MOTIVATING REVISIONS OF MANAGEMENT ACCOUNTING SYSTEMS: AN EXAMINATION OF ORGANIZATIONAL GOALS AND ACCOUNTING FEEDBACK By Tyler F. Thomas A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of Business Administration—Doctor of Philosophy 2013 ABSTRACT MOTIVATING REVISIONS OF MANAGEMENT ACCOUNTING SYSTEMS: AN EXAMINATION OF ORGANIZATIONAL GOALS AND ACCOUNTING FEEDBACK By Tyler F. Thomas Organizations often revise their management accounting systems (MAS) to adjust for changes in markets, products, organizational designs, and technologies. Successful revisions are likely to require input from accountants and other employees throughout an organization. These individuals, however, have competing demands on their time. These demands and the uncertainty of the benefits of MAS revisions can make it challenging to motivate these individuals to revise MAS. It has been suggested that providing individuals with feedback that small-scale and/or short-run MAS changes have been a success can motivate continued change efforts. My study provides theory, based on motivational psychology, and experimental evidence about why and when this strategy predictably either increases or decreases individual motivation to exert further effort in revising MAS. Experimental results indicate, as predicted, that accounting feedback about the success or failure of MAS revisions can increase or decrease motivation to exert additional effort on revising MAS depending on the goal(s) activated in individuals’ minds by language in the accounting report. DEDICATION I dedicate this work to my family. iii ACKNOWLEDGEMENTS I gratefully acknowledge the help and insight from my dissertation committee: Richard Deshon, Joan Luft (chair), Ranjani Krishnan, and Mike Shields. I would also like to thank Ranjani Krishnan, Eric Marinich, and Karen Sedatole for help in recruiting students and Ryan Bowles, Matt DeAngelis, Dan Lynch, Miles Romney, Joe Schroeder, and participants at Michigan State University, the 2012 Brigham Young University Accounting Research Symposium, the 2013 MAS Midyear Conference, Queen’s University, the Georgia Institute of Technology, the University of Waterloo, the University of Wisconsin, and Cornell University for their comments and assistance. I also appreciate the financial support of this study from the Department of Accounting and Information Systems at Michigan State University and the Institute of Management Accountants. My advisor Joan Luft has been an incredible mentor and I have learned so much from her. The skills and lessons that I have gained will greatly enrich my research and academic career and I hope that I can share these lessons with future PhD students that I will have the pleasure to work with. I have also greatly enjoyed the time I have been able to spend with my fellow PhD students and the faculty at Michigan State University. I thank them for all of their support and encouragement and wish them the best in their careers. I am ever so grateful for the support and love of my family. My wonderful wife Andrea and four amazing children, Bret, Kate, Adyn, and Jace, have given me smiles to brighten my day, hugs and kisses to lift me up, and the faith that everything would work out for the best. I cannot begin to express my gratitude for the never ending support of Andrea and the incredible companion that she is. She has been with me every step of the way, through the seemingly iv endless years of school, with a smile on her face and words of encouragement to push me forward. My family is such a blessing and they make me want to be better each and every day. I love them dearly. Lastly, I thank my Heavenly Father and His Son, Jesus Christ, for guiding me and my family, and strengthening us. We are eternally grateful for all that we have been blessed with. v TABLE OF CONTENTS LIST OF TABLES ………………………………………………………...…………………… vii LIST OF FIGURES …………………………………………………………………………… viii CHAPTER 1: INTRODUCTION ….……………………………………………………………. 1 CHAPTER 2: LITERATURE REVIEW & HYPOTHESIS DEVELOPMENT …....................... 7 MAS Revision ……………………………………………………………………….............. 7 Motivation and Goals ………………………………………………………………………... 9 Accounting Report Design and Priming …………………………………………………… 12 Effects of Goal Priming in the Accounting Context ……………………………………….. 15 Theoretical Predictions …………………………………………………………………….. 16 CHAPTER 3: EXPERIMENTAL METHOD …………………………………………………. 18 Participants and Design …………………………………………………………………….. 18 Task and Procedure ………………………………………………………………………… 18 CHAPTER 4: RESULTS .……………………………………………………………………… 22 Manipulation Check …..……………………………………………………………………. 22 Descriptive Statistics and Measures …................................................................................... 22 Hypothesis Tests …………………………………………………………………………… 25 Time ……………………………………………………………………………. 25 Number of Times Continued …………………………………………………... 27 CHAPTER 5: CONCLUSION….…………………………………………………………….... 30 APPENDICES ...……………………………………………………………………………..… 34 APPENDIX A Figures 1 - 5 ……........................................................................................... 35 APPENDIX B Tables 1 - 5 …………............…….……………………………………...… 42 APPENDIX C Experimental Task …………………………………………………………. 50 APPENDIX D Post-Experiment Questionnaire ……………………………………………. 75 REFERENCES ………………………………………………………………………………… 83 vi LIST OF TABLES Table 1: Feedback Following Cost Driver Choice for the Painting Activity ………………….. 43 Table 2: Cell Means (Standard Deviations) of Future Effort on MAS Revision ………………. 45 Table 3: Correlations and Descriptive Statistics of the Two Measures of the Dependent Variable, Time and #TimesContinued, and the Covariates …………………..………………..… 46 Table 4: Future Effort on MAS Revision Measured as the Natural Log of Time Spent on MAS Revision (LNTime) ………………….………………………………………………... 48 Table 5: Future Effort on MAS Revision Measured as the Number of Times Continued Revising MAS (#TimesContinued) …………………..…………………………………………. 49 Table 6: Production Process and Overhead Cost Information for the Painting Activity …….… 57 Table 7: Accounting Feedback Reports with Priming and Prior Success Manipulations ……... 60 Table 8: Production Process and Overhead Cost Information for the Fabrication Activity …… 66 Table 9: Production Process and Overhead Cost Information for the Finishing Activity …...… 68 vii LIST OF FIGURES Figure 1: Theoretical Predictions ………………………………………………………………. 36 Figure 2: Experimental Task Event Line ………………………………………………………. 37 Figure 3: Marginal Means (Adjusted for Covariates) of Future Effort on MAS Revision Following Initial Cost Driver Choice Measured as Time Spent on MAS Revision (Time) ……………………………………………………………………………..… 39 Figure 4: Marginal Means (Adjusted for Covariates) of Future Effort on MAS Revision Following Initial Cost Driver Choice Measured as the Natural Log of Time Spent on MAS Revision (LNTime) ………………………………………………………..…... 40 Figure 5: Marginal Means (Adjusted for Covariates) of Future Effort on MAS Revision Following Initial Cost Driver Choice Measured as the Number of Times Continued Revising MAS (#TimesContinued) …………………………………………………. 41 Figure 6: Michigan’s Cost Accounting Policy ……………………………………………….… 53 viii CHAPTER 1: INTRODUCTION Academic and practitioner literature has long stressed both the value and difficulty of 1 revising management accounting systems (MAS) in response to changes in markets, products, organizational designs, and technologies (Keegan, Eiler, and Jones 1989; Keegan and Eiler 1994; Tatikonda and Tatikonda 1998; Waldrup, MacArthur, and Michelman 2009; Wouters 2009; Chapman and Kern 2011). Successful revisions can improve accounting information quality, but MAS revisions can pose significant challenges for motivating the individuals involved with the revision. One challenge is that these individuals have competing demands on their time. Further, MAS revisions can be uncertain projects in which the path to the goal is not always clear and the expected benefits of these revisions and thus appropriate compensation for revision effort can be difficult to estimate. This uncertainty of achieving success can discourage effort. Research has examined a number of determinants of successful MAS revisions, such as top management support, adequate resources, and training (Cooper et al. 1992; Shields 1995; Krumwiede 1998; Balakrishnan, Hansen, and Labro 2012). This research has not disentangled the enabling and motivating effects of these factors, however, and little accounting research has addressed motivation as such. One suggested strategy to motivate continued effort in revision projects is that when possible, a major revision should be initiated with a small-scale activity that is likely to be a success, and these short-run successes should be prominently reported. Prior successes—for example, in a pilot or demonstration project or a previous revision—are expected to motivate further revision efforts by increasing enthusiasm and estimates of the probability of further success, reducing the uncertainty about the outcomes of effort (O’Hara, Watson, and Kavan 1999; Hankinson and Lloyd 2000; Horngren, Datar, and Rajan 2012). Feedback 1 Examples of MAS are product costing systems and performance measurement systems. 1 concerning success can be provided through accounting reports and research has shown that the presentation and format of accounting reports can affect individual attention to information and decision-making (Cardinaels 2008; Bloomfield et al. 2011). In my study I bring the two literature streams of MAS revision and accounting report presentation and format together by examining whether accounting reports about the success of prior MAS revision and language accompanying these reports can affect future effort on MAS revision. Report presentation and format can include language or graphics that influence effort by affecting how people interpret feedback provided by the report. I argue, based on theories from motivational psychology (Fishbach and Dhar 2005; Fishbach, Dhar, and Zhang 2006), that how individuals interpret feedback indicating success (or failure) on prior MAS revision can increase or decrease their effort on continued MAS revision, dependent on goal framing. If individuals interpret prior success in terms of goal commitment, then success can motivate further effort by invigorating individuals and increasing their belief that the goal can be achieved (Locke and Latham 2002; Ilies and Judge 2005). If the same feedback is interpreted as goal progress, however, then prior success can lead to less subsequent effort as individuals believe that the goal is (at least partially) fulfilled and feel free to exert effort elsewhere (Dhar and Simonson 1999; Fishbach et al. 2006). Whether feedback is interpreted in terms of goal commitment or goal progress can depend on whether a high-level goal (e.g., improving accounting for making better decisions) or a low-level goal (e.g., choosing a better cost driver for specific activities) is activated in individuals’ minds during MAS revision. Individuals often have goal hierarchies that consist of high-level goals that are abstract and long-run focused and low-level subgoals that are concrete and immediate. Even if individuals know their goals equally well at multiple levels of the hierarchy, one goal level is 2 often more strongly activated in their minds and is more likely to influence their behavior and decisions at a given time (Anderson, Reder, and Lebiere 1996). I argue that language accompanying accounting information, not in the standard body of the accounting report, can act as a prime that affects the goal that is activated in individuals’ minds and thus influence their subsequent behavior. Priming is an implicit memory effect in which exposure to a stimulus can affect behavior in response to a subsequent stimulus (Schacter 1994; Neely 2003), by activating affect, knowledge, or goals (Higgins 1996). If a high-level goal is more activated in individuals’ minds during MAS revision, for example through priming, then prior success compared to prior failure is expected to motivate more revision effort. Following success, individuals feel more committed to the high-level goal and want to exert more effort towards this goal (Fishbach et al. 2006; Zhang, Fishbach, and Dhar 2007). Conversely, when a low-level goal is more activated—as low-level goals often are in the absence of a high-level goal prime, because they are more concrete and immediate—prior failure will motivate more revision effort compared to prior success. With success, individuals will feel as though sufficient progress has been made on the low-level goal and thus their effort can be exerted elsewhere. Therefore, the effect of a small-scale and/or short-run success depends crucially on whether high-level or low-level goals are activated. I use an experiment to examine how the interaction between priming of the high-level goal of improving accounting for making better decisions and prior success/failure affects additional effort exerted on MAS revision. In this setting, individuals are involved in improving the organization’s costing system by recommending new cost drivers for multiple production activities. Following their first cost-driver choice, participants receive an accounting feedback report indicating whether their choice was successful or not (i.e., whether usage of the chosen 3 cost driver relates to overhead resource usage). Some participants receive a prime with this accounting feedback report in the form of an accounting slogan which is expected to activate the high-level goal, whereas others do not receive this prime with the accounting report. Participants then have the opportunity to continue revising the MAS, and I measure the effort they exert in additional revision of the costing system. Consistent with my hypothesis, the experimental results indicate that accounting reports influence motivation not only by providing feedback about a prior success or failure of MAS revision, but also by influencing the interpretation of this feedback through the presence or absence of language accompanying these reports that activates high-level goals. As predicted, prior success while priming the high-level goal leads to more effort in MAS revision than prior failure, but prior failure leads to more effort in MAS revision than prior success when the highlevel goal is not primed. Also as predicted, prior success in MAS revision leads to more effort when the high-level goal is primed than when it is not primed, but not priming the high-level goal with prior failure leads to more effort than if this goal is primed. This study provides insight to managers on how to motivate employees to be involved in and exert effort towards MAS improvement to support better decision-making throughout the organization. My results show that language as a peripheral communication (e.g., a slogan) can affect the goals that are activated in individuals’ minds, affecting their motivation to persist in MAS revision. As a project proceeds successfully managers can increase motivation by reporting this success and activating the overall goal of the project (e.g., using a slogan pertaining to this goal). If the project is not proceeding successfully, however, this feedback needs to be reported with a focus on the task at hand to keep the short-term goal at the forefront of employees’ minds to motivate them to continue in the revision process and make successful progress. 