Loading...

Journal Article Analysis

Open Posted By: ahmad8858 Date: 10/10/2020 Graduate Case Study Writing

The article analysis should be written on topic "Balanced Scorecard" based on the below PDF and must include the below 3 sections in APA format.


DEFINITION: a brief definition of the key term followed by the APA reference for the term; this does not count in the word requirement. 

 
SUMMARY: Summarize the article in your own words- this should be in the 150-200 word range. Be sure to note the article's author, note their credentials and why we should put any weight behind his/her opinions, research or findings regarding the key term. 

 
DISCUSSION: Using 300-350 words, write a brief discussion, in your own words of how the article relates to the selected chapter Key Term. A discussion is not rehashing what was already stated in the article, but the opportunity for you to add value by sharing your experiences, thoughts and opinions. This is the most important part of the assignment. 

Category: Mathematics & Physics Subjects: Mathematics Deadline: 12 Hours Budget: $120 - $180 Pages: 2-3 Pages (Short Assignment)

Attachment 1

JOURNAL OF MANAGEMENT ACCOUNTING RESEARCH American Accounting Association Vol. 32, No. 2 DOI: 10.2308/jmar-52574 Summer 2020 pp. 201–224

Causal Inference in Judgment Using the Balanced Scorecard

Kristian Rotaru Monash University

BrainPark

Dennis D. Fehrenbacher Monash University

Min Hui Liang Independent

Axel K.-D. Schulz La Trobe University

ABSTRACT: One of the potential threats to the effectiveness of the Balanced Scorecard (BSC) is that managers over- or underuse particular perspectives of the BSC. Specifically, we investigate the effects of (1) the presentation of strategic objectives (generic strategy map versus strategic objective list), and (2) the performance outcome patterns (positive versus negative outer perspective) across the performance measurement perspectives of the BSC and find support that is consistent with the violation of the causal independence assumption (VCIA) in the psychology literature (Rehder 2014). Our findings show that the presentation of the strategic objectives and the performance outcome patterns interact significantly affecting performance evaluation outcomes. Two follow-up experiments provide further support for the VCIA observed in the main experiment by ruling out an alternative explanation that managers simply place a greater emphasis on financial performance measures.

Data Availability: Data are available from the authors upon request.

Keywords: balanced scorecard; strategy map; causal inference; performance evaluation.

I. INTRODUCTION

K aplan and Norton (1992) developed the Balanced Scorecard (BSC) as a set of performance measures grouped into

four perspectives: Financial Performance, Customer Relations, Internal Business Processes, and Learning and

Growth. The BSC later included a strategy map, a type of causal chain linking the four perspectives, to facilitate the

implementation and evaluation of strategic actions (Kaplan and Norton 1996a, 1996b). Popularity and high level of managerial

satisfaction with the use of the BSC was documented by Rigby and Bilodeau (2011). This was further supported by a more

We gratefully appreciate the comments and suggestions made by Shane S. Dikolli (editor) and two anonymous referees. We also thank Steve Kaplan, Gary Hecht, Alexander Brüggen, Sean Peffer, Ralph Kober, David Smith, and Erin Hawkins, the participants of the 2016 American Accounting Association (AAA) Annual Meeting, the Accounting & Finance Association of Australia and New Zealand (AFAANZ) 2016 Annual Conference, the Monash University Forum on Managerial Accounting (MONFORMA 2016), the 2015 Americas Conference of Information Systems (AMCIS), and the La Trobe Business School research seminar for their valuable comments on earlier drafts of the paper. We acknowledge financial support from the Australian Research Council for funds associated with the equipment used in the Monash Business Behavioural Laboratory.

Kristian Rotaru, Monash University, Monash Business School, Department of Accounting, Victoria, Australia, BrainPark, The Turner Institute for Brain and Mental Health, School of Psychological Sciences and Monash Biomedical Imaging Facility, Monash University, Victoria, Australia; Dennis D. Fehrenbacher, Monash University, Monash Business School, Department of Accounting, Victoria, Australia; Min Hui Liang, Independent, Melbourne, Victoria, Australia; Axel K.-D. Schulz, La Trobe University, La Trobe Business School, Department of Accounting and Data Analytics, Victoria, Australia.

Editor’s note: Accepted by Shane S. Dikolli, under the Senior Editorship of Karen L. Sedatole.

Submitted: August 2016 Accepted: September 2019

Published Online: September 2019

201

recent international BSC Usage Survey (2GC Active Management 2018), where 75 percent of respondents indicated that the

BSC was extremely or very useful to their organization.

