could be accomplished through seniors’ coaching approaches, or through technology such as visualization. While some inefficiency may result, using complex processes for simple tasks could lead to engagement-level benefits such as incidental detection of fraud in low-risk areas and earlier and/or faster development of complex process habits in individual auditors. The latter would be beneficial not only for auditing complex estimates but for other complex tasks such as auditing revenues. The rest of the paper is organized as follows. Section 2 provides theory and hypotheses. Sections 3 and 4 describe the design and results of the experiment. Section 5 concludes. II. BACKGROUND AND HYPOTHESIS DEVELOPMENTIn this section, we first develop theory that simple cognitive processes (superficial, confirmatory, and piecemeal processing) allow for effective and efficient completion of the simple tasks that staff frequently perform. Drawing from psychology theory on habits, we then argue that repetition of the processes in the typical audit room context with rewards can lead staff to develop habits of varying strength to use the processes. For seniors carrying forward stronger (versus weaker) simple process habits, the typical context is more likely to activate use of those processes, irrespective of task demands. Consequently, we predict that simple process habit strength will negatively affect seniors’ performance of these tasks. Simple Tasks Performed by Staff Auditors and Cognitive ProcessesStaff auditors frequently perform tasks that are relatively simple (e.g., Power 2003; Westermann et al. 2015; Westermann et al. 2019). We define simple tasks as those that involve unambiguous information cues and only a few cues to process, individually, at a time (Bonner 1994). For example, vouching prices from sales invoices to a price list is simple because it involves prices that are quantified and objective. Further, because vouching consists of a series of “micro-tasks” in that the auditor vouches prices for one invoice at a time, there are only a few cues to process per audited “item” (Bonner 1994). Finally, staff can process each cue (price) individually because the necessary integration to project the likely misstatement in sales typically is performed later by software or a senior. We propose, therefore, that staff performing simple tasks will use simple cognitive processes, specifically superficial, confirmatory, and piecemeal processing, because these processes are appropriate for these tasks. They lead to sufficiently effective performance and are far more efficient than complex processes (deep, nonconfirmatory, and integrative processing) and, as argued below, we expect staff receive positive reinforcement for exhibiting this tradeoff between efficiency and effectiveness when performing simple tasks. Psychology research defines superficial processing as focusing on obvious, surface features of information cues, and deep processing as analyzing their meaning and implications (Craik and Lockhart 1972; Craik 2002). Superficial processing is sufficient for effective performance of simple audit tasks given their unambiguous cues. For example, when vouching sales prices, auditors only need to glance at the price list to see if invoice prices are correct. While deeper processing, such as thinking about whether prices make sense under current economic conditions, also would lead to effective performance in this task, it is unnecessary to achieve the objective of testing the accuracy assertion for sales. Consequently, deeper processing during this simple task would lead to inefficiency. Confirmatory processing is searching for evidence supporting, and/or interpreting evidence to be consistent with, what one expects or desires to see; nonconfirmatory processing entails searching for and accurately interpreting contradictory information (e.g., Kunda 1990). Confirmatory processing also is sufficient for effective performance of simple audit tasks because of their unambiguous cues. For the vouching task, identifying incorrect prices requires only noting mismatches between invoices and the price list. Searching for contradictory evidence is not required because the evidence needed for drawing conclusions appears as part of the task; also, there is no interpretation needed to identify mismatches. While nonconfirmatory processing (e.g., scrutinizing each invoice with a price mismatch to ensure that there are no errors or fraud in other terms) also would lead to effective performance, again, it is unnecessary and inefficient. Finally, piecemeal processing involves bringing one or a few cues into working memory, evaluating them, then “closing them out” by removing them from memory, while integrative processing involves considering the implications of cues jointly, which necessitates returning to earlier cues for re-processing each time a new cue is introduced (e.g., Anderson 1981). Piecemeal processing is sufficient for effective performance of simple audit tasks, at least the “micro-task” part that is performed by staff. For the vouching task, staff compare the prices for one invoice to the price list, record any mismatches, and move to the next invoice knowing that information integration will be done later. Although integrative processing also could lead to effective performance, it is unnecessary and inefficient as integration would occur (again) later. To summarize, superficial, confirmatory, and piecemeal processing are sufficiently effective and particularly efficient for the simple tasks that staff perform frequently. Thus, we posit that staff, on average, repeat these processes, but that they do so to varying degrees depending on factors such as their exposure to simple tasks. In turn, repetition of behavior in a stable context with experienced rewards leads to habit formation (e.g., Mazar and Wood 2018). Simple Cognitive Processes as Habits A habit is a mental association between a behavior (e.g., eating popcorn) and a stable context in which that behavior is enacted (e.g., movie theater), with “context” referring to one’s surroundings, including physical surroundings and people (Wood and Rünger 2016). Here, the behaviors are the simple processes, and the context is a conference room at a client site in which a team of auditors typically works. Audit staff and seniors work in a stable context as they tend to be assigned to the same clients over time; moreover, features of the typical context are fairly consistent across clients. Hence, auditors face conditions highly conducive to habit formation.Initially, a given behavior may occur in response to a goal, the achievement of which is rewarded. For example, at first, a person may eat popcorn at a movie theater because it tastes good. Indeed, research shows that experiencing rewards strengthens habits (Gardner and Lally 2013). Similarly, at first, staff likely use simple processes to perform simple tasks in an audit room because they feel rewarded (e.g., by positive reinforcement from a superior for efficiency). Over time, however, with rewarded repetition in a stable context, behaviors can become more strongly linked to the context than to the initial goal (or task demands). Thus, behavioral control shifts to context, such that a behavior is automatically and unconsciously activated by the context rather than by task demands and the related anticipation of rewards (Wood and Neal 2007; Wood 2017; Verplanken 2018). For example, moviegoers who repeatedly enjoy popcorn can develop an “eat popcorn at the movies” habit, such that elements of the theater context (e.g., theater seats) activate the behavior to eat popcorn, rather than the desirability of the popcorn itself (Wood, Quinn, and Kashy 2002; Wood 2017). Likewise, we expect that staff who repeatedly use simple processes in the typical audit room context can develop “simple process habits,” such that elements of this context (e.g., a conference table) activate their cognitive processing, rather than the appropriateness of the processes for the demands of the particular task they are performing. Habit strength is the strength of the association between the context and behavior in memory. As with other associations in memory, habit strength increases with the frequency of rewarded co-occurrence of two items (Deese 1960). Thus, people form strong popcorn-eating habits with extensive experience eating and enjoying popcorn while watching movies in a theater. We expect that staff likewise can form strong simple process habits with extensive rewarded repetition of simple processes while working in the typical audit room context. As we note above, the simple processes are effective and highly efficient for the simple tasks staff tend to perform, and so, on average, likely are repeated and rewarded. That is, most audit teams face time and deadline pressures (e.g., Bobek, Daugherty, Radtke 2012; Brown, Gissel, Neely 2016; Westermann et al. 2019). To manage such pressures, superiors tend to reduce budgeted audit hours where possible, which likely is easiest in the lower risk areas where staff tend to work (e.g., Houston 1999; Bierstaker and Wright 2001, 2005). Staff will thus use the simple processes for simple tasks to manage to their superiors’ preferences (e.g., Bagley 2010), receiving positive reinforcement for so doing (e.g., Nelson and Proell 2018). Importantly, however, repetition and rewards will vary across individuals, creating variation in habit strength. Repetition of the