4 My study further makes three contributions to the accounting literature. First, it provides theory and evidence on why and when the often-recommended strategy of seeking and reporting small-scale and/or short-run success actually does or does not increase individuals’ motivation to exert further effort on MAS revision. This strategy can be effective when individuals are primed with the high-level goal for the project. In the absence of such a prime, however, low-level goals are likely to be more activated in individuals’ minds, because they are more concrete and immediate; and if this is the case, then prior success can be counterproductive, leading to less subsequent effort on MAS revision. Second, my study adds to the literature on accounting report design. Prior studies have examined how report presentation and format, including the addition of graphics in reports, can affect decision-making (e.g., Cardinaels and Van Veen-Dirks 2010; Bloomfield et al. 2011; Jiang, Petroni, and Wang 2012). This research has focused on cognitive and attention-directing effects and the weights that individuals place on different cues. My study examines how a prime—a slogan—accompanying an accounting feedback report, can affect motivation and effort. Reports of accounting feedback can include not only information about success or failure but also language (or other information, such as graphics) that can make high- or low-level goals more activated in individuals’ minds. The combination of prior revision feedback and the more strongly activated goal can increase or decrease motivation in revising MAS. Third, my study contributes to the emerging literature on priming in accounting. Research in accounting has provided evidence of priming effects, generally investigating primes that provide information that is an essential part of the accounting task requiring thorough cognitive processing (e.g., Hammersley, Bamber, and Carpenter 2010; Lambert and Agoglia 2011). I add to this literature by evaluating the priming of goals and showing that a peripheral stimulus 5 (prime) that does not require attention for the task to be completed—an accounting slogan—can either increase or decrease motivation to improve accounting information. The remainder of my study is organized as follows: chapter 2 provides a literature review and hypothesis development. Chapter 3 describes the experiment. Chapter 4 presents the results of the experiment and chapter 5 concludes. 6 CHAPTER 2: LITERATURE REVIEW & HYPOTHESIS DEVELOPMENT I first review literature on MAS revision. Then I present in more detail the problem of motivating individual effort persistence in these projects and a solution often proposed in the literature, seeking out and communicating prior revision success. I then review the psychology literature that predicts that prior success can only motivate continued effort under certain conditions and explains how this motivating effect can be influenced by priming. Finally, I identify gaps between the psychology literature and the accounting problem that makes it uncertain whether or how the psychology theory will apply in the accounting context. MAS Revision MAS are valuable for decision-making when they produce accurate information. As organizations change, MAS need to be revised so as not to become obsolete and reduce their effectiveness for decision-making. To fully elicit the benefits from MAS, they need to be evaluated, monitored, and revised (Ittner and Larcker 2003; Wouters 2009; Chapman and Kern 2011), which requires time and effort to be successful (Keegan et al. 1989; Wouters and Wilderom 2008; Kelly 2010). Accounting research indicates that involving accountants and nonaccountants in MAS revision can lead to improvements in MAS and firm performance (Wouters 2009; Chapman and Kern 2011; Wouters and Roijmans 2011). Studies have also shown that user involvement during the revision process can improve user satisfaction and interaction with the revised MAS (Hunton and Gibson 1999; Abernethy and Bouwens 2005; Eldenburg et al. 2010). Motivating the individuals involved in revising MAS, however, can be challenging. One challenge is that the individuals that participate in the revision often have competing demands on their time (Shields and Young 2000; Simons 2000; Wouters 2009; Eldenburg et al. 2010; Chapman and Kern 2011). Further, MAS revisions can be uncertain projects in which the path to 7 the goal is not always clear. These projects often require trial-and-error learning and the development of prototypes for experimenting and testing (Wouters and Wilderom 2008; Wouters and Roijmans 2011), which may be abandoned or substantially altered as a result of the experiments and tests. This can often make it difficult to estimate the benefits of MAS revision activities (and thus the appropriate amount of compensation), and the uncertainty of achieving these benefits can discourage effort from risk-averse employees. Thus, linking significant compensation to these activities can be impracticable because of the inherent uncertainty in these projects. This difficulty in contracting and compensating for effort in these activities can lead to a preference for “consummate cooperation” in which individuals work together to achieve an end without contract or enforcement (Kay 1995). This cooperation can be short lived, however, as it gives an incentive to free-ride further discouraging individual effort in these types of projects. Research in a number of business disciplines suggests that individuals can be motivated to continue working on organizational and MAS change projects by providing information that a small-scale and/or short-run change was a success (Shields and Young 1989; O’Hara et al. 1999; Hankinson and Lloyd 2000; Kotter 2007). Anderson (1995) details GM’s process of implementing an activity-based costing system throughout the organization, and shows that the initial phases of the revision process consisted of running pilot studies involving employees with varying backgrounds. To help motivate effort and involvement in the project, individuals at successful pilot plants were asked to give testimonials of their successes to those at other pilot plants to help them get energized and move forward in the process. Feedback on the success or failure of prior MAS revision efforts can provide information about the likelihood that future efforts to revise the MAS will succeed or fail, and thus about the project’s expected net benefit. Feedback about success can also generate positive affect that 8 energizes individuals to exert additional effort to revise MAS (Ilies and Judge 2005). For example, Horngren et al. (2012) state that for successful implementation of an activity-based costing system (ABC), individuals should focus on “Seeking small short-run successes as proof that the ABC implementation is yielding results” and that this can motivate those involved “… to stay on course and build momentum” (155). Research provides some empirical support for this strategy (Kwon and Zmud 1987; Anderson 1995), but does not fully explain the mechanism of this strategy and the limitations to its effectiveness. Motivation and Goals Individual motivation can be influenced by many non-monetary factors such as affect, goals, self-efficacy, and social comparison (Bargh et al. 2001; Locke and Latham 2002; Mitchell and Daniels 2003). For decision-influencing purposes, accounting information is generally used to support monetary incentives to motivate desired behavior by affecting individuals’ conscious choice of actions. Accounting, however, can also affect behavior by other means, such as by influencing individuals’ frame of reference and their valuation of outcomes, even when expected monetary payoffs and risks are held constant (Luft 1994; Rowe 2004; Luft and Shields 2009) and providing social comparisons that influence future behavior (Frederickson 1992; Hannan, Krishnan, and Newman 2008). Evaluating the non-monetary motivating effects of reporting prior success in MAS revision is important because it can be difficult to tie significant compensation to success in uncertain MAS revision projects. Further, even if monetary incentives are provided, non-monetary motivation could still have an effect on motivation. If the non-monetary motivation is positive, then lower monetary rewards would be required to motivate a given level of effort. Negative non-monetary motivation, however, could reduce the effectiveness of monetary benefits, requiring more monetary reward to elicit a given level of effort. 9 Psychology research on motivation provides theory and evidence that feedback on the success (or failure) of a prior effort can be interpreted differently and thus can influence future motivation either positively or negatively (Fishbach et al. 2006). If prior success is interpreted in terms of goal progress—the progress made towards a previously defined goal (e.g., Carver and Scheier 1998)—then an individual can experience benefits associated with goal fulfillment and feel free to pursue other goals (Dhar and Simonson 1999; Fishbach and Dhar 2005; Fishbach et al. 2006). The opposite can occur if an individual interprets prior success (or failure) in relation to goal commitment—the importance or strength of the goal—as individuals often re-evaluate their commitment to a goal in light of current circumstances (Bem 1972; Fishbach and Dhar 2005). All else equal, goal commitment increases with success by emotionally invigorating individuals and increasing their belief that the goal is attainable, which motivates continued goal pursuit (Locke and Latham 2002; Ilies and Judge 2005). Conversely, goal commitment decreases with failure, demotivating continued effort. Whether an individual interprets feedback as goal progress or goal commitment can depend on whether the goal that is more activated in individuals’ minds is a high-level or low-level goal. Individuals often have a hierarchy of goals in which high-level goals are abstract and long-run focused, and subordinate low-level goals are concrete and short-run focused (Austin and Vancouver 1996; Bateman, O’Neill, and Kenworthy-U’Ren 2002). For example, in revising a costing system individuals can focus more on trying to improve accounting information quality for making better decisions (a high-level goal) or on improving the allocation of overhead costs for specific activities (a low-level goal). The low-level goal is one of many possible improvements that can be made to the MAS to achieve the high-level goal. 10 Even when individuals know their goals equally well at multiple levels, one goal level is often more strongly activated in their minds than another at any given time. When a goal is more strongly activated, this does not necessarily mean that it has a higher conscious priority; in fact, activation can be unconscious even when it influences behavior (Bargh et al. 2001). Rather, when a goal is more strongly activated than other goals, the goal itself and information relevant to it are more accessible in working memory and thus more likely to influence decision-making (sometimes but not always consciously), and conflicting goals and information are more likely to be suppressed (Anderson et al. 1996; Shah, Friedman, and Kruglanski 2002; Förster and Liberman 2007). Individual attention and behavior can thus be determined by active goals (Dijksterhuis and Aarts 2010). If a high-level goal is more strongly activated, then feedback about prior effort can affect individuals’ motivation to pursue the high-level goal as they evaluate their commitment to the goal (Fishbach et al. 2006). Prior success increases goal commitment, motivating continued effort towards the high-level goal, whereas prior failure decreases goal commitment and leads to a reduction of effort toward the high-level goal. Thus, if overall improvement in the decision usefulness of accounting information is the goal that is more strongly activated in individuals’ minds during MAS revision, then prior success will increase their commitment to this high-level goal and increase the effort devoted to making additional revisions to the MAS. If a low-level goal is more strongly activated—as low-level goals often are, because they are concrete and immediate and can pertain to the task at hand—then individuals tend to interpret outcome feedback about prior actions related to this goal in terms of goal progress and focus on attainment of the more defined low-level goal (Fishbach et al. 2006). If it is attained, then individuals can feel free to allocate their effort to other activities. But if the low-level goal is not 11 attained, then individuals will be motivated to continue their efforts towards achieving this goal. Thus, if a low-level goal is more strongly activated, then individuals are more likely to interpret their prior MAS revision success as an indication that they have made sufficient progress on revising the MAS for the time being and can turn their effort elsewhere. These arguments can provide insight into how and when the strategy of reporting prior success will be effective in motivating persistence in MAS revision. Prior success can be effective in motivating continued effort when the high-level goal is activated, which increases commitment and encourages effort, but will not be if the low-level goal is more activated as individuals are more likely to believe that sufficient progress has been made. If the prior MAS revision action is unsuccessful, then activating the high-level goal can be demotivating by decreasing goal commitment. If the low-level goal is more activated, however, then prior failure can actually be motivating as individuals are focused on achieving the low-level goal and continue to allocate effort to it. Accounting Report Design and Priming Accounting research has provided evidence that the presentation and format of accounting reports can influence individual behavior and decision-making. How accounting information is organized and presented has been shown to influence decisions made by users of financial statements (Maines and McDaniel 2000; Bloomfield et al. 2011; Clor-Proell, Proell, and Warfield 2013) and performance evaluations (Cardinaels 2008). Studies have also provided theory and evidence that graphics included on accounting reports can influence the weight that individuals place on different pieces of information, both in financial and management accounting reports (Malina and Selto 2001; Cardinaels and Van Veen-Dirks 2010; Jiang et al. 2012). These studies focus on the cognitive and attention-directing effects of the presentation and 12 format of accounting reports (influences on cue-weighting in judgments and decisions), but they do not focus on how presentation and format can influence motivation and individual effort. In my study I evaluate the motivational effects of accompanying language in accounting reports through the mechanism of priming. How strongly a high-level goal is activated in individuals’ minds can be influenced by exposure to a prime, that is, a stimulus that activates constructs in individuals’ minds (Bargh 2006). Priming is an implicit memory effect in which exposure to a stimulus affects the response to a subsequent stimulus (Schacter 1994; Neely 2003; Hsu and Schṻtt 2012). Exposure to the prime can unconsciously influence individual attention, behavior, and decision-making by activating affect, knowledge, or goals (Higgins 1996; Bargh et al. 2001; Linkenauger 2012). One example of a priming effect on behavior comes from a study of employees working in a call center (Shantz and Latham 2009). Employees who received shift instructions that included a small picture of a woman winning a race—to prime achievement—raised 60% more money during the work shift than those employees who received the same instructions but 2 without the picture. This prime activated the concept of achievement for employees, which then influenced their work motivation. Shantz and Latham’s (2009) study and other research in industrial-organizational psychology have demonstrated that, holding goals and monetary incentives for achieving these goals constant, priming the goals motivates additional effort both in laboratory and field settings (Latham, Stajkovic, and Locke 2010). Recent studies in accounting have investigated priming (Clor-Proell and Nelson 2007; Hammersley et al. 2010; Elliott, Hobson, and Jackson 2011; Lambert and Agoglia 2011). These 2 Those who received shift instructions with the prime raised $349 compared to $217 raised by those who received shift instructions without the prime, on average. 