A growing body of research in accounting looks at identifying, as well as overcoming, limitations of using the BSC for

judgment in performance evaluation (e.g., Lipe and Salterio 2000, 2002; Banker, Chang, and Pizzini 2004). The literature in

this area has also examined the role of causal relationships between distinct elements of the BSC for decision support (Tayler

2010; Cheng and Humphreys 2012; Humphreys and Trotman 2011).

How individuals draw causal inferences regarding implied cause-and-effect relationships between the perspectives of the

BSC (with or without a formal strategy map) is a topic that has been controversial (Nørreklit 2000; Ittner, Larcker, and Randall

2003). Following the BSC research tradition of investigating the impact of information organization and presentation in the

BSC on performance evaluation (e.g., Cardinaels and van Veen-Dirks 2010; Cheng and Humphreys 2012; Humphreys, Gary,

and Trotman 2016), we explore whether the implicit or explicit representation of causal relationships between BSC

perspectives are subject to cognitive biases associated with mental processing of causal models (Sloman 2005; Rehder 2014;

Hagmayer 2016; Rehder 2018).

The effectiveness of presenting high level fundamental objectives as a strategy map was investigated by Humphreys and

Trotman (2011) who assessed the ability of decision makers to achieve balance between common and unique measures.

Subsequently Cheng and Humphreys (2012) showed that the presentation of strategic objectives as a strategy map improved

information relevance and strategy appropriateness of judgments made by managers. While these studies have shown that

strategy maps can be used to overcome certain cognitive biases, particularly in relation to the common and unique measurement

bias, studies to date have not investigated the potential of strategy maps to enable or exacerbate biased behavior when it comes

to the interpretation and use of causally structured elements in the BSC.

While the term strategy map is commonly used in the literature, the implementation of strategy maps in both practice and

empirical academic studies has been modeled at relatively high levels of aggregation (e.g., Kaplan 2009; Humphreys et al.

2016). Commonly, such a design choice would be adopted to make the strategy maps more visually appealing to managers and

facilitate their buy in, or to facilitate empirical testing by academics. Some studies in the strategic management literature (e.g.,

Scholey 2005; Lueg 2015), have referred to these types of strategy maps as ‘‘generic’’ strategy maps to reflect the high level of aggregation. These types of ‘‘generic’’ strategy maps are the focus of our study.

We draw on literature that has examined specific tendencies exhibited by individuals in processing causal information. In

particular, we draw on a heuristic associated with placing greater emphasis on root causes and ultimate effects of a causal chain

model and thereby paying considerably less attention to the intermediary variables. This tendency has been referred to in prior

literature as the violation of the causal independence assumption (VCIA) or, more broadly, as the violation of the causal

Markov condition (Hagmayer 2016; Rehder 2014; Rehder and Burnett 2005; Walsh and Sloman 2007).

Our study investigates this tendency in a BSC setting by examining the context where performance information for the

‘‘outer perspectives’’ (Learning and Growth and Financial) compared to performance information for the ‘‘inner perspectives’’ (Internal Business Process and Customer) vary (positive outer and negative inner versus negative outer and positive inner). The

explicit representation of cause-and-effect relationships in a generic strategy map using unidirectional links versus using the

presentation of strategic objectives as a separate list without any connectors is a further treatment used in this study.

We conducted three (a main and two follow-up) controlled laboratory experiments where participants assumed the role of a

group manager tasked to undertake a performance evaluation of their department manager. For the main experiment

(Experiment 1), the information provided to the participants involved two different representations of the relationships between

strategic objectives (generic strategy map versus strategic objective list). Two performance outcome patterns (negative outer

perspectives versus positive outer perspectives) relative to target were presented. The manipulation of these patterns allowed us

to test and observe the differential weighting attributed by individuals to the perspectives of the BSC.

Considering prior accounting literature that has examined the role of the financial performance perspective in the BSC

(Cardinaels and van Veen-Dirks 2010), we conducted two follow-up experiments (Experiments 2 and 3) to rule out a potential

alternative explanation that our results in Experiment 1 are driven solely by managers placing greater emphasis on the financial

perspective due to the perceived importance of financial measures rather than on the position of the financial perspective in the

causal chain. The design of the follow-up experiments for the most part repeated the first experiment. A difference in

Experiment 2 was in the design of the performance pattern such that all three of the non-financial performance averages were

positive (negative), whereas the financial performance averages were negative (positive). In Experiment 3, the pattern was such

that the two middle perspectives were held neutral and the Financial perspective (FP) was positive (negative), whereas Learning

and Growth perspective (LGP) performance averages were negative (positive).