processes will vary based on factors such as how frequently staff perform simple tasks. Staff also will vary in how much they feel rewarded. For example, staff with a high need for approval (Kelley 1994; Malone and Roberts 1996) may more highly value praise. When the context is experienced, habit strength determines the probability that habitual behaviors are automatically and unconsciously activated in memory, then enacted, irrespective of current goals or task demands (e.g., Wood and Rünger 2016). Because habits are resistant to extinction (e.g., Wood and Neal 2016; Verplanken et al. 2018), we expect that staff will carry forward their simple process habits when promoted to senior. Seniors who developed stronger habits will be more likely to use the simple processes when in the typical audit room. This is problematic because most auditors stay in this context as they move from staff to senior, but task demands change. Thus, to the extent the processes that are activated by context are inappropriate for the task at hand, seniors’ effectiveness will suffer.Effect of Simple Process Habit Strength on Evaluation of Assumptions in Typical ContextIndeed, prior research supports that auditing complex estimates is a task that cannot be completed effectively with simple processing (Kadous and Zhou 2019; Griffith 2018; Joe et al. 2017; Griffith et al. 2015b; Austin et al. 2019). This task is complex in that it involves ambiguous cues, such as predictions and qualitative information, that require interpretation and scrutiny to detect potential management bias. It also contains cues that have implications for other cues, including those that support one assumption but contradict others (Griffith et al. 2015a, 2015b; Kadous and Zhou 2019). Thus, superficial and piecemeal processing could cause auditors to miss evidence indicative of problems. Confirmatory processing could lead auditors to fail to identify contradictory evidence or to “explain it away.” We predict that the likelihood of the simple processes being activated automatically by the typical audit room context, even though the task requires complex processes, increases with habit strength. As a result, we predict that seniors with stronger habits will be less effective at auditing an estimate than seniors with weaker habits in the typical context. Consistent with stronger habits causing behavior inconsistent with intention, Neal et al. (2011) find that people with strong popcorn-eating habits eat as much stale as fresh popcorn when in a movie theater despite indicating dislike of the former. This occurs because, while people can in theory inhibit a strongly habitual response and choose a different action, doing so requires a great deal of self-discipline (e.g., Wood 2017). Further, alternative actions are less accessible for people with strong habits (e.g., Lally and Gardner 2013). Thus, while seniors with strong habits may understand conceptually that simple processes are inappropriate for the complex tasks they perform, we predict they are likely to stick with the automatically triggered simple processes. By contrast, the typical context is less likely to activate the simple processes for weaker habits auditors. Even if activated, these auditors can inhibit their habits and access the alternative complex processes with greater ease, relative to auditors with stronger simple process habits. Thus, we predict that auditors with weaker simple process habits will perform better than auditors with stronger habits. As mentioned earlier, we capture the use of simple processes using the issues auditors identify with the assumptions underlying an estimate. Stated formally:H1: In the typical audit room context, auditors with stronger simple process habits will identify fewer issues in assumptions than those with weaker simple process habits. Context as a Moderator of Habit StrengthPsychology research has shown that altering context cues can jar people out of engaging in strongly habitual behaviors (e.g., Wood and Neal 2009). Since the context in which the habit was developed is no longer available to activate the response (e.g., Verplanken et al. 2018), people are left to more thoughtfully consider which action is most appropriate (Wood 2019). For example, Neal et al. (2011) show that people with strong habits to eat popcorn at a movie theater respond positively to being placed in an alternative context (watching a video in a meeting room); they eat less stale popcorn than fresh popcorn. In fact, they eat no more stale popcorn than do people with weak habits in the alternative context. Likewise, we predict that seniors with stronger habits will be more effective at auditing a complex estimate in an alternative context. Their simple process habits will not be activated and we expect that, freed up to consider which processes are appropriate, they will use the complex processes needed for the task. While we predict that auditors with stronger habits will respond positively to a different context, we expect those with weaker habits are less likely to do so. One possibility is that these auditors are processing intentionally, which implies little reaction to context. This possibility is consistent with how persons with weaker habits for a behavior generally are characterized in psychology studies on habits. For example, participants with weaker popcorn-eating habits, who eat less stale than fresh popcorn in both the typical and alternative contexts, are characterized as intentionally deciding how much popcorn to eat based on tastiness (Neal et al. 2011). However, the audit environment is more complex than those previously studied. First, it is depleting (e.g., Hurley 2015). Intentional decision-making requires executive control, such that auditors engaging in intentional cognitive processing on their tasks would be “constantly in the throes of heavy mental lifting” (Wood 2019, 9). Thus, while people with weak popcorn-eating habits can intentionally choose whether or not to eat popcorn on the occasion they see a movie in a theater, depleted auditors may not be able to make intentional choices each time they engage in cognitive processing. Second, in psychology studies, there generally is no single behavior that is the obvious “opposite” of the studied behavior and that occurs as a consequence of having a weaker habit for the studied behavior. For example, people with weak popcorn-eating habits may not go to the theater frequently or, instead, may go frequently but eat nothing. In contrast, auditors must perform audit tasks and must use cognitive processing to do so. Consequently, there is an obvious “opposite” to using simple cognitive processes: using complex processes. Taken together, these arguments suggest that seniors carrying forward weaker simple process habits may process habitually, but may instead have habits for complex processes.To summarize, we predict that auditors with stronger simple process habits will respond positively to context. However, we predict that auditors with weaker simple process habits will be less likely to respond positively to context. If they are processing intentionally, they should be relatively unaffected. Alternatively, if they are using complex process habits, those habits will not be automatically activated by the alternative context; thus, they must regroup and start afresh, which could negatively affect their performance (Wood 2019). For example, people with healthy eating habits who experience a context change make less healthy food choices in the new context (Lin, Wood, and Monterosso 2016). Collectively, our predictions for the effects of context on seniors with stronger and weaker simple process habits imply the following hypothesis: H2: Auditors with stronger simple process habits will identify more issues in assumptions in the alternative audit room context than in the typical audit room context, while auditors with weaker simple process habits will not.III. METHODParticipants Participants are 128 experienced senior auditors (mean experience of 40.3 months) from two large audit firms. Study administration was facilitated by the Center for Audit Quality, with participants completing the study during firm training sessions. The estimates task requires participants to evaluate management’s assumptions related to a goodwill impairment test. Seniors typically perform this task in practice (Griffith et al. 2015a). Design and ProceduresOur study employs a 2 x 2 between-participants design. We measure simple process habit strength and use a median split to classify participants as having either stronger or weaker simple process habits. We manipulate context by randomly assigning auditors either to the typical audit room or an alternative audit room that eliminates many cues from the typical room. Following psychology research (e.g., Neal et al. 2011; Neal et al. 2012), we manipulate context not to examine its effects per se but to validate the habit strength construct by showing differential context effects when simple process habits are activated (in the typical context) versus not (in the alternative context). Observing theory-consistent effects of context will support the predictive validity of our habit strength measure (see Rebar et al. 2018). The study proceeds as follows. Participants first view a photograph of either the typical or alternative context and write a paragraph to reinforce the manipulation. They then complete the estimates task, which includes listing issues of concern