13 studies provide evidence of priming effects in accounting tasks using primes that provide information that is an essential part of the accounting task requiring thorough cognitive processing. For example, Hammersley et al. (2010) ask participants to list specific firm fraud risks that were previously brainstormed, to prime a fraud frame, and Lambert and Agoglia (2011) have participants read reviewer notes that prime either a conclusion-frame (e.g., to reach an appropriate conclusion) or a documentation-frame (e.g., to have defensible documentation). 3 These studies evaluate how the priming of different information can affect decisions, but they do not, however, examine the priming of goals and its effect on effort. I argue that goal priming through language that accompanies accounting information for peripheral communication (e.g., a slogan) that does not require attention for completion of the task can influence individual behavior and either increase or decrease effort on MAS revision. Language with accounting information can help to determine which goal is more activated in individuals’ minds during MAS revision, and goal activation can then influence their interpretation of prior MAS revision outcomes and their future effort on MAS revision, holding monetary incentives constant. This language can prime the high-level goal (e.g., improving accounting for making better decisions) during MAS revision making it more activated, but if not the low-level goal (e.g., choosing a better cost driver for specific activities) is likely to be more activated since it is more concrete and immediate and the task itself can activate this goal. This effect on which goal is more likely to be activated will lead to the consequences of goal activation described in the previous section. 3 Hammersley et al. (2010) provides the clearest evidence of priming, and, similar to my study, indicates that its effects are not always in the intuitive direction. Their task, prime, and theory, however, are different than those in my study. 14 Effects of Goal Priming in the Accounting Context Psychology research has evaluated the effects of goal priming and the interpretation of success or failure on prior low-level goal actions (e.g., Fishbach and Dhar 2005; Fishbach et al. 2006). The accounting context, however, differs from that studied in psychology in several ways, and thus it is uncertain whether the findings in the psychology literature apply to accounting. The first and most important difference is the type of goals of interest. Studies in psychology focus on individual goals to which individuals are already committed, e.g., academic success or healthy living. My study examines the role of organizational goals on individual effort. Organizational or assigned goals could reduce the personal relevance of and commitment to the goal and motivation for an individual to achieve the goal. Employees from different areas of the firm involved in MAS revision might not view the benefit that it can provide or how the revision project positively affects them personally or their department. The second difference is the task used to examine motivation. Tasks in psychology studies include predicting the amount of exercise time in the next week or the selection of food. Shantz and Latham (2011, 291) note that similar tasks used in psychology priming studies can have “… questionable relevance for most work settings.” Further, many of these tasks rely on predictions of effort and not on actual effort exertion. The third difference is the feedback used for evaluation. Much of the feedback in the psychology studies focuses on social comparison (i.e., participants’ performance and effort compared to others). Although accounting feedback sometimes provides relative performance information, it does not always do so, as is the case in my study. There is considerable evidence that relative performance information can have stronger motivational effects than information 15 about an individual’s own performance only (Frederickson 1992; Hannan et al. 2008). Therefore, feedback about MAS revision outcomes could have less of an effect on individuals’ motivation. Theoretical Predictions Figure 1 illustrates the hypothesis of my study. I predict that if the accounting feedback report primes the high-level goal of improving accounting for making better decisions, then a report of prior success in MAS revision will motivate more effort on subsequent MAS revision than a report of prior failure, (1 > 3 in Figure 1). Conversely, I predict that if the report does not prime this high-level goal (and thus a more concrete and immediate low-level goal is more likely to be activated), then a report of prior failure in MAS revision will motivate more effort than a report of prior success (4 > 2). Thus, I argue that reporting success can prompt high goal commitment and invigorate and encourage further effort when the high-level goal is activated. The reporting of success can also be counterproductive, however, when the low-level goal is activated, giving individuals the belief that sufficient progress has been made and that their effort can be allocated elsewhere. These two predictions specify the different signs of the slopes of the two lines in Figure 1, indicating that prior success does not always increase effort. They do not specify the relative vertical locations of the lines, however. For example, priming a high-level goal is predicted to motivate more effort with prior success than with prior failure, but this does not address whether priming will have a positive effect on effort even with prior failure. Psychology theory also supports predictions about the relative location of the two lines. Thus, I predict that feedback indicating a prior success will motivate more revision effort when a slogan priming the highlevel goal is included than when it is not (1 > 2). In contrast, if feedback indicates a prior failure, not including the slogan will motivate more revision effort than will including the slogan (4 > 3). 16 That is, I propose that language accompanying accounting reports (e.g., a slogan) can have a positive motivating effect which stimulates future effort on MAS revision when prior revision effort is successful. The same language, however, can backfire if prior efforts fail, thus reducing future effort on MAS revision. H: Priming of the high-level goal will interact with the success of prior MAS revision to affect the effort exerted on subsequent MAS revision such that: a) When the high-level goal of improving accounting for making better decisions is primed, prior MAS revision success will lead to more effort than prior failure (1 > 3), b) When the high-level goal is not primed, prior MAS revision failure will lead to more effort than prior success (4 > 2), c) When the prior MAS revision is a success, priming the high-level goal will lead to more effort than not priming the goal (1 > 2), d) When the prior MAS revision is a failure, not priming the high-level goal will lead to more effort than priming the goal (4 > 3). 17 CHAPTER 3: EXPERIMENTAL METHOD Participants and Design Participants in the experiment are 118 upper-level undergraduate accounting and fulltime MBA students currently taking management accounting courses. Their mean age is 26 (range 20 – 42) and 25% are female. Participants have taken an average of 3.6 accounting courses and have a mean of two years of professional work experience. All participants receive 4 $15 for participation in the experiment, which takes about 20 minutes to complete. The experiment has a 2 x 2 between-subjects design. The independent variables are the priming or not of the high-level goal of improving accounting for making better decisions (Priming) and feedback of success or failure on prior MAS revision (PriorSuccess). The dependent variable is future effort on MAS revision. Task and Procedure The experiment is conducted in a computer lab at a large university using a web-based computer program which assigns participants randomly to the priming and prior success/failure conditions. Participants receive two envelopes that contain pertinent information for different production activities and are informed that they should only open the envelopes when requested by the computer program. The production activity information is provided on paper for ease of reference during the task. Figure 2 presents a graphic representation of the sequence of events and participants’ decisions in the experiment. Participants in the experiment assume the role of production managers for Michigan Inc. (Michigan), an industrial firm that produces two products, Green (its basic product) and White (a 4 To examine the motivating effects of the independent variables separately and independently of the motivating effects of monetary incentives, participants receive a fixed payment for participation without any performance-dependent compensation. 18 more complex product), in three manufacturing activities: Painting, Fabrication, and Finishing 5 (see Appendix C for the experimental instrument used in the study). They are provided with information describing Michigan’s cost accounting policy, which explains that manufacturing overhead costs are allocated based on direct labor costs along with an example of how product costing can affect firm profit. Along with this information all participants are presented with the slogan “Better Accounting for Better Decisions” at the top right corner of the page which displays the cost accounting policy. The policy information indicates the importance that Michigan places on the quality of accounting in the firm and identifies Michigan’s high-level 6 goal as improving accounting to facilitate making better decisions. Participants are then told that a competitor is offering a product similar to Michigan’s Green at a lower price, and top management wants to determine whether Michigan can have sufficient long-run profits at this lower sales price. Michigan might not be able to compete at this price because this more focused competitor could be more efficient at making the product. Another possibility is that Michigan is as efficient as the competitor but is over-allocating manufacturing overhead costs to Green. To determine if Green can be profitable at a lower price, top management has requested that managers assist in revising Michigan’s costing system. Participants receive information about the production process and about overhead cost for the first manufacturing activity 7 (Painting), which uses a significant amount of manufacturing overhead resources. They then choose a cost driver from a list of potential drivers that can be tracked using Michigan’s current 5 The experimental setting is adapted from Booker (2000). 6 This information makes the high-level goal available to participants and thus able to be activated (Higgins 1996). 7 This information is contained on paper in envelope #1. 19 8 manufacturing information system. Participants are informed that Michigan’s management accountants will observe production in the Painting activity to determine if there is a relationship between their selected cost driver and manufacturing overhead resource usage. In order to provide some uncertainty about whether revision to the costing system will be successful, participants are told that the cost driver they choose may or may not provide a better estimate of product costs. A cost driver that provides highly accurate estimates might require information that the manufacturing information system does not currently capture; or the overhead costs in the Painting activity may be too diverse to be captured by a single cost driver, and data limitations make tracking smaller cost activity pools and multiple cost drivers for these activity pools infeasible in the near term. After selecting a cost driver for the Painting activity, participants receive feedback in an accounting report with data gathered by the management accountants during their observation of the production process, which compares the percentage change in the use of their chosen cost driver and the percentage change in the use of each manufacturing overhead resource to determine if a relationship exists (shown in Table 1). Half of the participants receive feedback indicating that their choice of cost driver was successful, (i.e., there is a relationship between their chosen cost driver and overhead resource usage, Table 1, Panel A) and the other half receive feedback indicating that their choice of cost driver was unsuccessful (i.e., there is no relationship between their chosen cost driver and overhead resource usage, Table 1, Panel B). The slogan “Better Accounting for Better Decisions,” presented earlier in the experiment to all participants, is now used to prime the high-level goal of improving accounting for making better decisions when the feedback is given. For half of the participants, this slogan (prime) is 8 Participants are made aware that modifying Michigan’s information system to collect different information is not feasible in the immediate future. 20 now presented at the top right corner of the accounting feedback report, and for the other half of 9 the participants this slogan is absent from the accounting feedback report. After receiving the accounting feedback report, participants are informed that several weeks have passed since they made a decision about the Painting cost driver and that they have returned to their regular production management responsibilities, which “… continue to fill your entire work week.” They then learn that top management would like to determine whether revisions should be made to the cost drivers for the other two production activities (Fabrication and Finishing). Participants receive production process and overhead cost information for these two activities 10 and then decide how much time they want to use to continue revising the costing system. This decision has four levels: Participants can (i) make no further revisions, (ii) examine and recommend (or not) a suggestion to use one cost driver for both activities, (iii) examine more overhead and activity information to select a cost driver for one of the two activities, or (iv) examine still more overhead and activity information to select cost drivers for both activities. When participants choose to stop revising the costing system or at the completion of all the possible costing system revisions in the experiment, they are directed to answer post-experiment questions that collect demographic information and explanations for participant decisions (see Appendix D for the post-experiment questionnaire). 9 The slogan is exactly the same as that was previously displayed to all participants with the description of the firm’s cost accounting policy earlier in the experiment. 