The findings of Experiment 1 show that the observed performance evaluation scores are driven by the performance

outcomes in the outer perspectives of the BSC. Specifically, when two identically structured BSCs with equivalent overall

performance outcomes but with different distributions of outcomes across perspectives are provided to decision makers, their

performance evaluation scores are higher when above (versus below) target performance outcomes are presented in the outer

202 Rotaru, Fehrenbacher, Liang, and Schulz

Journal of Management Accounting Research Volume 32, Number 2, 2020

(versus inner) perspectives and when below (versus above) target performance outcomes are presented in the inner (versus

outer) perspectives. This difference in performance evaluation scores indicates that the decision makers attributed a greater

weight to the outer perspectives, where they are perceived as the root cause and the ultimate effect in the BSC causal chain. Our

findings also show that when a generic strategy map (with explicit arrows to indicate the cause-and-effect relationships between

objectives) is presented the difference of performance evaluation scores between the two performance outcome patterns is

greater than when an ordered list of objectives without arrows is presented.1 This indicates a relatively greater weight being

attributed to the outer perspectives when a generic strategy map is used.

Experiments 2 and 3 show that the FP is weighted lower than the three non-FPs combined and not weighted differently

when compared to the LGP, thereby ruling out an alternative explanation that managers simply place a greater emphasis on

financial performance measures.

The remainder of our study is structured as follows: in Section II a brief literature review is presented followed by the

presentation of the hypotheses. In Section III the details of Experiment 1 are outlined. In Section IV the findings of the main

experiment are reported, followed by a brief outline of Experiment 2, Experiment 3, and their results. Section V section

concludes by discussing the implications of the findings within the context of performance evaluation.

II. BACKGROUND AND HYPOTHESES DEVELOPMENT

Cognitive Biases in Performance Information Processing

A growing body of research in accounting has been concerned with better understanding how individuals use the BSC in

performance evaluation (Wong-On-Wing, Guo, Li, and Yang 2007) and the impact of how information is organized and

presented in the BSC (Cardinaels and van Veen-Dirks 2010). Research has investigated whether the use of information

presented in the BSC is subject to various types of information processing distortion that may limit the effectiveness of the BSC

(Tayler 2010; Lipe and Salterio 2000, 2002; Liedtka, Church, and Ray 2008).

Prior literature has also investigated potential solutions to BSC cognitive biases such as the common measures bias (Libby,

Salterio, and Webb 2004; Dilla and Steinbart 2005; Humphreys and Trotman 2011; Cheng and Humphreys 2012). For

example, decision makers affected by common measures bias inappropriately overweigh measures that are common to a

number of individual divisions in the organization compared to measures that are unique to a particular division (Lipe and

Salterio 2000).

Potential solutions to the common measures bias have included invoking process accountability (Libby et al. 2004) and

using strategy maps (Banker et al. 2004; Cheng and Humphreys 2012; Humphreys and Trotman 2011). Researchers in this area

have been primarily concerned with finding remedies to rebalance the weighting individuals give to the measures in the BSC in

response to the common measures bias. The strategy maps used in these previous studies were generic in nature due to the fairly

high level of abstraction used to represent the causal relationships among the perspectives.

With the introduction of the strategy map as a visual representation of the causal relationships inherent in the BSC and its

linkage to organizational strategy, the literature has highlighted potential shortcomings in managerial judgment in regard to

causal relationships. For example, Ittner et al. (2003) provided evidence of companies’ tendency to overlook the validity of the

causal links between leading and lagging BSC performance measures and to ignore the underlying strategically-linked causal

business models.

We investigate the impact of information organization and presentation in the BSC on performance evaluation outcomes.

The potential cognitive limitations in processing graphical causal information presented via strategy maps and their effect on

performance evaluation have not been investigated in prior literature. Our study examines whether individuals exhibit

unintended systematic tendencies, or cognitive biases, in processing causal relationships between the key variables in the BSCs

and generic strategy maps. We use the psychological (as opposed to artificial intelligence) theories of causal Bayes nets2 to

build our predictions regarding causal information processing when using BSC.

BSC Causal Links and the Violation of the Causal Independence Assumption (VCIA)

Inherent in the BSC design is the assumption that individuals take all perspectives into account as ultimately each of these

perspectives plays a role in the implementation of certain aspects of the company’s vision and strategy (Kaplan and Norton

1996a). The four performance perspectives in the BSC—Learning and Growth perspective (LGP), Internal Business Process

1 The order of the list is consistent with that followed in prior studies (e.g., Humphreys and Trotman 2011) and reflects the order the perspectives were presented in Kaplan and Norton (2001).