regarding the assumptions. Next, they answer post-experimental questions. Finally, all participants view a photo of the typical context when completing the habit strength measure. We use this context when measuring habit strength because, as mentioned earlier, this is the context in which staff spend most of their time and, thus, are likely to develop habits. Use of the typical context for both the manipulation (to activate habits with differential probability based on strength) and during measurement of habit strength is consistent with prior research (e.g., Neal et al. 2011; Neal et al. 2012).Simple Process Habit Strength We measure simple process habit strength using a reaction-time measure. Implicit measures are advantageous for capturing the strength of mental associations between items, such as “birthday” and “cake” (De Houwer, Teige-Mocigemba, Spruyt, and Moors 2009). That is, the strength of a link in memory affects not only the probability, but also the speed, of activation of the second item when the first item is activated. Because habits are defined as associations in memory between a context and a behavior, a reaction time measure directly captures habit strength (Labrecque and Wood 2015; Mazar and Wood 2018; Rebar et al. 2018). While other measures such as frequency of past behavior in a stable context (Mazar and Wood 2018) capture individual determinants of habit strength, the reaction-time measure captures the resulting association in memory, which incorporates all determinants, including experienced rewards. Yet other measures capture outcomes of habit strength, such as the extent to which a behavior feels automatic (Verplanken and Orbell 2003). Our reaction-time measure captures automaticity using speed, but also overcomes limitations with outcome-based measures. For example, automaticity can resist conscious reflection (e.g., Hagger et al. 2015; Gardner 2015; Rebar et al. 2018). More fundamentally, we do not measure habit strength using outcomes because they can reflect other processes; that is, they do not capture the context-dependency of habits. Our habit strength measure assesses the speed of auditors’ recognition of words reflective of the simple processes after experiencing the context in which we expect most have developed their processing habits (i.e., the typical audit room). This general context-behavior reaction time approach has been used in other psychology studies. For example, Neal et al. (2012) assess the speed of participants’ recognition of words reflecting running after viewing words reflecting the typical context in which they run (e.g., “gym”). Adriaanse et al. (2011) assess the speed of participants’ recognition of words reflecting habitual snacks after being primed with “home” and of habitual drinks after being primed with “bar.” Similar implicit measures have been used to capture habits for transportation (Danner et al. 2008) and hand hygiene (Hargadon 2017).We implement our measurement procedures as follows. Near the end of the study, we prime (i.e., activate in memory, see, e.g., Doyen, Klein, Simons, and Cleerman 2014) the typical context by having all auditors complete a one-minute “spot the difference” exercise involving two photographs: the typical room and the same room with five small differences created using Photoshop. We then collect reaction times by having auditors complete word fragments that are reflective of the simple processes while the typical context remains activated (the photograph of the typical room remains displayed). We instruct auditors to type the entire word represented by each fragment as quickly as possible (see Figure 1, Panel A for instructions), tracking completion time in milliseconds. The simple process words (e.g., scan, glance) are shown in Figure 1, Panel B, and were chosen based on the previously mentioned interviews with four seniors. Auditors for whom the simple processes are strongly habitual (assuming those habits were developed in the typical context) should, ceteris paribus, more quickly complete the related word fragments. However, because reaction times also can be affected by idiosyncratic factors such as typing or reading speed, we subtract reaction times to non-audit control words (as in Neal et al. 2012). The control words (e.g., bark; see Panel B) are roughly matched to the simple process words on syllables and length. We classify auditors with faster (slower) completion times relative to the median as having stronger (weaker) simple process habits.Audit Room Context Our second independent variable is audit room context. We manipulate context by asking auditors to imagine they are at a client site when viewing a picture of either a typical audit room or an alternative room that removes many elements of the former context (see Panels A and B of Figure 2). Use of a photo to manipulate context is consistent with psychology research (e.g., Neal et al. 2012; Weinstein, Przybylski, and Ryan 2009). As discussed earlier, the typical room contains common elements auditors encounter (e.g., chairs and equipment for multiple team members). The alternative room omits several cues while maintaining a realistic working environment (see Panel B). We staged the two audit rooms in a conference room of a participating firm’s office, and a research assistant photographed the rooms. After viewing the pictures, auditors in the typical (alternative) context condition are given the instructions shown in Panel C (D) of Figure 2. They are asked to “describe what you have imagined in 5-7 sentences,” to activate the context in memory and enable spreading activation to related concepts (here, cognitive processes) and also to make them feel like they are in the room (MacInnis and Price 1987). The probability with which simple process habits are activated when completing the estimates task should (should not) vary with habit strength in the typical (alternative) context.Task, Dependent Variables, and Other Measures Estimates Task The task requires that auditors evaluate management’s assumptions underlying an estimate of goodwill as part of the client’s step-one analysis of an impairment test and is adapted from Kadous and Zhou (2019). The task includes background information, analysis, and evidence related to management’s assumptions. The client uses a discounted cash flow model to estimate the fair value of the reporting unit, which indicates that it passes the impairment test. The task includes sections for three key assumptions: five-year projections of revenue, operating expenses, and capital expenditures. We embed seven issues that, as in the real-world task, are less likely to be identified if auditors use simple cognitive processes (see the Appendix). Dependent Variable and Other MeasuresBecause our focus is on how habit strength affects cognitive processing in the estimates task, our dependent variable is the number of embedded issues a participant identifies (Issues Identified). We ask participants to “list any specific concerns” they have about the estimate based on their evaluation. Auditors also complete a number of other post-experimental questions to capture determinants of habit strength, potential moderators, and potential noise variables. IV. RESULTSTests of HypothesesValues for Issues Identified range from zero to six of the seven embedded issues. Descriptive statistics are tabulated in Table 1, Panel A. We test hypotheses using an Analysis of Variance (ANOVA) model with Issues Identified as the dependent variable and independent variables indicating whether the participant’s Simple Process Habit Strength is stronger or weaker and whether the assigned Context is typical or alternative (see Panel B).H1 predicts that, within the typical context, auditors with stronger simple process habits will identify fewer issues than auditors with weaker habits. Simple effects analyses are displayed in Panel C of Table 1. In support of H1, there is a negative effect of Simple Process Habit Strength on Issues Identified (one-tailed p = 0.014) in the typical context. In this context, seniors with stronger habits identify fewer issues than those with weaker habits (means = 1.09 vs. 1.91). Consistent with research indicating that strong habits are particularly “hard to break” (Wood and Neal 2016), an additional analysis (untabulated) reveals that months of experience does not reduce the habit strength effect (p > 0.500). Thus, while, experience could in theory reduce the negative effects of habits (e.g., through knowledge acquisition), our findings suggest otherwise.H2 predicts that auditors with stronger habits will identify more issues when in the alternative versus typical context, while auditors with weaker habits are less likely to be positively influenced by context. In support of H2, the interaction between Simple Process Habit Strength and Context is significant (two-tailed p = 0.011) (see Panel B). As shown in Panel C, among auditors classified as having high Simple Process Habit Strength, there is a positive simple effect of Context on Issues Identified (one-tailed p = 0.024) (means = 1.09 and 1.83 for the typical and alternative contexts, respectively). Performance of stronger simple process habits auditors in the alternative context is indistinguishable from that of auditors with weaker simple process habits in the typical context (untabulated, two-tailed p = 0.868). Turning to auditors with weaker habits, we observe a marginally significant negative simple effect of Context (two-tailed