10 This information is contained on paper in envelope #2. 21 CHAPTER 4: RESULTS Manipulation Check In this section I check whether the manipulation of prior success/failure on MAS revision was successful. In the post-experiment questions participants were asked how successful they believed their first cost-driver choice was at improving Michigan’s costing system and responses were given on an 10-point Likert scale with 1 = Extremely Unsuccessful and 10 = Extremely Successful. Participants in the prior success condition rated their success at improving the costing system higher (mean = 7.65) than those in the prior failure condition (mean = 5.53). This difference is significant (t = 7.99, p < .01, one-tailed) providing evidence that the manipulation was successful. Descriptive Statistics and Measures Table 2 presents descriptive statistics on the dependent variable, future effort on MAS revision. This variable is measured two ways. Time is the time in seconds that participants spend on revising the costing system after receiving accounting feedback on their first cost driver choice. #TimesContinued is the number of times participants choose to continue revising the costing system after receiving accounting feedback about their first cost-driver choice, measured 11 on a scale of zero to three. In preliminary tests, I found that these measures of the dependent variable were significantly (p’s < .10) affected by a number of the participants’ characteristics, three of which differ across treatment conditions: the number of accounting courses taken, knowledge of ABC, 11 This measure corresponds to the participants’ four possible revision decisions; 0 = make no further revisions, 1 = examine and recommend (or not) a suggestion to use one cost driver for both activities, 2 = examine more overhead and activity information to select a cost driver for one of the two activities, and 3 = examine still more overhead and activity information to select cost drivers for both activities. 22 and interest in the experiment. 12 These variables are thus included as covariates in the empirical models used to test the hypothesis. Table 3, Panel A provides a correlation matrix for the measures of future effort on MAS revision, Time and #TimesContinued, and the covariates. General descriptive statistics for these variables are shown on Table 3, Panel B, and Table 3, Panel C provides descriptive statistics by cell for the covariates in the study. I also test whether participant type (PartType, undergraduate accounting or full-time MBA students) influenced the measures of the dependent variable. 13 In a general linear model with Time as the dependent variable and Priming, PriorSuccess, and PartType as independent variables, the p-value (two-tailed) for the three-way interaction of PartType with the experimental manipulations is .10. The natural log of Time (LNTime), however, is used for the hypothesis tests (discussed later) and using either LNTime or #TimesContinued as the dependent variable in the general linear model no interaction with PartType is significant (p’s > .10). Further, there are no significant interactions with PartType (p’s > .10) when the covariates are added to the general linear model using Time, LNTime, or #TimesContinued as the dependent variable. Thus, PartType is not included in the tests of the hypothesis. 12 Interest in the experiment could proxy for participants’ general interest in the task and/or how their interest is affected from the experimental treatment. Therefore, including this variable as a covariate could remove some of the treatment effect which would bias against significant findings. For accounting courses and knowledge of ABC, differences in interpretation of these questions could have driven some of the cell differences. For example, even though the mean number of accounting courses taken was less than four, five participants noted that they had taken 10 or more accounting courses and four of these students were in cell four, potentially increasing the cell mean differences for this variable. 13 The psychology theory that forms the basis for my hypothesis does not require specialized knowledge or experience and thus should apply to a diverse population. 23 The first covariate is the time spent on the experimental materials before participants view the feedback report containing the experimental manipulations, including the time spent on the first cost-driver choice. This measure captures a variety of factors other than the experimental manipulations that could affect the two dependent variables such as reading speed, facility in using an unfamiliar website, time pressure (e.g., whether the participant wanted to finish quickly in order to go on to another activity), and intrinsic motivation (unaffected by the manipulations) to perform the task. It is measured as the total time spent on the experiment before viewing the feedback screen (TimeBefore). Accounting knowledge and work experience also have a significant effect on the dependent variable. Dearman and Shields (2001) shows that an individual’s level of accounting knowledge, and more specifically ABC knowledge, can influence costing decisions. Further, work experience can affect the involvement and familiarity that participants have with costing systems, which in turn can affect the time needed or taken to make the cost driver decisions in the experiment. Accounting knowledge is measured as the number of accounting courses completed (AcctgCourses), and ABC knowledge (ABCNonClass) is captured from a self-rated question asking participants, “How familiar are you with activity-based costing in a nonclassroom setting?” This rating is on a 10-point Likert scale with 1 = Extremely Unfamiliar and 10 = Extremely Familiar. 14 Work experience (WorkExp) is measured as participants’ number of months of professional experience. 14 Participants are also asked on the same 10-point Likert scale “How familiar are you with activity-based costing in a classroom setting?” This variable, ABCClass, is not significantly related to the dependent variable in any of the analyses performed and does not change any inferences when included. Thus, it is not included in the tests of the hypothesis. 24 Lastly, how interesting participants find the experiment can have an impact on the amount of time they spend on the experimental task and the required cost-driver decisions. This measure, Interesting, is also captured through a self-rating by asking participants on a 10-point Likert scale, “How interesting did you find this exercise?” with end points as 1 = Extremely Uninteresting and 10 = Extremely Interesting. This self-rating of interest in the exercise can potentially be affected by the experimental manipulations. To remove this influence I regressed Interesting on the independent variables Priming and PriorSuccess and saved the standardized residual from this regression. This residual, InterestRes, was then included in the tests of the hypothesis to control for interest in the exercise unrelated to the experimental manipulations. Hypothesis Tests Time Figure 3 reports the marginal means of Time, adjusted for the covariates of TimeBefore, AcctgCourses, ABCNonClass, WorkExp, and InterestRes. 15 The form of the interaction is consistent with my prediction, however, the distribution of Time is slightly skewed with the potential for outlier influence, and thus the natural log of Time (LNTime) is used in the tests of the hypothesis. Figure 4 reports the marginal means of LNTime, adjusted for the covariates of TimeBefore, AcctgCourses, ABCNonClass, WorkExp, and InterestRes. To first test the significance of the interaction I use a general linear model (ANOVA) in which the dependent variable is LNTime and the independent variables are Priming and PriorSuccess, without the covariates included (untabulated). In this model the main effects of Priming and PriorSuccess are not significant (p’s > .10, two-tailed), but their interaction is 15 Marginal means adjusted for covariates are means of the dependent variable as predicted by the empirical model used in hypothesis testing, with the covariates held at their mean levels. 25 marginally significant (p = .08, one-tailed), supporting the hypothesis. This model has an Rsquared of .017. To further test the significance of the interaction I use a general linear model (ANCOVA) in which the dependent variable is LNTime, the independent variables are Priming and PriorSuccess, and the covariates are TimeBefore, AcctgCourses, ABCNonClass, WorkExp, and InterestRes (Table 4, Panel A). In this model the main effects of Priming and PriorSuccess are again not significant (p’s > .10, two-tailed), but their interaction is significant (p = .01, onetailed, partial η2 = .052), which provides support for the hypothesis. The ANCOVA shows that the covariates do have significant effects on the dependent variable and this model has an Rsquared of .282, demonstrating a substantial increase in explanatory power over the simple ANOVA. Thus, the covariates are included in the remaining tests of the hypothesis. The hypothesis makes specific pairwise predictions and thus pairwise contrasts within the general linear model are used to test these predictions. P-values of the pairwise tests are 16 Bonferroni-adjusted to provide a family-wise alpha of .05 for the four tests. Table 4, Panel B provides a summary of the pairwise contrast results. My hypothesis predicts that when the highlevel goal of improving accounting for making better decisions is primed, prior MAS revision success will lead to more effort than will prior failure, but when this goal is not primed prior MAS revision failure will lead to more effort than will prior success. Results show that, as predicted, when the goal is primed participants spend more time on subsequent MAS revision following prior success (adjusted marginal mean of the natural log of time in seconds = 4.948) than they do following prior failure (mean = 4.621). This difference is significant (p = .03, onetailed, Cohen’s d = .52). Also as predicted, when the goal is not primed, however, participants spend more time on subsequent MAS revision following prior failure (mean = 4.916) than they 16 That is, the original p-values are multiplied by four, providing more conservative p-values than if family-wise error was not considered. 26 do following prior success (mean = 4.674). This difference is marginally significant (p = .07, one-tailed, d = .39). I also predict that when the prior MAS revision is a success, priming the high-level goal of improving accounting for making better decisions will lead to more effort than will not priming this goal, whereas when the prior MAS revision is not a success not priming the goal of improving accounting for making better decisions will lead to more effort than will priming this goal. As predicted, when the prior MAS revision is a success, participants spend significantly more time on subsequent MAS revision (p = .05, one-tailed, d = .44) when the goal of improving accounting for making better decisions is primed (mean = 4.948) than when this goal is not primed (mean = 4.674). Conversely, also as predicted following prior MAS revision failure, participants spend significantly more time on subsequent MAS revision (p = .04, one-tailed, d = .47) when the goal is not primed (mean = 4.916) than when the goal is primed (mean = 4.621). Number of Times Continued 17 Figure 5 reports the marginal means of #TimesContinued, adjusted for the covariates TimeBefore, AcctgCourses, ABCNonClass, WorkExp, and InterestRes. The form of the interaction is consistent with my hypothesis. To test the significance of the interaction I use a general linear model with #TimesContinued as the dependent variable, Priming and PriorSuccess as independent variables, and TimeBefore, AcctgCourses, ABCNonClass, WorkExp, and InterestRes as covariates (Table 5, Panel A). Similar to the model using LNTime as the dependent variable, the main effects of Priming and PriorSuccess are not significant (p’s > .10, 17 #TimesContinued has a relatively small range as participants’ continuation choices can range from zero to three. 27 two-tailed), but their interaction is significant (p = .03, one-tailed, partial η2 = .031), which provides support for my hypothesis. 18 Pairwise contrasts are also performed with this measure of the dependent variable to test the pairwise predictions, with p-values of these tests Bonferroni-adjusted to provide a familywise alpha of .05 for the four tests. Table 5, Panel B provides a summary of the pairwise contrast results with #TimesContinued. Consistent with the predicted effect, when the high-level goal of improving accounting for making better decisions is primed participants choose to continue more with subsequent MAS revision following prior revision success (mean = 1.43) than prior failure (mean = 0.98), and the difference is marginally significant (p = .10, one-tailed, d = .35). Also as predicted, when the goal of improving accounting for making better decisions is not primed individuals choose to continue more with MAS revision following prior revision failure (mean = 1.71) than prior success (mean = 1.28) and this difference is marginally significant as well (p = .10, one-tailed, d = .34). Consistent with my predictions, when prior MAS revision is a success, participants choose to continue more with subsequent MAS revision when the goal is primed (mean = 1.43) than when it is not primed (mean = 1.28); but this difference is not statistically significant (p = .32, one-tailed, d = .12). On the other hand, when prior MAS revision is a failure participants choose to continue more with subsequent MAS revision when the goal of improving accounting for making better decisions is not primed (mean = 1.71) than when it is primed (mean = 0.98), and this difference is significant (p = .02, one-tailed, d = .57), supporting the prediction. The 18 Due to the fact that #TimesContinued is an ordinal variable with a restricted range (four possible values) I also use an ordered logit model to test the hypothesis using #TimesContinued as the dependent variable, Priming and PriorSuccess as independent variables, and TimeBefore, AcctgCourses, ABCNonClass, WorkExp, and InterestRes as covariates. In this regression model Priming (p = .01, one-tailed), PriorSuccess (p = .09, one-tailed), and their interaction (p = .02, one-tailed) are all significant (at least marginally), providing further support for the hypothesis. 28 pairwise comparison results using #TimesContinued are somewhat weaker in support of the hypothesis than the results using LNTime, possibly due to the restricted range of #TimesContinued (four possible values ranging from zero to three). Even with the restricted range, however, the interaction is significant of the predicted form and three of the four pairwise comparisons are (at least marginally) significant after the Bonferroni adjustment. 29 CHAPTER 5: CONCLUSION Successful MAS revision can improve decision-making and organization performance, and research studies have examined different determinants of successful implementation of new MAS. Research, however, has not fully addressed the effects of non-monetary factors on individual motivation to exert effort on MAS revision. One suggested strategy is to motivate MAS revision by seeking and providing feedback on short-run successes. The motivating effects of this strategy have not been directly tested. The results of my study provide strong support for my hypothesis that priming the highlevel goal of improving accounting for making better decisions and providing accounting feedback concerning prior success of MAS revision interact to affect subsequent MAS revision effort. These findings provide evidence that the strategy of encouraging and reporting short-run success in MAS revision can be successful in motivating future effort in MAS revision when those involved are primed with the high-level goal of the project. The results also indicate, however, that when the high-level goal is not primed, the strategy of encouraging and reporting short-run success can actually backfire and demotivate continued efforts in MAS revision. Further, there is evidence that language accompanying accounting feedback reports (e.g., an accounting slogan) that activates the high-level goal of the project can be beneficial to organizations during MAS revision to promote continued effort in the revision process if prior revision is successful. If those involved believe that their prior revision efforts were a failure, however, this same language can be counterproductive, resulting in less subsequent effort. This study provides insight to managers on motivating employees to exert effort towards MAS revision to support better decision-making within the organization. As a revision project proceeds successfully managers can motivate those involved by reporting this success and 30 activating the project’s overall goal (e.g., including a slogan related to the goal). If the project is not proceeding successfully, however, managers can motivate continued effort in the revision process by reporting this feedback while focusing on the task at hand to keep the short-term goal at the forefront of employees’ minds. My study also makes three contributions to the accounting literature. First, it provides theory and evidence on how reporting small-scale and/or short-run success can affect individuals’ motivation to exert additional effort on MAS revision projects either positively or negatively. This reporting strategy can be effective when employees are primed with the project’s high-level goal, but without such a prime this strategy can backfire and lead to a reduction in effort on MAS revision as individuals focus more on lower-level goals. Second, my study adds to the literature on accounting report design. Prior studies (e.g., Maines and McDaniel 2000; Cardinaels 2008; Jiang et al. 2012) have examined how report presentation and format, including the addition of graphics in reports, can influence decision making by affecting how individuals direct their attention and weight information. My study examines how including a prime—a slogan—on an accounting feedback report can affect effort, providing evidence that feedback about prior revision success and goal activation together can increase or decrease motivation to exert additional effort in revising MAS. Third, my study contributes to the emerging literature on priming in accounting. Research in accounting (e.g., Hammersley et al. 2010; Lambert and Agoglia 2011) has provided evidence of priming effects, generally evaluating primes essential to the task and requiring cognitive processing. My study examines goal-priming and shows that a peripheral stimulus (prime), not requiring attention for the task to be complete (i.e., an accounting slogan), can unconsciously affect whether individuals either increase or decrease their effort on MAS revision. 31 A limitation of my study is that it evaluates effort exerted on MAS revision, which can be a long-run process, using a laboratory setting of shorter duration. There has been evidence, however, that results from short-run experiments in a laboratory setting are consistent with decisions and behavior displayed over a longer time period outside of the laboratory. For example, Norman, Requate, and Waichman (2012) compare the generalizability of short-run experiments lasting about an hour in duration to treatments lasting a month and find no significant differences in decisions. Generalizability of short-run findings to treatments of longer durations has also been shown with unconscious effects similar to the priming effect examined in my study. In a laboratory setting, Stajkovic, Locke, and Blair (2006) provide evidence that unconsciously priming a goal leads to higher task performance. To evaluate whether these results can generalize to the workplace and last for a longer duration, Shantz and Latham (2009) and Latham and Piccolo (2012) test similar priming effects using employees in a call center. The authors find comparable results, showing that the unconscious priming of a goal can lead to significantly higher task performance in a workplace setting over a longer duration. In my study I examine continued individual effort on MAS revision projects following feedback on prior revision efforts. There are opportunities for future research on this topic. First, these projects can involve diverse individuals working as a team to revise the MAS (Shields and Young 2000; Wouters and Roijmans 2011) and studies have demonstrated that the involvement of accountants and non-accountants in MAS revision can lead to system improvements, increased user satisfaction, and organization performance (Hunton and Gibson 1999; Wouters 2009; Eldenburg et al. 2010; Chapman and Kern 2011). Future research might examine the effect that priming a high-level accounting goal and feedback on prior success can have on team effort to revise MAS and how team composition (e.g., accountants, managers, or consultants) can 32 influence this effect. Second, during a MAS revision project individuals can receive multiple feedback reports concerning the success of prior revisions. Studies in psychology have shown that both the direction and the velocity (or rate) of progress towards (or away from) a goal can influence individual motivation (Hsee and Ableson 1991; Kluger and DeNisi 1996). Future research could examine the effect that multiple feedback reports during the revision process with varying levels of performance success (or failure) can have on continued effort on MAS revision. 33 APPENDICES 34 APPENDIX A Figures 1 - 5 35 Figure 1 Theoretical Predictions Future Effort on MAS Revision 4 1 High-Level Goal of Improving Accounting for Better Decisions Primed 3 2 Not Primed Failure Success Accounting Feedback H: Priming of the high-level goal will interact with the success of prior MAS revision to affect the effort exerted on subsequent MAS revision such that: a) When the high-level goal of improving accounting for making better decisions is primed, prior MAS revision success will lead to more effort than prior failure (1 > 3), b) When the high-level goal is not primed, prior MAS revision failure will lead to more effort than prior success (4 > 2), c) When the prior MAS revision is a success, priming the high-level goal will lead to more effort than not priming the goal (1 > 2), d) When the prior MAS revision is a failure, not priming the high-level goal will lead to more effort than priming the goal (4 > 3). 36 Figure 2 Experimental Task Event Line Managers view background information and Michigan’s cost accounting policy, which includes the accounting slogan. A competitor begins to offer a lower price for a comparable product to Michigan’s basic product, Green. Top management asks managers to choose a better cost driver for the Painting activity, to determine if costs are being over-allocated to Green. Managers choose a cost driver based on production and overhead information. Managers view a feedback report showing their cost driver choice to be a success or failure. (Half include the accounting slogan and half do not.) (This figure is continued on the next page.) 37 Figure 2 (cont’d) Top management asks managers to make cost driver decisions for the Fabrication and Finishing activities. Managers select a cost driver for the Fabrication activity. Managers choose whether to use direct labor hours for both activities. Managers choose whether to end or continue their involvement in revising the costing system. Yes No End Managers choose whether to use machine hours for both activities, recommended by a new manager. Yes Continue Managers select a cost driver for the Finishing activity. No Post-Experiment Questions 38 Figure 3 Marginal Means (Adjusted for Covariates) of Future Effort on MAS Revision Following Initial Cost Driver Choice Measured as Time Spent on MAS Revision (Time) Time (in seconds) Spent on MAS Revision 220 180 170.54 160.82 140 136.60 High-Level Goal of Improving Accounting for Better Decisions Primed Not Primed 123.01 100 60 Failure Success Accounting Feedback Covariates are TimeBefore, AcctgCourses, ABCNonClass, WorkExp, and InterestRes. 39 Natural Log of Time (in seconds) Spent on MAS Revision Figure 4 Marginal Means (Adjusted for Covariates) of Future Effort on MAS Revision Following Initial Cost Driver Choice Measured as the Natural Log of Time Spent on MAS Revision (LNTime) 5.3 4.9 4.916 4.948 4.674 4.5 4.621 High-Level Goal of Improving Accounting for Better Decisions Primed Not Primed 4.1 Failure Success Accounting Feedback Covariates are TimeBefore, AcctgCourses, ABCNonClass, WorkExp, and InterestRes. 40 Number of Times Continued Revising MAS Figure 5 Marginal Means (Adjusted for Covariates) of Future Effort on MAS Revision Following Initial Cost Driver Choice Measured as the Number of Times Continued Revising MAS (#TimesContinued) 2.40 1.43 High-Level Goal of Improving Accounting for Better Decisions 1.28 Primed 1.71 1.60 0.80 Not Primed 0.98 0.00 Failure Success Accounting Feedback Covariates are TimeBefore, AcctgCourses, ABCNonClass, WorkExp, and InterestRes. 41 APPENDIX B Tables 1 - 5 42 “Better Accounting for Better Decisions” Table 1 Feedback Following Cost Driver Choice for the Painting Activity Panel A: Feedback Received if Cost Driver Choice Was Successful Management accountants have collected sample observations of the Painting activity to test whether your chosen cost driver of (Selected Cost Driver) is related to the use of manufacturing overhead resources in this activity. The quantity of (Selected Cost Driver) has been documented over the last several weeks while monitoring the quantity of manufacturing overhead resources used. Actual resource usage was recorded and calculated in dollars for ease of comparison. During these observations it was noted that the cost driver increased by 10%, and thus the corresponding observations for overhead resource use were compared to determine whether the overhead costs changed as the cost driver changed. Indirect Cost Category Indirect Labor Supplies Machining Maintenance Utilities Other Resource Use at Initial Level of Cost Driver $750 $2,000 $4,500 $1,800 $1,500 $1,250 Resource Use at 10% Higher Level of Cost Driver $800 $2,200 $5,000 $2,075 $1,650 $1,350 Percent Change in Cost of Indirect Resources Used 7% 10% 11% 15% 10% 8% From this test data the management accountants believe that (Selected Cost Driver) do relate to manufacturing overhead costs in the Painting activity and should be used to allocate these costs to the two products. Using this cost driver to allocate manufacturing overhead costs does reduce the estimated cost per unit for Green, and suggests that a competitive reduction in price is possible, while still making a profit. (This table is continued on the next page.) 43 Table 1 (cont’d) “Better Accounting for Better Decisions” Panel B: Feedback Received if Cost Driver Choice Was Unsuccessful Management accountants have collected sample observations of the Painting activity to determine whether your chosen cost driver of (Selected Cost Driver) is related to the use of manufacturing overhead resources in this activity. The amount of (Selected Cost Driver) has been documented over the last several weeks while monitoring the amount of manufacturing overhead resource used. Actual resource usage was recorded and the corresponding costs were calculated in dollars for ease of comparison and relative significance. During these observations it was noted that the cost driver increased by 10%, and thus the corresponding observations for overhead resource use were compared to determine whether the overhead costs changed as the cost driver changed. Indirect Cost Category Indirect Labor Supplies Machining Maintenance Utilities Other Resource Use at Initial Level of Cost Driver $750 $2,000 $4,500 $1,800 $1,500 $1,250 Resource Use at 10% Higher Level of Cost Driver $1,000 $1,950 $4,500 $2,000 $1,300 $1,800 Percent Change in Cost of Indirect Resources Used 33% -3% 0% 11% -13% 44% From this test data the management accountants believe that (Selected Cost Driver) do not relate to manufacturing overhead costs in the Painting activity and should not be used to allocate these costs to the two products. Thus, the estimated cost per unit for Green has not changed, providing no information as to whether a competitive reduction in price is possible, while still making a profit. In the accountants’ opinion it appears that one cost driver will not sufficiently relate to all of these manufacturing overhead costs, and thus this activity would need to be split into separate activity cost pools with separate cost drivers. This, however, is not feasible in the immediate future, since this would entail a significant revision of Michigan’s information systems. ______________________________________ Participants receive the feedback in either Panel A or B depending on feedback condition. If participants are in a condition with priming, then the feedback report they receive includes the slogan “Better Accounting for Better Decisions” as illustrated. 44 Table 2 Cell Means (Standard Deviations) of Future Effort on MAS Revision Panel A: Time PriorSuccess Primed Priming Not Primed Success 161.49 (108.08) n = 30 Failure 130.33 (67.55) n = 28 141.70 (90.55) n = 30 158.01 (81.70) n = 30 Panel B: #TimesContinued PriorSuccess Success Failure Primed 1.30 (1.34) n = 30 1.14 (1.33) n = 28 Not Primed 1.30 (1.32) n = 30 1.67 (1.22) n = 30 Priming 45 Table 3 Correlations and Descriptive Statistics of the Two Measures of the Dependent Variable, Time and #TimesContinued, and the Covariates Panel A: Correlation Matrix Time Time #TimesContinued 0.546 TimeBefore 0.295 AcctgCourses 0.093 ABCNonClass -0.026 WorkExp -0.205 Interesting 0.290 InterestRes 0.293 #TimesContinued TimeBefore AcctgCourses ABCNonClass WorkExp Interesting 0.075 0.111 -0.127 -0.095 0.191 0.203 -0.189 0.136 -0.092 0.145 0.131 0.073 -0.231 -0.083 -0.049 -0.046 0.228 0.217 All bolded correlations are significant at the 0.05 level (two-tailed). Variables Time Definition Time in seconds spent on future MAS Revision. #TimesContinued The number of times participants continue revising the MAS. TimeBefore Time spent on the experimental materials before feedback. AcctgCourses The number of accounting courses completed. Self-rating of “How familiar are you with activity-based costing in a non-classroom setting?” (Likert scale, 1-10). The number of months of professional experience. ABCNonClass WorkExp Interesting Self-rating of “How interesting did you find this exercise?” (Likert scale, 1-10). InterestRes Residual of Interesting regressed on Priming and PriorSuccess. (This table is continued on the next page.) 46 -0.002 0.013 0.978 Table 3 (cont’d) Panel B: Descriptive Statistics Variables Time #TimesContinued TimeBefore AcctgCourses ABCNonClass WorkExp Interesting Mean 148.18 1.36 476.77 3.59 2.59 23.66 5.95 Median 142.35 1.00 458.41 3.00 2.00 6.50 6.00 SD 88.27 1.30 139.38 2.98 1.91 27.96 2.15 Min 13.98 0 113.21 0 1 0 1 Max 469.78 3 969.49 20 9 108 10 Panel C: Cell Means (Standard Deviations) of Covariates Primed Not Primed Primed Not Primed TimeBefore Success Failure 472.54 463.99 (125.23) (118.65) 498.63 471.06 (124.71) (182.58) WorkExp Success Failure 24.60 24.91 (28.56) (27.02) 19.60 25.60 (27.84) (29.31) Primed Not Primed Primed Not Primed AcctgCourses Success Failure 2.83 4.07 (2.15) (3.87) 3.23 4.27 (2.64) (2.97) Interesting Success Failure 6.20 5.50 (2.02) (2.05) 6.53 5.53 (2.29) (2.15) 47 Primed Not Primed ABCNonClass Success Failure 3.03 1.75 (2.36) (1.24) 2.53 3.00 (2.15) (1.44) Table 4 Future Effort on MAS Revision Measured as the Natural Log of Time Spent on MAS Revision (LNTime) Panel A: General Linear Model Source Model Intercept Priming PriorSuccess Priming * PriorSuccess TimeBefore AcctgCourses ABCNonClass WorkExp InterestRes Error df 8 1 1 1 1 1 1 1 1 1 109 Mean Square 2.01 F 5.35 Sig. <.01 103.01 0.00 0.05 2.23 4.62 1.74 2.71 1.00 6.30 274.31 0.01 0.13 5.94 12.31 4.64 7.22 2.66 16.76 <.01 .93 .72 .01 <.01 .03 .01 .11 <.01 Panel B: Pairwise Contrasts within the General Linear Model Predictions Time on Subsequent Revision P-Value, One-Tailed Accounting Prime and Prior Success (1) > Accounting Prime and Prior Failure (3) No Accounting Prime and Prior Failure (4) > No Accounting Prime and Prior Success (2) Prior Success and Accounting Prime (1) > Prior Success and No Accounting Prime (2) 4.948 > 4.621 4.916 > 4.674 4.948 > 4.674 0.03 0.07 0.05 Prior Failure and No Accounting Prime (4) > Prior Failure and Accounting Prime (3) 4.916 > 4.621 0.04 P-values for all pairwise comparisons are Bonferroni-adjusted. 