p = 0.096) (means = 1.91 vs. 1.28), which may suggest their use of complex process habits.Additional AnalysesIn this section, we first provide further evidence about the validity of our habit strength measure (as suggested by Rebar et al. 2018). Following, we explore implications of our finding that weaker habits auditors identify marginally fewer issues in the alternative context. Finally, we present robustness tests to show that effects of context are not driven by features of the typical vs. alternative context per se, but are best explained by theory on habits. Validation of Habit Strength MeasureDeterminants of simple process habit strength. Our theory proposes that auditors build simple process habits as staff, but these habits vary in strength. We use a structural equations model to examine determinants of habit strength (see Figure 3). Theory predicts that habit strength increases as a function of repetition of behaviors and rewards felt from such repetition in a stable context (Mazar and Wood 2018). We measure Simple Process Repetition with four questions capturing seniors’ use of approaches to simple tasks, as staff, that reflect use of the simple processes on a scale from 1 “Never Used This Approach” to 9 “Always Used This Approach”. These questions are akin to those used in psychology research to capture habit strength using frequency of repetition of past behavior (e.g., Danner et al. 2008; Neal et al. 2011; Neal et al. 2012). We measure auditors’ tendency to seek their team’s approval as a proxy for Rewards. Staff valuing approval more would likely feel more rewarded by positive reinforcement from superiors for use of the simple processes. We elicit agreement (ranging from 1 “Strongly Disagree” to 7 “Strongly Agree”) that, as staff making day-to-day decisions on their engagements, they generally chose the option they thought their audit team would approve of. We also measure antecedents of Simple Process Repetition. These include auditors’ exposure to simple tasks as staff, which creates opportunities for using the simple processes, and personal effectiveness preferences, which could lead staff to use complex processes even for simple tasks. We measure Simple Task Exposure as the percentage of time that auditors report having worked on simple versus complex audit tasks as staff. We measure professional identity as a proxy for Effectiveness Preference. Prior research finds that stronger professional identity predicts a focus on audit effectiveness (Bamber and Iyer 2007; Bauer 2015), consistent with effective audit task performance being an integral part of the self-concept (Britt 1999, 2005). Standard measures support a good model fit (χ2(22) = 24.57, p = 0.318; CFI = 1.00; RMSEA = 0.03). The four indicators of Simple Process Repetition discussed above load on one latent factor (largest one-tailed p = 0.047). As expected, Simple Process Repetition positively affects Simple Process Habit Strength (one-tailed p = 0.029), as does Rewards (one-tailed p = 0.003). Turning to antecedents of Simple Process Repetition, as expected, there is a positive effect of Simple Task Exposure (one-tailed p = 0.027); completing more simple tasks leads to greater repetition of simple processes. However, this effect is moderated by a negative Simple Task Exposure X Effectiveness Preference interaction (one-tailed p = 0.033), indicating that this relationship is less pronounced for auditors with higher effectiveness preferences. That these theoretical determinants predict our habit strength measure validates the measure. The findings also reinforce a key benefit of our measure of habit strength: the memory association reflects effects of frequency of repetition of behavior and experienced rewards. Moderation of context effects for strong habits auditors who also develop habits in the alternative context. Because the audit environment provides ample opportunities for rewarded repetition of behaviors for habit development, we assume that auditors develop habits in the typical context. However, some may also develop habits in our alternative context if they spend enough time working in that context. While a change in context can prevent habits from being activated, this is only true if the person has not also developed habits in the alternative context (Verplanken et al. 2018). Thus, if auditors with stronger simple process habits also have developed their habits in the alternative context, the alternative context is unlikely to improve performance. In that case, the alternative context will activate the simple processes just as the typical context does. Table 2, Panel A reports results of a regression (in the full sample) with Context, Simple Process Habit Strength, and Alternative Context Experience as independent variables, and Issues Identified as the dependent variable. We measure Alternative Context Experience using auditors’ time spent working in contexts like our alternative context. The three-way interaction is significant (two-tailed p = 0.048), and we explore the results for weaker habits auditors in a subsequent section. Consistent with theory, regression estimates (see Panel B) reveal a positive effect of Context for stronger habits auditors with low and medium Alternative Context Experience (largest one-tailed p = 0.049). However, this effect is insignificant at high Alternative Context Experience (one-tailed p = 0.432), consistent with the idea that auditors with stronger simple process habits who also developed their habits in the alternative context have the simple processes activated in both contexts. Moderation of context effects for stronger habits auditors who exert self-discipline. Because inhibiting strong habits is effortful (Wood 2016, 2017), we assume that auditors with stronger habits use the automatically activated simple processes unless the alternative context prevents their activation. However, if auditors with stronger simple process habits effortfully inhibit the simple processes with self-discipline, the alternative context will have less opportunity to improve performance. Table 3, Panel A reports results of a regression (in the full sample) with Context, Simple Process Habit Strength, and Self-Discipline as independent variables, and Issues Identified as the dependent variable. Self-Discipline is auditors’ agreement on a 7-point scale that they exerted self-discipline while working on the estimates task. Among auditors with stronger habits, a regression reveals the expected Context X Self-Discipline interaction (two-tailed p = 0.045; Panel B). Consistent with theory, estimates from the regression (Panel C) reveal a positive effect of Context for stronger habits auditors exerting low and medium Self-Discipline (largest one-tailed p = 0.018). However, there is an insignificant effect at high Self-Discipline (one-tailed p = 0.497), consistent with the idea that auditors with stronger habits who effortfully inhibit their habits have less need for the alternative context.Exploring the Nature of Processing Among Weaker Habits AuditorsIn the development of H2, we considered the possibility that at least some auditors with weaker simple process habits may have complex process habits. We consider this possibility further in this section. First, our analysis of determinants of simple process habits (see Figure 3) supports that at least some staff had opportunities to develop complex process habits. To the extent that complex processes are the opposite of the four simple process indicators that we measured, this analysis indicates the following. Seniors who, as staff, engaged in more repetition of complex processes developed weaker simple process habits. This finding suggests that seniors could have developed complex process habits as staff due to performing more complex tasks. The related finding that staff with high effectiveness preferences tended to use complex processes even on simple tasks implies that these staff felt rewarded from repeating the complex processes. Next, we examine whether our findings suggest these seniors are using complex process habits (versus intentional processing) on the estimates task. If weaker habits auditors are using complex process habits, we expect them to perform well in the typical context because their complex process habits are activated and are effective for the task. However, we expect they may have difficulty in the alternative context because these habits are not activated and they may need to regroup unless, as discussed earlier, they have formed habits in both contexts. Consistent with our expectations, estimates from a regression model (Table 2, Panel C) show that the effect of Context is significant and negative for weaker habits auditors with low and medium Alternative Context Experience (two-tailed p’s of 0.017 and 0.067), consistent with complex process habits not being activated in the alternative context. The insignificant effect at high Alternative Context Experience (p > 0.500) is consistent with these auditors having complex process habits activated in both contexts. Overall, the findings support complex process habits being at play, especially considering that intentional processing would lead to similar performance across contexts and irrespective of experience in the alternative context. However, future research is needed to more directly study such habits.Robustness Checks for Differences Across the Typical and Alternative Contexts In this section we present analyses supporting that