48 Table 5 Future Effort on MAS Revision Measured as the Number of Times Continued Revising MAS (#TimesContinued) Panel A: General Linear Model Source Corrected Model Intercept Priming df 8 1 1 Mean Square 3.50 8.91 2.40 PriorSuccess Priming * PriorSuccess TimeBefore AcctgCourses ABCNonClass 1 1 1 1 1 0.00 5.33 1.63 3.30 11.23 0.00 3.44 1.05 2.13 7.24 .97 .03 .31 .15 .01 1 1 109 0.97 11.92 1.55 0.62 7.69 .43 .01 WorkExp ZRE_IntNew Error F 2.26 5.74 1.54 Sig. .03 .02 .22 Panel B: Pairwise Contrasts within the General Linear Model Predictions Number of Times Continued P-Value, One-Tailed Accounting Prime and Prior Success (1) > Accounting Prime and Prior Failure (3) No Accounting Prime and Prior Failure (4) > No Accounting Prime and Prior Success (2) Prior Success and Accounting Prime (1) > Prior Success and No Accounting Prime (2) 1.43 > 0.98 1.71 > 1.28 1.43 > 1.28 0.10 0.10 0.32 Prior Failure and No Accounting Prime (4) > Prior Failure and Accounting Prime (3) 1.71 > 0.98 0.02 P-values for all pairwise comparisons are Bonferroni-adjusted. 49 APPENDIX C Experimental Task 50 Instructions In this decision-making exercise, you will assume the role of a manager making decisions based on information for a manufacturing firm. Please perform this task carefully as you make your decisions. Different participants in this exercise will be making different decisions than you will be, so do not be surprised to see others completing this exercise before or after you do. Please follow instructions carefully and work independently. Please fully consider each page and decision before clicking Next as you will not be able to go back after clicking Next. On the sheet of paper given to you, please write your chosen ID and keep the sheet of paper to remember your ID. You will use this ID later to collect your pay. Enter your chosen ID: . 51 Background Information You are a production manager at Michigan, Inc. (Michigan), an industrial firm based in the Midwest. As a production manager you are responsible for the production process, including personnel, inventory, equipment, and production output. You coordinate personnel needs for the production line and are required to make adjustments to personnel as needed. You track inventory levels, monitor production equipment, keeping it maintained and in proper operation, and ensure that products meet specifications throughout the production process. Further, much of your time is spent in discussions with customers concerning shipments, issues with products, special order requests, etc. Michigan manufactures two products, Green and White, for finished-goods suppliers. Green has been Michigan’s primary product for many years. White is a more complex product that Michigan has been producing for only a few years. Michigan contracts with its customers to supply these products, setting prices based on a percentage above the full production cost of each product. These products pass through three main activities during the production process. First, in the Painting activity direct materials purchased from suppliers are painted with a primer, then in the Fabrication activity these materials are combined to form the final product, and finally in the Finishing activity they receive finishing coats of paint and final inspections before being shipped to customers. 52 Figure 6 Michigan’s Cost Accounting Policy Better Accounting for Better Decisions Michigan’s Cost Accounting Policy Michigan strives for accurate cost accounting to provide high-quality cost and profit measures for decision-making throughout the firm including: inventory valuation, product pricing, budgeting, performance measurement, and performance evaluation. To be accurate, product costs should be based on resource consumption by products, with more overhead costs allocated to products that consume higher amounts of overhead resources. Accurate cost accounting measures are important as they can affect firm performance, as shown below. Currently, manufacturing overhead costs incurred during the manufacturing process are allocated to products based on direct labor costs. That is, total manufacturing overhead costs (such as supervision, supplies used, machining, etc.) are divided by total direct labor costs to determine the allocation rate. This rate is multiplied by the direct labor costs used by a specific product to determine the amount of manufacturing overhead to be allocated to that product. The full product cost is the sum of direct material costs, direct labor costs, and the allocated manufacturing overhead costs. Example An example of cost accounting affecting firm performance is that product costs estimated too low could encourage managers to set product prices lower, potentially resulting in winning unprofitable contracts, or product costs estimated too high could make managers believe they need to set product prices higher, potentially resulting in the loss of profitable contracts. For interpretation of the references to color in this and all other figures, the reader is referred to the electronic version of this dissertation. 53 Background Information Michigan has had a strong market presence for many years, which has allowed it to maintain its policy of basing products’ selling prices on full production cost plus a standard markup. Recently, Michigan has become aware that a competitor, Ohio, Inc. (Ohio), will be bidding to supply one of Michigan’s large customers with a product similar to Green when Michigan’s contract ends at the end of the current year. Available market information indicates that Ohio will be offering a price lower than Michigan’s current price. Because this contract is important to Michigan, top management has decided to reduce the price of Green for this customer in order to keep the contract for next year. Ohio offers comparable prices in other markets and makes a profit, so it can likely sustain its low price. Therefore top management at Michigan is considering: how can we compete profitably in the longer term? Can we still make a profit while aggressively meeting Ohio's prices, or should we move quickly into the less competitive markets for more complex products like White? One important factor in answering these questions is a better understanding of product costs at Michigan. It is possible that because Ohio focuses on a single product, it is significantly more efficient at making this product than Michigan is. It is also possible, however, that Michigan is equally efficient but is misestimating its own cost of producing Green, because it has not devoted much attention to how it allocates manufacturing overhead costs between Green and White. If Green is actually as costly, or even more costly, to produce than current cost estimates indicate, then Michigan must re-examine its strategy in terms of manufacturing practices, product mix, and product prices. If Green is actually less costly to produce than current cost estimates indicate, then Michigan can profitably pursue an aggressive pricing policy on Green, and can consider the possibility of increasing prices in the less competitive White market. 54 Background Information Currently, all manufacturing overhead costs are allocated using direct labor costs. It is uncertain, however, whether the use of manufacturing overhead resources really changes as direct labor costs change. It is also uncertain what other cost drivers–if any–would do a better job of predicting products’ use of these overhead resources. Creating a new costing system would be a long and costly process and would not produce the information top management needs as soon as they would like to have it. It may be possible, however, to get more accurate product cost estimates in the near term by choosing a different cost driver on which information is already collected and using it to allocate overhead costs. 55 Background Information Due to your familiarity with the production process as the production manager, top management has asked you to choose a better cost driver for one of the manufacturing activities, the Painting activity. This activity has been selected since it uses significantly more manufacturing overhead resources compared to the other activities, and there appears to be a lot of diversity in the overhead resources used by Green and White in this activity. Top management thus believes that changes in how overhead costs are allocated in this activity could significantly improve cost estimation for Michigan’s two products. Once you recommend a cost driver, a team from the accounting department will then test the usefulness of this driver by observing production closely and determining whether actual overhead resource use (indirect labor, supplies, maintenance, machining, etc.) changes as this cost driver changes. It is possible that this test will show that the new cost driver will more accurately allocate costs, and thus provide a better estimate of product costs. It is also possible, however, that it will not. An accurate cost driver for this activity might require information that the manufacturing information system does not currently capture. Or it might be that the overhead costs incurred in this activity are diverse and do not increase or decrease together with any one cost driver. Dividing this activity cost pool into more activity cost pools and tracking these activity cost pools and cost drivers is not feasible in the immediate future, because this would require significant revision of Michigan’s information systems. You have been presented with the budgeted costs of direct materials, direct labor, and manufacturing overhead for the Painting activity in the coming year. Please open envelope #1, then click Next. 56 Table 6 Production Process and Overhead Cost Information for the Painting Activity Painting Activity Each product is composed of several pieces that are molded by suppliers. In the Painting activity molded pieces receive priming coats that are machine-applied to each piece. This activity is not labor intensive. Similar pieces can be painted in sequence without adjustments to the machines. Molds differ in the type and number of prime coats that must be applied, however, and machines must be stopped and set up more often for molds requiring special prime coats. Further, molds requiring more priming coats require more machine time for painting to ensure that the coats are evenly applied to the entire piece. Finally, the pieces are inspected for any blemishes. Green White Total Number of Molds (per unit) 6 9 15 Priming Coats (per unit) 18 30 48 Painting Machine Hours (per 100 units) 2.50 5 7.50 Painting Direct Labor Hours (per 100 units) 1.50 1.50 3 Painting Setups (Green – Batch of 750 units White – Batch of 250 units) 14 12 26 0.75 1.25 2 Painting Inspection Hours (per unit) (This table is continued on the next page.) 57 Table 6 (cont’d) DM Painting Green $236,500 White $145,850 DL Green $51,750 White $14,250 Manufacturing Overhead: Painting Indirect Labor $13,400 Supplies Machining Maintenance Utilities $45,000 $84,000 $32,500 $25,000 Other Overhead $31,750 Total $231,650 These budgeted costs are based on an estimated production of 10,000 units of Green and 3,200 units of White, which is the current production volume and is approximately the total capacity usage desired by top management. 58 Please select the ONE indirect cost driver that you believe will best allocate Painting overhead costs to products. POSSIBLE COST DRIVERS Number of Molds Number of Priming Coats Painting Machine Hours Painting Direct Labor Hours Number of Painting Setups Painting Inspection Hours 59 Table 7 Accounting Feedback Reports with Priming and Prior Success Manipulations “Better Accounting for Better Decisions” Management accountants have collected sample observations of the Painting activity to test whether your chosen cost driver of (Selected Cost Driver) is related to the use of manufacturing overhead resources in this activity. The quantity of (Selected Cost Driver) has been documented over the last several weeks while monitoring the quantity of manufacturing overhead resources used. Actual resource usage was recorded and calculated in dollars for ease of comparison. During these observations it was noted that the cost driver increased by 10%, and thus the corresponding observations for overhead resource use were compared to determine whether the overhead costs changed as the cost driver changed. Indirect Cost Category Indirect Labor Supplies Machining Maintenance Utilities Other Resource Use at Initial Level of Cost Driver $750 $2,000 $4,500 $1,800 $1,500 $1,250 Resource Use at 10% Higher Level of Cost Driver $800 $2,200 $5,000 $2,075 $1,650 $1,350 Percent Change in Cost of Indirect Resources Used 7% 10% 11% 15% 10% 8% From this test data the management accountants believe that (Selected Cost Driver) do relate to manufacturing overhead costs in the Painting activity and should be used to allocate these costs to the two products. Using this cost driver to allocate manufacturing overhead costs does reduce the estimated cost per unit for Green, and suggests that a competitive reduction in price is possible, while still making a profit. _____________________________________ Participants in the Prior Success/Primed condition receive this feedback report which includes the slogan “Better Accounting for Better Decisions” as illustrated. 60 Table 7 (cont’d) “Better Accounting for Better Decisions” Management accountants have collected sample observations of the Painting activity to determine whether your chosen cost driver of (Selected Cost Driver) is related to the use of manufacturing overhead resources in this activity. The amount of (Selected Cost Driver) has been documented over the last several weeks while monitoring the amount of manufacturing overhead resource used. Actual resource usage was recorded and the corresponding costs were calculated in dollars for ease of comparison and relative significance. During these observations it was noted that the cost driver increased by 10%, and thus the corresponding observations for overhead resource use were compared to determine whether the overhead costs changed as the cost driver changed. Indirect Cost Category Indirect Labor Supplies Machining Maintenance Utilities Other Resource Use at Initial Level of Cost Driver $750 $2,000 $4,500 $1,800 $1,500 $1,250 Resource Use at 10% Higher Level of Cost Driver $1,000 $1,950 $4,500 $2,000 $1,300 $1,800 Percent Change in Cost of Indirect Resources Used 33% -3% 0% 11% -13% 44% From this test data the management accountants believe that (Selected Cost Driver) do not relate to manufacturing overhead costs in the Painting activity and should not be used to allocate these costs to the two products. Thus, the estimated cost per unit for Green has not changed, providing no information as to whether a competitive reduction in price is possible, while still making a profit. In the accountants’ opinion it appears that one cost driver will not sufficiently relate to all of these manufacturing overhead costs, and thus this activity would need to be split into separate activity cost pools with separate cost drivers. This, however, is not feasible in the immediate future, since this would entail a significant revision of Michigan’s information systems. ______________________________________ Participants in the Prior Failure/Primed condition receive this feedback report which includes the slogan “Better Accounting for Better Decisions” as illustrated. 61 Table 7 (cont’d) Management accountants have collected sample observations of the Painting activity to test whether your chosen cost driver of (Selected Cost Driver) is related to the use of manufacturing overhead resources in this activity. The quantity of (Selected Cost Driver) has been documented over the last several weeks while monitoring the quantity of manufacturing overhead resources used. Actual resource usage was recorded and calculated in dollars for ease of comparison. During these observations it was noted that the cost driver increased by 10%, and thus the corresponding observations for overhead resource use were compared to determine whether the overhead costs changed as the cost driver changed. Indirect Cost Category Indirect Labor Supplies Machining Maintenance Utilities Other Resource Use at Initial Level of Cost Driver $750 $2,000 $4,500 $1,800 $1,500 $1,250 Resource Use at 10% Higher Level of Cost Driver $800 $2,200 $5,000 $2,075 $1,650 $1,350 Percent Change in Cost of Indirect Resources Used 7% 10% 11% 15% 10% 8% From this test data the management accountants believe that (Selected Cost Driver) do relate to manufacturing overhead costs in the Painting activity and should be used to allocate these costs to the two products. Using this cost driver to allocate manufacturing overhead costs does reduce the estimated cost per unit for Green, and suggests that a competitive reduction in price is possible, while still making a profit. _____________________________________ Participants in the Prior Success/Not Primed condition receive this feedback report which does not include the slogan “Better Accounting for Better Decisions”. 62 Table 7 (cont’d) Management accountants have collected sample observations of the Painting activity to determine whether your chosen cost driver of (Selected Cost Driver) is related to the use of manufacturing overhead resources in this activity. The amount of (Selected Cost Driver) has been documented over the last several weeks while monitoring the amount of manufacturing overhead resource used. Actual resource usage was recorded and the corresponding costs were calculated in dollars for ease of comparison and relative significance. During these observations it was noted that the cost driver increased by 10%, and thus the corresponding observations for overhead resource use were compared to determine whether the overhead costs changed as the cost driver changed. Indirect Cost Category Indirect Labor Supplies Machining Maintenance Utilities Other Resource Use at Initial Level of Cost Driver $750 $2,000 $4,500 $1,800 $1,500 $1,250 Resource Use at 10% Higher Level of Cost Driver $1,000 $1,950 $4,500 $2,000 $1,300 $1,800 Percent Change in Cost of Indirect Resources Used 33% -3% 0% 11% -13% 44% From this test data the management accountants believe that (Selected Cost Driver) do not relate to manufacturing overhead costs in the Painting activity and should not be used to allocate these costs to the two products. Thus, the estimated cost per unit for Green has not changed, providing no information as to whether a competitive reduction in price is possible, while still making a profit. In the accountants’ opinion it appears that one cost driver will not sufficiently relate to all of these manufacturing overhead costs, and thus this activity would need to be split into separate activity cost pools with separate cost drivers. This, however, is not feasible in the immediate future, since this would entail a significant revision of Michigan’s information systems. ______________________________________ Participants in the Prior Failure/Not Primed condition receive this feedback report which does not include the slogan “Better Accounting for Better Decisions”. 63 Current Responsibilities It has been weeks since the test of the cost driver for the Painting activity and you have returned to your regular responsibilities of managing and monitoring production. You are currently revising and finalizing production schedules, determining personnel needs, working on a project to increase production quality, and evaluating equipment maintenance requirements. You are also dealing with delivery issues and other questions from customers. These duties continue to fill your entire work week. 64 Top management would like to consider whether it would now be valuable to make revisions to how manufacturing overhead costs are allocated in the other two activities in the production process, Fabrication and Finishing. Even though the manufacturing overhead costs in the Painting activity did not relate to the one chosen cost driver, this does not mean it will be true in these other two activities. Selecting new cost drivers for these activities may or may not be beneficial. New cost drivers for these activities could improve cost estimation for Michigan’s products. Cost driver selection, however, takes time which would take you away from your numerous production management responsibilities, and new drivers must be tested, requiring time that could be used elsewhere in the firm. Further, the revision process may not lead to better cost allocations. More accurate cost drivers for these activities might not be available due to data limitations or it might be that the costs incurred in these activities are diverse and cannot be captured with only one cost driver for each activity. Dividing these activity cost pools into more activity cost pools and tracking these activity cost pools and cost drivers is not feasible in the immediate future, since this would require a significant revision to Michigan’s information systems. You have been presented with the budgeted costs of direct materials, direct labor, and manufacturing overhead for these activities in the coming year. Please open envelope #2, then click Next. 65 Table 8 Production Process and Overhead Cost Information for the Fabrication Activity Fabrication Activity Painted molds are received into the Fabrication process. These molds are assembled in this activity in multiple steps. As molds are combined, workers transfer them through different stations until the molds are assembled into a completed product. The more molds that are required for a product, the more stations are used to combine them. Workers combine many of these molds by hand, but some have to be combined using machinery. Setup for these machines is relatively straightforward and does not require much time to complete. Once all of the molds are combined, products are inspected for any defects. Following these inspections, products are transferred to Finishing. Green White Total 6 9 15 1.50 2.50 4 Fabrication Direct Labor Hours (per 100 units) 3 8 11 Fabrication Setups (Green – Batch of 1,750 units White – Batch of 300 units) 6 10 16 Fabrication Inspection Hours (per unit) 1.25 1.75 3 Number of Molds (per unit) Fabrication Machine Hours (per 100 units) (This table is continued on the next page.) 66 Table 8 (cont’d) DM Fabrication Green $60,500 DL White $48,650 Green $128,250 White $51,750 Manufacturing Overhead: Fabrication Indirect Labor $29,250 Supplies $18,500 Machining $51,850 Maintenance $15,750 Utilities $18,500 Other Overhead $12,500 Total $146,350 These budgeted costs are based on an estimated production of 10,000 units of Green and 3,200 units of White, which is the current production volume and is approximately the total capacity usage desired by top management. 67 Table 9 Production Process and Overhead Cost Information for the Finishing Activity Finishing Activity Products transferred from Fabrication receive finishing coats of paint. Workers clean these products to prepare them for finishing, and finishing coats are then machine applied to the cleaned products. Machine attachments are used to apply product-specific finishes. These are easily switched and adjusted so that little setup time is required. Different products require different numbers of finishes, resulting in varying requirements for run time on the finishing machines. After the finishes are applied, the products are set to dry. Once dry, products pass through final inspections and are machine tested to see if the finish coats are smooth and the final products are within specifications. Inspection requires more time for products with more finishes. Once the products clear inspection they are shipped by common carrier to customers. Green White Total Number of Finishes (per unit) 4 7 11 Finishing Machine Hours (per 100 units) 2 3 5 Finishing Direct Labor Hours (per 100 units) 1.50 2 3.50 Finishing Setups (Green – Batch of 2,500 units White – Batch of 500 units) 4 6 10 Finishing Inspection Hours (per unit) 2 3 5 (This table is continued on the next page.) 68 Table 9 (cont’d) DM Finishing Green $58,000 DL White $40,500 Green $45,000 White $9,000 Manufacturing Overhead: Finishing Indirect Labor $12,750 Supplies $14,750 Machining $49,500 Maintenance $11,750 Utilities $13,250 Other Overhead $10,700 Total $112,700 These budgeted costs are based on an estimated production of 10,000 units of Green and 3,200 units of White, which is the current production volume and is approximately the total capacity usage desired by top management. 69 Direct labor costs are being used to allocate manufacturing overhead costs in Fabrication and Finishing to products. If you recommend keeping this cost driver, then you will not be involved with additional revisions to the costing system, and you will return to your responsibilities as the production manager. If you do not recommend keeping this cost driver, then you will continue to be involved in revising the costing system. Recommend using direct labor costs to allocate manufacturing overhead costs for Fabrication and Finishing, and return to your production responsibilities. Do not recommend using direct labor costs to allocate manufacturing overhead for these activities, and continue to be involved in revising the costing system. 70 One manager, who is new to Michigan, has suggested using machine hours to allocate the manufacturing overhead costs for Fabrication and Finishing, based on her knowledge of the production process. She believes that machine hours would provide for a more accurate allocation of manufacturing overhead costs than direct labor costs. If you choose to recommend machine hours as the basis for allocating manufacturing overhead costs for both of these activities, then you will not be involved with additional revisions to the costing system, and you will return to your responsibilities as the production manager. If you choose not to recommend this cost driver, then you will continue to be involved in revising the costing system. Recommend using machine hours to allocate manufacturing overhead for these activities, and return to your production responsibilities. Do not recommend using machine hours to allocate manufacturing overhead for these activities, and continue to be involved in revising the costing system. 71 To select a cost driver for Fabrication, please refer to the activity information you have been provided in envelope #2 for Fabrication. Please select the ONE indirect cost driver that you believe will best allocate Fabrication overhead costs to products. POSSIBLE COST DRIVERS Number of Molds Fabrication Machine Hours Fabrication Direct Labor Hours Number of Fabrication Setups Fabrication Inspection Hours 72 If you choose to not be involved with additional revisions to the costing system, you will return to your responsibilities as the production manager. If you choose to proceed, then you will continue to be involved in revising the costing system. Complete your involvement in the revision of the costing system, and return to your production responsibilities. Continue your involvement in revising the costing system by recommending a cost driver for the Finishing activity. 73 To select a cost driver for Finishing, please refer to the activity information you have been provided in envelope #2 for Finishing. Please select the ONE indirect cost driver that you believe will best allocate Finishing overhead costs to products. POSSIBLE COST DRIVERS Number of Finishes Finishing Machine Hours Finishing Direct Labor Hours Number of Finishing Setups Finishing Inspection Hours 74 APPENDIX D Post-Experiment Questionnaire 75 1. Why did you decide to continue revising the costing system? Allocate 100 points among the reasons below based on how important each reason was to why you decided to continue revising the costing system, where 0 = Extremely Unimportant and 100 = Extremely Important. Points Reason Revising the costing system was important to me Product cost information was not sufficiently improved to support better decisions My strong cost accounting knowledge I was required to continue revising the costing system I was interested in revising the costing system It was not important for me to return to my production management responsibilities My extensive knowledge of how to revise the costing system I had no other uses of my time I valued what I learned from revising the costing system Other: Total Points 76 2. After choosing a cost driver for Painting, how successful did you believe that this revision to Michigan’s costing system was at improving the costing system? Extremely Unsuccessful 1 2 3 4 5 6 7 8 Extremely Successful 10 9 3. Did you believe that additional revision of the costing system after selecting a cost driver for Painting would lead to costing system improvement? No Improvement 1 2 3 4 5 6 7 Extreme Improvement 9 10 8 4. After you put effort into choosing a cost driver for Painting, did you believe you had made progress towards increasing the accuracy of allocating overhead costs to products? No Progress 1 2 3 4 5 6 7 8 9 Extreme Progress 10 5. How enjoyable was the task of revising the costing system? Extremely Unenjoyable 1 6. 3 4 5 6 7 8 9 After you put effort into choosing a cost driver for Painting, were you committed to increasing the accuracy of allocating overhead costs to products? Not Committed 1 7. 2 Extremely Enjoyable 10 2 3 4 5 6 7 8 9 Extremely Committed 10 How easy or difficult did you find the task of revising the costing system? Extremely Easy 1 2 3 4 5 6 77 7 8 9 Extremely Difficult 10 8. After you put effort into choosing a cost driver for Painting, did you believe you had improved the accuracy of allocating overhead costs to products? No Improvement 2 1 9. 