the observed effects of context are not driven by differing features of the typical vs. alternative context per se. Table 4, Panel A reports auditors’ thoughts about the context, coded from the paragraphs they wrote during the manipulation. Chi-square analyses (see Panel B) indicate that six thoughts differ across contexts. Auditors in the typical (vs. alternative) context are more likely to think about: coaching or supervising staff, feeling stressed (i.e., due to clutter or being cramped), interruptions, time pressure, busy season, and a personal recollection (e.g., from their audit experience). For the last two thoughts, only six and seven auditors showed evidence of them. For these two, we reperform hypotheses tests within two subsamples that exclude these six and seven auditors, respectively. Results are significant at the same critical levels, except that the decline in performance for weaker habits auditors falls below significance; however, additional analysis reveals a significant decline for those with less Alternative Context Experience.We examine the robustness of our results to the other four thoughts in three ways. First, we include them as covariates in the ANOVAs used to test hypotheses; results are significant at the same critical levels, except that the decline for weaker habits auditors is significant only for those with less Alternative Context Experience. Second, in lieu of using Context as an independent variable, we use the context-related thoughts. These thoughts do not interact with Simple Process Habit Strength in the form of our observed interaction. Third, assuming the thoughts invoke the simple processes, to explain our interaction, stronger habits auditors would need to show fewer of the thoughts in the alternative (vs. typical) context, while weaker habits auditors would not. ANOVAs using Simple Process Habit Strength and Context and their interaction, with the thoughts as dependent variables, show only a main effect of Context. V. DISCUSSION AND CONCLUSIONSThis paper examines simple process habit strength as a potential reason why some audit seniors use simple processes on complex tasks, including audits of estimates, despite that those processes are inappropriate for such tasks. Using a reaction-time measure of habit strength and manipulating imagined context, we show that, in the typical audit room context, seniors with stronger simple process habits identify fewer issues in a complex estimate than do seniors with weaker habits. Moreover, seniors with stronger habits who are placed in an alternative context identify more issues, while auditors with weaker habits identify marginally fewer issues in that context. The observed interaction is consistent with theory and supports that habit strength is the causal construct. However, we strengthen this inference by reporting evidence of theory-consistent determinants of habit strength and moderators of context effects. Finally, we provide evidence that auditors with weaker simple process habits may be using complex process habits. Prior research and PCAOB inspection reports have attributed difficulties in audits of complex accounts to auditors’ use of simple processes. Our study provides evidence that the strength of simple process habits may be a root cause. Habits are “hardwired” links between the context and processes in auditors’ memory. Because the context in which seniors work does not change, seniors with strong simple process habits may use the simple processes for complex tasks in spite of firm training and decision aids, changes in professional standards, and recurring audit deficiencies. For example, the requirement in both judgment frameworks and the standard on auditing estimates (PCAOB 2019) to search for contradictory evidence likely would be ineffective for seniors with strong habits; although they likely would comply, their habits may lead them to find ways to explain away such information (e.g., Verplanken 2018). Instead, psychology theory suggests that, to be effective, interventions must either alter context or be targeted toward breaking or overcoming the habits. These strategies present difficulties, such that firms may find preventing strong simple process habits from forming among staff to be more promising. Alternatively, a more proactive approach could be helping staff develop complex process habits. Accounting firms are discovering the power of building positive habits for their employees’ health and well-being (PwC 2020). They may be able to use similar approaches to build cognitive processing habits among staff that will be appropriate when they are promoted and performing senior-level tasks. Cultivating complex process habits could be highly beneficial: if seniors have carried forward complex process habits, this would free up bandwidth for using intentional processes on other, new tasks, such as coaching and reviewing. Our analysis of determinants suggests that staff may use more complex processes when performing more complex tasks, such that “pushing down” more complex tasks could be beneficial. Firms could further leverage the habits literature for additional insight into how they can facilitate formation of complex process habits. For example, firms could speed up formation of complex process habits by innovating ways for staff to find repetition of those processes enjoyable and rewarding (e.g., Gardner and Lally 2013). They could reduce frictions that make complex processing difficult to prevent backsliding (Lally and Gardner 2013), perhaps by encouraging complex tasks to be performed when there are few distractions. Finally, firms can vividly highlight valued outcomes of using effective processing on complex tasks to build satisfaction and encourage repetition of complex processes (Lally and Gardner 2013). Complex processes also could be made more rewarding through thoughtful design of firm technology such as apps (Carden and Wood 2018), and by making staff feel that they have autonomy and self-direction in choosing processes (Lally and Gardner 2013). Finally, firms may be able to facilitate auditors’ recognition of simple versus complex tasks and the appropriate processes for each. With this approach, auditors could build memory associations between type of task and processes, such that task instead of context would activate processing. Research with tax professionals indicates that at least some appear to distinguish between simple and complex tasks and use the appropriate cognitive processing (Magro 1999, 2005). However, future research is needed to determine how this can best be accomplished. Our study contributes to the auditing literature by demonstrating the importance of habit strength to auditor performance of a critical task. Our work also may contribute to the literature on auditor skepticism (see Nelson 2009; Nolder and Kadous 2018) by providing an alternative view that problems attributed to lack of skepticism also could arise from habits. Also, while we focus on audits of complex estimates, we expect that simple process habits impair auditor performance of other audit tasks that require complex processing as well. Our work also provides a methodology for measuring cognitive processing habits. Given the ubiquitous nature of habits, we expect researchers in auditing and other areas of accounting could adapt and use our measure. Our study has limitations that offer ideas for future research. Our findings suggest that some seniors with weaker simple process habits may have complex process habits, but this conclusion relies on the assumption that complex processes are the alternative to simple processes. Future research can examine complex process habits directly. Future research also can examine whether seniors develop habits for other behaviors, such as reviewing staff’s work, habits that also could mitigate the negative effects of depletion (Mullis and Hatfield 2018). ReferencesAdriaanse, M., P. Gollwitzer, D. De Ridder, J. Wit, and F. Kroese. 2011. Breaking Habits With Implementation Intentions: A Test of Underlying Processes. Personality and Social Psychology Bulletin 37(4): 502-513. Anderson, N. 1981. Foundation of Information Integration Theory. New York, NY: Academic Press. Austin, A., J. Hammersley, and M. Ricci. 2019. Improving Auditors’ Consideration of Evidence Contradicting Management’s Estimate Assumptions. Contemporary Accounting Research, Forthcoming. Bagley, P. 2010. Negative Affect: A Consequence of Multiple Accountabilities in Auditing. AUDITING: A Journal of Practice and Theory 29(2): 141-157. Bamber, E. and V. Iyer. 2007. Auditors’ Identification with their Clients and its Effect on Auditors’ Objectivity. AUDITING: A Journal of Practice and Theory 26(2): 1-24. Bauer, T. 2015. The Effects of Client Identity Strength and Professional Identity Salience on Auditor Judgments. The Accounting Review 90: 95-114. Bierstaker, J. and A. Wright. 2001. The Effects of Fee Pressure and Partner Pressure on Audit Planning Decisions. Advances in Accounting 18:25-46. Bierstaker, J. and A. Wright. 2005. The Effect of Partner Preferences on the Development of Risk-Adjusted Program Plans. Advances in Accounting 21: 1-23.Bobek, D., B. Daugherty, and R. Radtke. 2012. Resolving Audit Engagement Challenges Through Communication. AUDITING: A Journal of Practice and Theory 31(4): 21-45. Bonner, S. 1994. A Model of the Effects of Audit Task Complexity. Accounting, Organizations and Society 19: 213-234. Bratten, B., L. Gaynor, L. McDaniel, N. Montague, and G. Sierra. 2013. The Audit of Fair Values and Other Estimates: The Effects of Underlying Environmental, Task, and Auditor-Specific Factors. AUDITING: A Journal of Practice and Theory 32: 7-44. Britt, T. 1999. Engaging the Self in the Field: Testing the Triangle Model of Responsibility. Personality and Social Psychology Bulletin 25: 698-708. Britt, T. 2005. The Effects of Identity-Relevance and Task Difficulty on Task Motivation, Stress, and Performance. Motivation and Emotion 29: 189-202. Brown, V., J. Gissel, and D. Neely. 2016. Audit Quality Indicators: Perceptions of Junior-Level Auditors. Managerial Auditing Journal 31(8-9): 949-980. Cannon, N. and J. Bedard. 2017. Auditing Challenging Fair Value Measurements: Evidence from the Field. The Accounting Review 92: 81-114. Carden, L. and W. Wood. 2018. Habit Formation and Change. Current Opinion in Behavioral Sciences 20: 117-122. Collins, A. and E. Loftus. 1975. A Spreading-Activation Theory of Semantic Processing.Psychological Review 82: 407-428. Craik, F. and R. Lockhart. 1972. Levels of Processing: A Framework for Memory Research. Journal of Verbal Learning and Verbal Behavior 11: 671-684. Craik, F. 2002. Levels of Processing: Past, Present…and Future? Memory 10: 305-318. Danner, U., H. Aarts, and N. de Vries. 2008. Habit vs. Intention in the Prediction of Future Behaviour: The Role of Frequency, Context Stability, and Mental Accessibility of Past Behaviour. British Journal of Social Psychology 47(2): 245-265. De Houwer, J., S. Teige-Mocigemba, A. Spruyt, and A. Moors. 2009. Implicit Measures: A Normative Analysis and Review. Psychological Bulletin 135(3): 347-368. Deese, J. 1960. Frequency of Usage and Number of Words in Free Recall: The Role of Association. Psychological Reports 7: 337-344. Doyen, S., O. Klein, D. Simons, and A. Cleermans. 2014. On the Other Side of the Mirror: Priming in Cognitive and Social Psychology. Social Cognition 32: 12-32.Evans, J. and K. Stanovich. 2013. Dual-Process Theories of Higher Cognition: Advancing the Debate. Perspectives on Psychological Science 8(3): 223-241. Fournier, M., F. d’Arripe-Longueville, and R. Radel. 2017. Testing the Effect of Text Messaging Cues to Promote Physical Activity Habits: A Worksite-Based Exploratory Intervention. Scandinavian Journal of Medicine and Science in Sports 27(10): 1157-1165. Gardner, B. and P. Lally. 2013. Does Intrinsic Motivation Strengthen Physical Activity Habit? Modeling Relationships Between Self-Determination, Past Behaviour, and Habit Strength. 2013. Journal of Behavioral Medicine 36: 488-497. Gardner, B. and V. Tang. 2014. Reflecting on Non-Reflective Action: An Exploratory Think-Aloud Study of Self-Report Habit Measures. British Journal of Health Psychology 19(2): 258-273.Gardner, B. 2015. A Review and Analysis of the Use of “Habit” in Understanding, Predicting, and Influencing Health-Related Behavior. Health Psychology Review 9(3): 277-295. Gardner, B. and P. Lally. 2018. “Modelling Habit Formation and Its Determinants,” in B. Verplanken (Ed.) The Psychology of Habit: Theory, Mechanisms, Change, and Contexts. Cham, Switzerland: Springer Publishing. Griffith, E., J. Hammersley, and K. Kadous. 2015a. Audits of Complex Estimates as Verification of Management Numbers: How Institutional Pressures Shape Practice.Contemporary Accounting Research 32: 833-863. Griffith, E., J. Hammersley, K. Kadous, and D. Young. 2015b. Auditor Mindsets and Audits of Complex Estimates. Journal of Accounting Research 53: 49-77.Griffith, E. 2018. When Do Auditors Use Specialists’ Work to Improve Problem Representations of and Judgments about Complex Estimates? The Accounting Review 93: 177-202. Hagger, M., A. Rebar, B. Mullan, O. Lipp, and N. Chatzisarantis. 2015. The Subjective Experience of Habit Captured by Self-Report Indexes May Lead to Inaccuracies in the Measurement of Habitual Action. Health Psychology Review 9(3): 296-302. Hargadon, D. 2017. Developing an Implicit Measure of Habit Strength: The Habit IAT. Kingston, ON: Queen’s University. Houston, R. 1999. The Effects of Fee Pressure and Client Risk on Audit Seniors’ Time Budget Decisions. AUDITING: A Journal of Practice and Theory 18(2): 70-86. Hurley, P. 2015. Ego Depletion: Applications and Implications for Auditing Research. Journal of Accounting Literature 35: 47-76. IFIAR (International Forum of Independent Audit Regulators). 2017. Survey of Inspection Findings. Available at: IFIAR. 2019. Survey of Inspection Findings. Available at: Itzchakov, G., L. Uziel, and W. Wood. 2018. When Attitudes and Habits Don’t Correspond: Self-Control Depletion Increases Persuasion But Not Behavior. Journal of Experimental Social Psychology 75: 1-10. Joe, J., Y. Wu, and A. Zimmerman. 2017. Overcoming Communication Challenges: Can Taking the Specialist’s Perspective Improve Auditors’ Critical Evaluation and Integration of the Specialist’s Work? Working Paper. Kadous, K. and Y. (D.) Zhou. 2019. How Does Intrinsic Motivation Improve Auditor Judgment in Complex Audit Tasks? Contemporary Accounting Research. 36: 108-131.Kelley, T. 1984. Time-Budget Pressures and Auditor Behavior. Houston, TX: University of Houston. Kunda, Z. 1990. The Case for Motivated Reasoning. Psychology Bulletin 108: 480-498. Labrecque, J. and W. Wood. 2015. What Measures of Habit Strength to Use? Comment on Gardner (2015). Health Psychological Review 9: 303-310. Lally, P., C. van Jaarsveld, H. Potts, and J. Wardle. 2010. How Habits are Formed: Modelling Habit Formation in the Real World. European Journal of Social Psychology 40(6): 998-1009.Lally, P. and B. Gardner. 2013. Promoting Habit Formation. Health Psychology Review 7(1): 137-158. Lin, P., W. Wood, and J. Monterosso. 2016. Healthy Eating Habits Protect Against Temptations. Appetite 103(1): 432-440. MacInnis, D. And L. Price. 1987. The Role of Imagery in Information Processing: Review and Extensions. Journal of Consumer Research 13: 473-491. Magro, A. 1999. Contextual Features of Tax Decision-Making Settings. The Journal of the American Taxation Association 21: 63-73. Magro, A. 2005. Knowledge, Adaptivity, and Performance in Tax Research. The Accounting Review 80: 703-722. Malone, C. and R. Roberts. 1996. Factors Associated with the Incidence of Reduced Audit Quality Behaviors. AUDITING: A Journal of Practice and Theory 15(2): 49-64. Mazar, A. and W. Wood. 2018. “Defining Habit in Psychology,” in B. Verplanken (Ed.) The Psychology of Habit: Theory, Mechanisms, Change, and Contexts. Cham, Switzerland: Springer Publishing. McNamara, T. 1992. Theories of Priming: I. Associative Distance and Lag. Journal of Experimental Psychology: Learning, Memory, and Cognition 18: 1173-1190. Mullis, C. and R. Hatfield. 2018. The Effects of Multi-tasking on Auditors’ Judgment Quality. Contemporary Accounting Research 35(1): 314-333. Neal, D., W. Wood, M. Wu, and D. Kurlander. 2011. The Pull of the Past: When Do Habits Persist Despite Conflict with Motives? Personality and Social Psychology Bulletin 37: 1428-1437. Neal, D., W. Wood, J. Labrecque, and P. Lally. 2012. How Do Habits Guide Behavior? Perceived and Actual Triggers of Habit in Daily Life. Journal of Experimental Social Psychology 48: 492-498. Nelson, M. 2009. A Model and Literature Review of Professional Skepticism in Auditing. AUDITING: A Journal of Practice and Theory 28: 1-34. Nelson, M. and C. Proell. 2018. Is Silence Golden? Audit Team Leader Reactions to Subordinates Who Speak Up “In the Moment” and at Performance Appraisal. The Accounting Review 93: 281-300. Nickerson, R. 1998. Confirmation Bias: A Ubiquitous Phenomenon in Many Guises. Review of General Psychology 2: 175-220. Nolder, C. and K. Kadous. 2018. Grounding the Professional Skepticism Construct in Mindset and Attitude Theory: A Way Forward. Accounting, Organizations and Society 67: 1-14. Potthoff, S., N. McCleary, F. Sniehotta, and J. Presseau. 2018. “Creating and Breaking Habits in Healthcare – Professional Behaviors to Improve Healthcare and Health,” in B. Verplanken (Ed.) The Psychology of Habit: Theory, Mechanisms, Change, and Contexts. Cham, Switzerland: Springer Publishing. Power, A. 2003. Auditing and the Production of Legitimacy. Accounting, Organizations and Society 28: 379-394. PricewaterhouseCoopers (PwC). 2020. Be Well, Work Well Habit Bank. Available at: Company Accounting Oversight Board (PCAOB). 2019. AS 2501: Auditing Accounting Estimates. Available at: , A., B. Gardner, R. Rhodes, and B. Verplanken. 2018. “The Measurement of Habit,” in B. Verplanken (Ed.) The Psychology of Habit: Theory, Mechanisms, Change, and Contexts. Cham, Switzerland: Springer Publishing. Verplanken, B. and S. Orbell. 2003. Reflections on Past Behavior: A Self-Report Index of Habit Strength. Journal of Applied Social Psychology 33(6): 1313-1330. Verplanken, B., O. Friborg, C. Wang, D. Trafimow, and K. Woolf. 2007. Mental Habits: Metacognitive Reflection on Negative Self-Thinking. Journal of Personality and Social Psychology 92(3): 526-541.Verplanken, B. and Y. Tangelder. 2011. No Body is Perfect: The Significance of Habitual Negative Thinking About Appearance for Body Dissatisfaction, Eating Disorder Propensity, Self-Esteem, and Snacking Psychological Science 6: 685-701.Verplanken, B. and N. Fisher. 2014. Habitual Worrying and Benefits of Mindfulness. Mindfulness 5: 566-573.Verplanken, B. and W. Wood. 2016. Interventions to Break and Create Consumer Habits. Journal of Public Policy and Marketing 25(1): 90-103. Verplanken, B. 2018. “Introduction,” in B. Verplanken (Ed.) The Psychology of Habit: Theory, Mechanisms, Change, and Contexts. Cham, Switzerland: Springer Publishing. Verplanken, B., D. Roy, and L. Whitmarsh. 