8 7 6 5 4 3 Extreme Improvement 10 9 How important is cost allocation to the success of Michigan’s short-term product decisions, such as decisions concerning current contracts? Extremely Unimportant 1 2 3 4 5 6 7 8 9 Extremely Important 10 10. How important is cost allocation to the success of Michigan’s long-term product decisions, such as decisions concerning strategy? Extremely Unimportant 1 2 3 4 5 6 7 8 9 Extremely Important 10 11. After you put effort into choosing a cost driver for Painting, did you care about increasing the accuracy of allocating overhead costs to products. Did Not Care 1 2 3 4 5 6 7 8 9 Cared Extremely 10 12. How important do you believe revising costing systems is for firms’ success in general? Extremely Unimportant 1 2 3 4 5 6 78 7 8 9 Extremely Important 10 1. What is your age? 2. What is your gender? Male Female 3. Is English your native language? Yes No 4. What is your GPA in the MBA program at Michigan State University? 5. What is your GMAT score? 6. What is your concentration in the MBA program? 7. How many university-level accounting courses have you completed in current or previous degree programs? 8. When you received test data concerning the cost driver you recommended for Painting, did the management accountants recommend using your selected cost driver? Yes No 9. How many months of post-baccalaureate work experience do you have? 10. How many months of work experience do you have doing accounting? 11. How many months of management experience do you have? 79 12. How often have you made (or been part of a group that made) product pricing decisions? Never 1 2 3 4 5 6 7 8 9 Extremely Often 10 13. How often have you made (or been a part of a group that made) decisions about product mix or volume? Never 1 2 3 4 5 6 7 8 9 Extremely Often 10 14. How familiar are you with activity-based costing in a classroom setting? Extremely Unfamiliar 1 2 3 4 5 6 7 8 9 Extremely Familiar 10 15. How familiar are you with activity-based costing in a non-classroom setting (e.g., in a firm)? Extremely Unfamiliar 1 2 3 4 5 6 7 8 9 Extremely Familiar 10 16. How familiar are you with product costing systems in a non-classroom setting (e.g. in a firm)? Extremely Unrealistic 1 2 3 4 5 6 80 7 8 9 Extremely Realistic 10 17. How much effort did you put into completing this exercise? No Effort 1 2 3 4 5 6 7 8 9 Extreme Effort 10 18. How interesting did you find this exercise? Extremely Uninteresting 1 2 3 4 5 6 7 8 Extremely Interesting 9 10 8 9 Extremely Relevant 10 9 Extremely Satisfied 10 9 Extremely Realistic 10 19. How relevant did you find this exercise? Extremely Irrelevant 1 2 3 4 5 6 7 20. How satisfied are you with your participation in this exercise? Extremely Unsatisfied 1 2 3 4 5 6 7 8 21. How realistic did you find this exercise? Extremely Unrealistic 1 2 3 4 5 6 81 7 8 Please verify that your chosen ID is written clearly on your payment sheet. Please keep this payment sheet with you as you will need it to collect your pay. Thank you for your participation. 82 REFERENCES 83 REFERENCES Abernethy, M. A. and J. Bouwens. 2005. Determinants of accounting innovation implementation. Abacus 41 (3): 217-240. Anderson, J. R., L. M. Reder, and C. Lebiere. 1996. Working memory: Activation limitations on retrieval. Cognitive Psychology 30: 221-256. Anderson, S. W. 1995. A framework for assessing cost management system changes: The case of activity based costing implementation at General Motors, 1986-1993. Journal of Management Accounting Research 7: 1-51. Austin, J. T. and J. B. Vancouver. 1996. Goal constructs in psychology: Structure, process, and content. Psychological Bulletin 120 (3): 338-375. Balakrishnan, R., S. C. Hansen, and E. Labro. 2012. Evaluating heuristics used when designing product costing systems. Management Science (Forthcoming). Bargh, J. A. 2006. What have we been priming all these years? On the development, mechanisms, and ecology of nonconscious social behavior. European Journal of Social Psychology 36 (2): 147-168. Bargh, J. A., P. M. Gollwitzer, A. Lee-Chai, K. Barndollar, and R. Trötschel. 2001. The automated will: Nonconscious activation and pursuit of behavioral goals. Journal of Personality and Social Psychology 81 (6): 1014-1027. Bateman, T. S., H. O’Neill, and A. Kenworthy-U’Ren. 2002. A hierarchical taxonomy of top managers’ goals. Journal of Applied Psychology 87 (6): 1134-1148. Bem, D. J. 1972. Self-perception theory. Advances in Experimental Social Psychology 6: 1-62. Bloomfield, R, F. D. Hodge, P. Hopkins, and K. Rennekamp. 2011. Does enhanced disaggregation and cohesive classification of financial information help credit analysts identify firms’ operating structures? Johnson School Research Paper Series No. 14-2011. Booker, D. M. 2000. An experimental investigation of activity-based costing, knowledge, and judgment accuracy. Unpublished dissertation, Michigan State University. Cardinaels, E. 2008. The interplay between cost accounting knowledge and presentation formats in cost-based decision-making. Accounting, Organizations and Society 33: 582-602. Cardinaels, E. and P. M. G. Van Veen-Dirks. 2010. Financial versus non-financial information: The impact of information organization and presentation in a balanced scorecard. Accounting, Organizations and Society 35: 565-578. Carver, C. S. and M. F. Scheier. 1998. On the Self-Regulation of Behavior. New York, NY: Cambridge University Press. 84 Chapman, C. and A. Kern. 2011. Developing accurate costing systems. Working paper, Imperial College London. Clor-Proell, S. and M. W. Nelson. 2007. Accounting standards, implementation guidance, and example-based reasoning. Journal of Accounting Research 45 (4): 699-730. Clor-Proell, S., C. Proell, and T. Warfield. 2013. The Effects of Presentation Salience and Measurement Subjectivity on Nonprofessional Investors’ Fair Value Judgments. Contemporary Accounting Research (Forthcoming). Cooper, R., R. S. Kaplan, L. S. Maisel, E. Morrissey, and R. M. Oehm. 1992. Implementing Activity-Based Cost Management: Moving from Analysis to Action. Montvale, NJ: Institute of Management Accountants. Dearman, D. T. and M. D. Shields. 2001. Cost knowledge and cost-based judgment performance. Journal of Management Accounting Research 13 (1): 1-18. Dhar, R. and I. Simonson. 1999. Making complementary choices in consumption episodes: Highlighting vs. balancing. Journal of Marketing Research 36 (1): 29-44. Dijksterhuis, A. and H. Aarts. 2010. Goals, attention, and (un)conscious. Annual Review of Psychology 61: 467-490. Eldenburg, L., N. Soderstrom, V. Willis, and A. Wu. 2010. Behavioral changes following the collaborative development of an accounting information system. Accounting, Organizations and Society 35 (2): 222-237. Elliott, W. B., J. S. Hobson, and K. E. Jackson. 2011. Disaggregating management forecasts to reduce investors’ susceptibility to earnings fixation. The Accounting Review 86 (1): 185-208. Fishbach, A. and R. Dhar. 2005. Goals as excuses or guides: The liberating effect of perceived goal progress on choice. Journal of Consumer Research 32: 370-377. Fishbach, A., R. Dhar, and Y. Zhang. 2006. Subgoals as substitutes or complements: The role of goal accessibility. Journal of Personality and Social Psychology 91 (2): 232-242. Förster, J. and N. Liberman. 2007. Knowledge activation. In Social Psychology: Handbook of Basic Principles (2nd ed.), edited by A. W. Kruglanski and E. T. Higgins. New York, NY: Guilford. Frederickson, J. R. 1992. Relative performance information: The effects of common uncertainty and contract type on agent effort. The Accounting Review 67 (4): 647-669. Hammersley, J. S., E. M. Bamber, and T. D. Carpenter. 2010. The influence of documentation specificity and priming on auditors’ fraud risk assessments and evidence evaluation decisions. The Accounting Review 85 (2): 547-571. Hankinson, H. and S. P. Lloyd. 2000. Change management. In Handbook of Cost Management, edited by J. Shank. New York, NY: Warren, Gorham, and Lamont. 85 Hannan, R. L., R. Krishnan, and A. H. Newman. 2008. The effects of disseminating relative performance feedback in tournament and individual performance compensation plans. The Accounting Review 83 (4): 893-913. Higgins, E. T. 1996. Knowledge activation: Accessibility, applicability, and salience. In Social Psychology: Handbook of Basic Principles, edited by E. T. Higgins and A. W. Kruglanski. New York, NY: Guilford Press. Horngren, C. T., S. M. Datar, and M. Rajan. 2012. Cost Accounting: A Managerial Emphasis. 14th ed. Boston, MA: Prentice Hall. Hsee, C. K. and R. P. Abelson. 1991. Velocity relation: Satisfaction as a function of the first derivative of outcome over time. Journal of Personality and Social Psychology 60 (3): 341347. Hsu, N. and Z. Schṻtt. 2012. The Psychology of Priming. Hauppauge, NY: Nova Publishers. Hunton, J. E. and D. Gibson. 1999. Soliciting user-input during the development of an accounting information system: Investigating the efficacy of group discussion. Accounting, Organizations and Society 24 (7): 597-618. Ilies, R. and T. A. Judge. 2005. Goal regulation across time: The effects of feedback and affect. Journal of Applied Psychology 90 (3): 453-467. Ittner, C. D. and D. F. Larcker. 2003. Coming up short on nonfinancial performance measurement. Harvard Business Review 81 (11): 88-95. Jiang, J., K. R. Petroni, and I. Y. Wang. 2012. Did stop signs stop investor trading? Investor attention and liquidity in the pink sheets tiers of the OTC market. Working paper, Michigan State University. Kay, J. 1995. Why Firms Succeed. New York, NY: Oxford University Press. Keegan, D. P., R. G. Eiler, and C. R. Jones. 1989. Are your performance measures obsolete? Management Accounting 70 (12): 45-50. Keegan, D. P. and R. G. Eiler. 1994. Let’s reengineer cost accounting. Strategic Finance 76 (2): 26. Kelly, K. O. 2010. Feedback and incentives on nonfinancial value drivers: Effects on managerial decision making. Contemporary Accounting Research 24 (2): 523-556. Kluger, A. N. and A. DeNisi. 1996. The effects of feedback interventions on performance: A historical review, a meta-analysis, and a preliminary feedback intervention theory. Psychological Bulletin 119 (2): 254-284. Kotter, J. P. 2007. Leading change: Why transformation efforts fail. Harvard Business Review 85 (1): 96-103. 86 Krumwiede, K. R. 1998. The implementation stages of activity-based costing and the impact of contextual and organizational factors. Journal of Management Accounting Research 10: 239277. Kwon, T. H. and R. W. Zmud. 1987. Unifying the fragmented models of information systems implementation. In Critical Issues in Information Systems Research, edited by R. J. Boland and R. Hirscheim. New York, NY: John Wiley. Lambert, T. A. and C. P. Agoglia. 2011. Closing the loop: Review process factors affecting audit staff follow-through. Journal of Accounting Research 49 (5): 1275-1306. Latham, G. P. and R. F. Piccolo. 2012. The effect of context-specific versus nonspecific subconscious goals on employee performance. Human Resource Management 51 (4): 511524. Latham, G. P., A. D. Stajkovic, and E. A. Locke. 2010. The relevance and viability of subconscious goals in the workplace. Journal of Management 36: 234-255. Linkenauger, S. 2012. You’ll golf better if you think Tiger has used your clubs. Harvard Business Review 90 (7/8): 32-33. Locke, E. A. and G. P. Latham. 2002. Building a practically useful theory of goal setting and task motivation. American Psychologist 57 (9): 705-717. Luft, J. 1994. Bonus and penalty incentives contract choice by employees. Journal of Accounting and Economics 18 (2): 181-206. Luft, J. and M. D. Shields. 2009. Psychology models of management accounting. Foundations and Trends in Accounting 4 (3-4): 199-345. Maines, L. A. and L. S. McDaniel. 2000. Effects of Comprehensive-Income Characteristics on Nonprofessional Investors’ Judgments: The Role of Financial-Statement Presentation Format. The Accounting Review 75 (2): 179-207. Malina, M. A. and F. H. Selto. 2001. Communicating and controlling strategy: An empirical study of the effectiveness of the balanced scorecard. Journal of Management Accounting Research 13: 47-90. Mitchell, T. R. and D. Daniels. 2003. Motivation. In Handbook of Psychology, v. 12, edited by I. B. Weiner. Hoboken, NJ: John Wiley & Sons. Neely, J. H. 2003. Priming. In Encyclopedia of Cognitive Science, v. 3, edited by Lynn Nadel. London: Nature Publishing Group. Nonaka, I. 1991. The knowledge-creating company. Harvard Business Review 69 (6): 162-171. Norman, H., T. Requate, and I. Waichman. 2012. Do short-term laboratory experiments provide valid descriptions of long-term economic interactions? Working paper, SSRN 2057675. O’Hara, M. T., R. T. Watson, and C. B. Kavan. 1999. Managing the three levels of change. Information Systems Management 16 (3): 63-70. 87 Rowe, C. 2004. The effect of accounting report structure and team structure on performance in cross-functional teams. The Accounting Review 79 (4): 1153-1180. Schacter, D. L. 1994. Priming and multiple memory systems: Perceptual mechanisms of implicit memory. In Memory Systems, edited by D. L. Schacter and E. Tulving. Cambridge, MA: The MIT Press. Shah, J. Y., R. Friedman, and A. W. Kruglanski. 2002. Forgetting all else: On the antecedents and consequences of goal shielding. Journal of Personality and Social Psychology 83 (6): 1261-1280. Shantz, A. and G. P. Latham. 2009. An exploratory field experiment of the effect of subconscious and conscious goals on employee performance. Organizational Behavior and Human Decision Processes 109 (1): 9-17. Shantz, A. and G. P. Latham. 2011. The effect of primed goals on employee performance: Implications for human resource management. Human Resource Management 50 (2): 289299. Shields, M. D. 1995. An empirical analysis of firms’ implementation experiences with activitybased costing. Journal of Management Accounting Research 7: 148-166. Shields, M. D. and M. Young. 1989. A behavioral model for implementing cost management systems. Journal of Cost Management 3: 17-27. Shields, M. D. and M. Young. 2000. Managing behavioral and organizational factors over the cost management system life cycle. In Handbook of Cost Management, edited by J. Shank. New York, NY: Warren, Gorham, and Lamont. Simons, R. 2000. Performance Measurement & Control Systems for Implementing Strategy. Upper Saddle River, NJ: Prentice Hall. Stajkovic, A. D., E. A. Locke, and E. S. Blair. 2006. A first examination of the relationships between primed subconscious goals, assigned conscious goals, and task performance. Journal of Applied Psychology 91 (5): 1172-1180. Tatikonda, L. U. and R. J. Tatikonda. 1998. We need dynamic performance measures. Strategic Finance 80 (3): 49-53. Thornberry, N. 1997. A view about vision. European Management Journal 15 (1): 28-34. Waldrup, B. E., J. B. MacArthur, and J. E. Michelman. 2009. Does your costing system need a tune-up? Strategic Finance 90 (12): 47-51. Wouters, M. 2009. A developmental approach to performance measures - Results from a longitudinal case study. European Management Journal 27: 64-78. Wouters, M. and D. Roijmans. 2011. Using prototypes to induce experimentation and knowledge integration in the development of enabling accounting information. Contemporary Accounting Research 28 (2): 708-736. 88 Wouters, M. and C. Wilderom. 2008. Developing performance-measurement systems as enabling formalization: A longitudinal field study of a logistics department. Accounting, Organizations and Society 33 (4-5): 488-516. Zhang, Y., A. Fishbach, and R. Dhar. 2007. When thinking beats doing: The role of optimistic expectations in goal-based choice. Journal of Consumer Research 34 (4): 567-578. 89