2018. “Cracks in the Wall: Habit Discontinuities as Vehicles for Behaviour Change,” in B. Verplanken (Ed.) The Psychology of Habit: Theory, Mechanisms, Change, and Contexts. Cham, Switzerland: Springer Publishing. Weinstein, N., A. Przybylski, and R. Ryan. 2009. Can Nature Make Us More Caring? Effects of Immersion in Nature on Intrinsic Aspirations and Generosity. Personality and Social Psychology Bulletin 35: 1315-1329. Westermann, K., J. Bedard, and C. Earley. 2015. Learning the ‘Craft’ of Auditing: A Dynamic View of Auditors’ On-the-Job Learning. Contemporary Accounting Research 32: 864-896. Westermann, K., J. Cohen, and G. Trompeter. 2019. PCAOB Inspections: Public Accounting Firms on “Trial.” Contemporary Accounting Research 36: 694-731. Whelan, R. 2008. Effective Analysis of Reaction Time Data. Psychological Record 58: 475-482. Wood, W., J. Quinn, and D. Kashy. 2002. Habits in Everyday Life: Thought, Emotion, and Action. Journal of Personality and Social Psychology 83: 1281-1297. Wood, W., L. Tam, and M. Witt. 2005. Changing Circumstances, Disrupting Habits. Journal of Personality and Social Psychology: Attitudes and Social Cognition 88: 918-933. Wood, W. and D. Neal. 2007. A New Look at Habits and the Habit-Goal Interface. Psychological Review 114: 843-863. Wood, W. and D. Neal. 2009. The Habitual Consumer. Journal of Consumer Psychology 19: 579-592. Wood, W. 2016. “The Role of Habits in Self-Control,” in K. Vohs and R. Baumeister (Eds.) Handbook of Self-Regulation. New York, NY: The Guilford Press. Wood, W. and D. Neal. 2016. Healthy Through Habit: Interventions for Initiating and Maintaining Health Behavior Change. Behavioral Science and Policy 2(1): 71-83. Wood, W. and D. Rünger. 2016. Psychology of Habit. Annual Review of Psychology 67: 289-314. Wood, W. 2017. Habits in Personality and Social Psychology. Personality and Social Psychology Review 21: 389-403. Wood, W. 2019. Good Habits, Bad Habits: The Science of Making Positive Changes That Stick. New York, NY: Farrar, Straus and Giroux. FIGURE 1 Simple Process Habit Strength Measure Panel A: Instructions: On the following screens, you will see a series of word fragments. The number of blanks indicates the number of missing letters.For example, if there is one blank visible, this means there is one letter missing. As another example, if there are three blanks visible, this means there are three letters missing.Please type the word (the entire word, not just the missing letters)?as quickly as possible?and then hit the arrow to continue to the next word fragment. As an example to get you started, view the two word fragments below:?WO __ D __ __ AGMEN __ As soon as you figure out the word,?you would type the word in the textbox below the word fragment.?For example, once you realized the first word was “WORD,”?you would type?“WORD”?in the textbox.?As another example, once you realized the second word was "FRAGMENT," you would type "FRAGMENT" in the textbox.??Please remember to type in the word and hit the arrow button?as quickly as possible.?Please click the arrow below to begin.Panel B: Control Word FRAGMENTSSTAFF process WORD FRAGMENTSB __ RK (Bark)__ HOC __ __ ATE (CHOCOLATE) FANC __ (FANCY )MAGAZ __ __ ES (MAGAZINES) MA __ SHMA __ __ OW (MARSHMALLOW)PAINT __ __ USH (PAINTBRUSH) __ __ TERMELON (WATERMELON) CHE __ __ LIST (CHECKLIST) GLAN __ E (GLANCE) INSPE __ __ (INSPECT) PROC __ __ D (PROCEED) __ __ ASONABLE (REASONABLE) SC __ N (SCAN) __ KIM (SKIM) SPREA __ __ __ EET (SPREADSHEET) VE __ __ FY (VERIFY) Panel A displays the instructions that participants read prior to completing the word fragments. Panel B displays the word fragments that we use to measure Simple Process Habit Strength. We calculate the average of each auditor’s completion times for the simple process fragments minus the average of each auditor’s completion times for the control, baseline (i.e., non-audit related) word fragments. We transform each reaction time by taking the reciprocal, which is a transformation commonly used in psychology studies using reaction-time measures to adjust for skewness (see Whelan 2008). We then classify auditors with completion times faster (slower) than the median as having stronger (weaker) simple process habits. FIGURE 2 Audit Room Context Manipulation Panel A: Typical Audit Room Context Panel B: Alternative Audit Room Context Panel C: Instructions Read by Participants in the Typical Audit Room Context: Please imagine that you are working in this audit room today. You are working in here all day, and your intern and three staff are also all here today.Look around and take in the room. Imagine how the day would progress as you are working in this room. For example: Imagine yourself sitting in the chair (yours is the gray one to the far right) and your staff and intern sitting in the other chairs.?Imagine yourself using your laptop and other tools/supplies as you are doing your audit work.?Imagine what might be happening throughout the day as you do your work in this room. Please describe what you have imagined in 5-7 sentences in the box below. Panel D: Instructions Read by Participants in the Alternative Audit Room Context: Please imagine that you are working in this audit room today. You are working by yourself in here all day, as your intern and three staff are all at a full day training in the local office today.Look around and take in the room. Imagine how the day would progress as you are working in this room. For example: Imagine yourself sitting in the chair.?Imagine yourself using your laptop and other tools/supplies as you are doing your audit work.?Imagine what might be happening throughout the day as you do your work in this room. Please describe what you have imagined in 5-7 sentences in the box below. Figure 2 provides details of the manipulation of (imagined) context. Auditors assigned to the typical (alternative) audit room context view the picture of the audit room in Panel A (Panel B), and read the instructions displayed in Panel C (and Panel D). After reading through the instructions, auditors then write a paragraph imagining themselves working in the context in the picture. FIGURE 3 Determinants of Simple Process Habit Strength2143760-635Simple Task Exposure Simple Task Exposure 37185609525Effectiveness Preference Effectiveness Preference 52933609525Simple Task Exposure X Effectiveness Preference Simple Task Exposure X Effectiveness Preference 420560590360500203203616325Simple Process Indicator 4Simple Process Indicator 4091440Simple Process Indicator 10Simple Process Indicator 1244856013716000564896096520? = - 0.01 p = 0.03300? = - 0.01 p = 0.0334175760116840? = + 0.60 p = 0.03500? = + 0.60 p = 0.0352651760118745? = + 0.05 p = 0.02700? = + 0.05 p = 0.027144272052070? = + 0.96 p = 0.04000? = + 0.96 p = 0.04047955193175000114808013970002032040640Simple Process Indicator 20Simple Process Indicator 21320800120650? = + 1.03 p = 0.02900? = + 1.03 p = 0.029313944031750Simple Process Repetition0Simple Process Repetition1229360787400020320135890Simple Process Indicator 30Simple Process Indicator 31341120132080? = + 1.49 p = 0.04700? = + 1.49 p = 0.0471493520140970005283200110490Rewards Rewards 4165600133985334264062865? = + 0.17 p = 0.02900? = + 0.17 p = 0.02912903205270500134112089535Fixed at 100Fixed at 1358229643666Simple Process Habit Strength 0Simple Process Habit Strength 48158409842500521208053975? = + 0.09 p = 0.00300? = + 0.09 p = 0.003The above structural equations model examines determinants of auditors’ simple process habit strength. The chi-squared test for this model reveals good fit (χ2(22) = 24.57, p = 0.318), as do other standard measures (CFI = 1.00; RMSEA = 0.03). Figure 2 defines Simple Process Habit Strength. We measure Simple Task Exposure using auditors’ self-reported percentage of time (spent as a staff working on audit procedures) that they worked on simple (versus complex) audit procedures (from 0 to 100 percent, in increments of 10). We measure Effectiveness Preference using the professional identity measure from Bauer (2015). We measure Simple Process Repetition with four questions that elicit the extent to which (when the seniors were staff and performing simple tasks), they used approaches that are indicative of the simple processes (on a scale from 1 “Never Used This Approach” to 9 “Always Used This Approach.”). Simple Process Indicator 1 is “When reviewing terms of a transaction or item, closely examined the details (e.g., checked additional details outside of just the key terms” (reverse-scored). Simple Process Indicator 2 is “When performing an audit procedure, considered how my findings within this procedure related to each other (e.g., if there were issues with multiple invoices in the sample, I thought about whether the same problem could be causing the issues)” (reverse-scored). Simple Process Indicator 3 is “When checking invoice terms, searched to see that there was evidence that agreed to what I was looking for (e.g., if I was looking for a dollar amount of ‘38,’ looked on the invoice to see if there was a ‘38’ anywhere.” Simple Process Indicator 4 is “When evaluating the findings of audit procedures, considered results from each procedure separately (i.e., did not think about connections among results).” We measure Rewards using auditors’ agreement (on a scale from 1 “Strongly Disagree” to 7 “Strongly Agree”) that, when staff, and making day-to-day decisions on their engagements, they generally chose the option that they thought their audit team members would approve of. P-values are one-tailed for directional predictions.TABLE 1: Auditor Performance (Issues Identified) by Simple Process Habit Strength and ContextPanel A: Descriptive Statistics Simple Process Habit StrengthContextStronger Simple Process HabitsWeaker Simple Process HabitsTypical Audit Room1.09(1.19) n=341.91(1.84)n=32Alternative Audit Room 1.83(1.58)n=301.28(1.30)n=32Panel B: Analysis of Variance for Issues Identified df SS MS F-statistic p-valueSimple Process Habit Strength 1 0.57 0.57 0.25 0.615 Context 1 0.12 0.12 0.05 0.820Simple Process Strength X Context 1 14.9914.99 6.73 0.011 Error 124276.09 2.23Panel C: Simple Effects Comparisons df t-statistic p-valueContext for Stronger Simple Process Habits Auditors124 1.99 0.024*Context for Weaker Simple Process Habits Auditors124-1.68 0.096Simple Process Habit Strength in the Typical Context 124-2.23 0.014*Simple Process Habit Strength in the Alternative Context124 1.46 0.148We conduct an ANOVA to test our hypotheses. Independent variables are defined in the notes to Figures 1 and 2. The dependent variable is Issues Identified, which is the total number of issues, out of seven embedded issues, that the auditor identifies in the goodwill impairment case. Descriptive statistics are reported in Panel A. Panel C reports our test of H1, that is, the simple effect of Simple Process Habit Strength on Issues Identified, considering the typical audit room context. Panel B and Panel C report our test of H2, including the interaction between Simple Process Habit Strength and Context on Issues Identified, as well as the effect of Context for auditors with stronger and weaker simple process habits. P-values with * are one-tailed, and all other p-values are two-tailed. TABLE 2: Performance (Issues Identified) for Auditors by Context, Simple Process Habit Strength, and Experience Working in the Alternative ContextPanel A: Regression Model (Full Sample) df Coeff. Std. Error t-statistic p-valueIntercept119 2.310.42 5.56<0.001Context119-1.390.57-2.42 0.017Simple Process Habit Strength119-1.250.59-2.10 0.038Alternative Context Experience119-0.010.01-1.17 0.243Context X Simple Process Habit Strength 119 2.650.83 3.19 0.002Context X Alternative Context Experience119 0.020.01 1.70 0.093Simple Process Habit Strength X Alternative Context Experience 119 0.010.01 0.82 0.416Context X Simple Process Habit Strength X Alternative Context Experience119-0.040.02-2.00 0.048Panel B: Simple Effects Estimates for Auditors with Stronger Simple Process Habits df Coeff. Std. Error t-statistic p-valueContext for Low Alternative Context Experience Auditors 119 1.170.54 2.15 0.017*Context for Medium Alternative Context Experience Auditors1190.640.38 1.67 0.049*Context for High Alternative Context Experience Auditors1190.110.62 0.17 0.432*Panel C: Simple Effects Estimates for Auditors with Weaker Simple Process Habits df Coeff. Std. Error t-statistic p-valueContext for Low Alternative Context Experience Auditors 119 -1.280.53-2.43 0.017Context for Medium Alternative Context Experience Auditors119 -0.700.38-1.85 0.067Context for High Alternative Context Experience Auditors119 -0.110.49-0.23 0.819We conduct a regression model to examine effects of Context, Simple Process Habit Strength, and Alternative Context Experience (and their interactions) on Issues Identified. Simple Process Habit Strength is defined in the notes to Figure 1. Context is defined in the notes to Figure 2. Alternative Context Experience is measured as auditors’ self-reported indication of the amount of time they spend working in contexts like our alternative context. Issues Identified is defined in the notes to Table 1. The results of the regression model are reported in Panel A. Estimates from the model for the effect of Context at low (one standard deviation below the mean), medium (the mean), and high (one standard deviation above the mean) levels of experience working in the alternative context are reported in Panel B (for auditors with stronger simple process habits) and in Panel C (for auditors with weaker simple process habits). P-values with * are one-tailed, and all other p-values are two-tailed. TABLE 3: Performance (Issues Identified) for Auditors by Context, Simple Process Habit Strength, and Self-Discipline While Working on the Estimates TaskPanel A: Regression Model (Full Sample) df Coeff. Std. Error t-statistic p-valueIntercept120 1.770.91 1.96 0.053Context120-0.501.21-0.41 0.679Simple Process Habit Strength120-2.241.30-1.72 0.088Self-Discipline120 0.030.19 0.16 0.875Context X Simple Process Habit Strength 120 4.112.01 2.04 0.043Context X Self-Discipline120-0.030.25-0.11 0.915Simple Process Habit Strength X Self-Discipline 120 0.290.26 1.11 0.268Context X Simple Process Habit Strength X Self-Discipline120-0.550.40-1.38 0.170Panel B: Regression Model (with Stronger Simple Process Habits Auditors) df Coeff. Std. Error t-statistic p-valueIntercept60-0.470.85-0.55 0.583Context60 3.611.46 2.47 0.016Self-Discipline 60 0.320.17 1.91 0.061Context X Self-Discipline60-0.580.28-2.05 0.045Panel C: Simple Effects Estimates for Stronger Simple Process Habits Auditors df Coeff. Std. Error t-statistic p-valueContext for Low Self-Discipline Auditors 60 1.470.51 2.90 0.003*Context for Medium Self-Discipline Auditors60 0.740.34 2.15 0.018*Context for High Self-Discipline Auditors60 0.000.48 0.01 0.497*We conduct a regression model to examine effects of Context, Simple Process Habit Strength, and Self-Discipline (and their interactions) on Issues Identified. Simple Process Habit Strength is defined in the notes to Figure 1. Context is defined in the notes to Figure 2. Self-Discipline is measured as auditors’ agreement on a 7 point scale that they exerted self-discipline while working on the goodwill task. Issues Identified is defined in the notes to Table 1. The results of the regression model are reported in Panel A. Panel B reports results from running the regression within strong simple process habits auditors. Panel C reports estimates from the model for the effect of Context at low (one standard deviation below the mean), medium (the mean), and high (one standard deviation above the mean) levels of Self-Discipline. P-values with * are one-tailed, and all other p-values are two-tailed. TABLE 4: Thoughts Reported by Auditors in Written Paragraphs (by Context) Panel A: Descriptive Statistics – Frequencies (Percentages)Mentioned the Presence of the InferenceMentioned the Absence of the InferenceNo Mention of the InferenceInferenceTypical ContextAlternative ContextTypical ContextAlternative ContextTypical ContextAlternative ContextCoaching or Supervising of Staff 31(47.0%)10(16.1%)0(0.0%)13(21.0%)35(53.0%)39(62.9%)Stress (i.e., due to clutter, cramped) 35(53.0%)4(6.5%)1(1.5%)9(14.5%)30(45.5%)49(79.0%)Interruptions 33(50.0%)2(3.2%)5(7.6%)33(53.2%)28(42.4%)27(43.6%)Busy Season 6(9.1%)0(0.0%)0(0.0%)0(0.0%)60(90.9%)62(100.0%)Contact from the Client 16(24.2%)21(33.9%)0(0.0%)0(0.0%)50(75.8%)41(66.1%)Time Pressure 4(6.1%)1(1.6%)2(3.0%)14(22.6%)60(90.9%)47(75.8%)Personal Recollection 6(9.1%)1(1.6%)N/A N/A60(90.9%)61(98.4%)Panel B: Chi-Square Test for Differences Across Context Conditions Inference Chi-Square Test ResultCoaching or Supervising of Staff χ2(2) = 23.87, p < 0.001 Stress (i.e., due to clutter, cramped) χ2(2) = 35.52, p < 0.001Interruptions χ2(2) = 48.03, p < 0.001Busy Season χ2(1) = 5.91, p = 0.015Contact from the Client χ2(1) = 1.44, p = 0.230Time Pressureχ2(2) = 12.27, p = 0.002Personal Recollectionχ2(1) = 3.46, p = 0.063The above table displays our analysis of thoughts auditors had in their written paragraphs by Context condition. Panel A displays frequencies and percentages (coded from auditors’ written paragraphs during the manipulation of context, i.e., in which they imagined themselves working in the particular context). Panel B displays results of a chi-squared test for differences in frequencies across Context conditions. APPENDIX – EMBEDDED ISSUES IN GOODWILL IMPAIRMENT CASEAssumptionDescription of the issueRevenue projectionsThe company consistently over-projected growth in the past, which casts doubt on the accuracy of the current projections Revenue projectionsThe projected revenue growth of Product C is not guaranteed due to uncertainties (a new competing product, delays in production, and/or reliance on synergies with existing product lines)Revenue projections There is an outlier in the benchmarking analysis for the projected revenue growth rate, so while the client’s rate is below the peer average, it would not be if this outlier were excluded Revenue projections Projected revenue growth is inconsistent with the overall market/economy/industry outlookOperating expense projectionsThe company plans to increase sales staff by 10 percent in the next three years, resulting in a significant increase in employment expense. This is not factored into the client’s operating expense assumptionCapital expenditures projectionsThe company is building a new $14 million office building, which is not included in the capital expenditures forecast. Capital expenditures projectionsThe company’s forecasted capital expenditures exhibit slower growth than industry analysts’ projections. ................
................

In order to avoid copyright disputes, this page is only a partial summary.

Google Online Preview   Download