2 For further review of the differences between the two approaches, see Hagmayer (2016).

Causal Inference in Judgment Using the Balanced Scorecard 203

Journal of Management Accounting Research Volume 32, Number 2, 2020

perspective (IBP), Customer perspective (CP), and Financial perspective (FP)—are considered as causally related (Kaplan and

Norton 1996b; Kaplan 2009). The BSC includes operational measures associated with the LGP and the IBP as leading

indicators, and measures associated with the CP and the FP as lagging indicators or outcomes (Kaplan 2009).

The causal chain formed by the BSC perspectives (see Figure 1) can be conceptualized in a way that perspectives such as

the LGP causally affect the FP, directly as well as indirectly, through the IBP and the CP (Kaplan and Norton 1996a, 2000,

2001, 2006). The question, however, is whether the weights attributed to the BSC perspectives, when using BCS as a decision

support tool, are dictated by their position in the causal mental model thereby structuring the variables according to the implied

hierarchical order.

In line with causal Bayes net theory (Hagmayer 2016), a causal chain presupposes the unconditional dependence

(correlation) of the constituent variables (see Figure 1, Panel B). According to the causal Markov condition underlying

normative causal reasoning, when the state of the mediator (intermediary) variable in a causal chain is known, the state of the

root cause variable should be fully mediated by the intermediary variable (Rottman and Hastie 2014). Strategy map design

(Kaplan and Norton 2000, 2004; Kaplan 2009) follows the logic of causal chains and as such are subject to the causal Markov

condition (Cohen, Thiraios, and Kandilorou 2008).

However, prior literature has demonstrated that people do not necessarily conform to some of the deterministic rules of

causal information processing imposed by causal Bayes net theory. In fact, the literature has found that human decision makers

systematically violate one of the fundamental principles of causal Bayes net theory, which postulates that, when presented with

a causal chain model, the first cause and the ultimate effect are independent, conditional on the intermediate variables (e.g.,

Rehder 2006, 2014; Rehder and Burnett 2005). Therefore, the intermediate variables ‘‘screen off’’ the first cause from the

ultimate effect (Glymour 1998; Hagmayer, Sloman, and Steven 2009; Hagmayer 2016). Individuals’ non-conformity to this

condition is commonly referred to as the violation of the causal independence assumption (VCIA), also known in the literature

as the violation of the Markov condition (Rottman and Hastie 2014). This condition is commonly attributed to the adoption of

associative reasoning by the human decision makers when drawing inferences using the presented causal models. As noted by

FIGURE 1 The BSC Causal Relationships and the Causal Graphical Models

Panel A suggests a causal chain structure between the four BSC perspectives expressed in a graphical form.

Panel B presents a corresponding graphical representation of a causal chain model. The straight arrows connecting the variables indicate the causal relationships between variables.

Panel C presents a case of the violation of the causal independence assumption (VCIA). The attribution of a greater attention toward the endpoints of such a chain is known in the literature as VCIA.

204 Rotaru, Fehrenbacher, Liang, and Schulz

Journal of Management Accounting Research Volume 32, Number 2, 2020

Hagmayer (2016), associative reasoning is based on observed or expected covariations (see Cheng 1997) and therefore entails

the VCIA bias.

Rehder and Burnett (2005) provided experimental evidence for the VCIA when drawing causal inference based on the

presented chain of cause-and-effect relationships. In both predictive and diagnostic inference tasks (i.e., inferring cause from

effect and vice versa) administered in the laboratory setting, participants were more likely to draw direct inference between the root cause and the ultimate effect of the chain, even when the state of the mediating variable was presented. The reported

evidence indicates that the knowledge of the mediating variable did not screen off or reduce the perceived relevance of the

direct relationship between the root cause and the ultimate effect.

In the context of the BSC considered in this study, the validity of the causal independence assumption may also be

violated. For example, individuals making judgments based on the causal graphical models3 (such as causal chain models) have

been found to be influenced by additional variables not explicitly introduced into the causal model (Rehder and Burnett 2005).

The risk of VCIA is particularly high when these individuals perceive the variables in the causal chain model as incomplete or

imperfect measures of the external environment (Hausman and Woodward 2004; Rehder and Burnett, 2005; Walsh and Sloman

2007).

Based on the aforementioned results of their experiments, Rehder and Burnett (2005) inferred that participants in the

experiments drew causal inference based on and beyond the information provided and formed an additional unobserved

‘‘mechanism’’ that had direct causal links with each variable in the chain. The initial hypothesized reason for the unobserved ‘‘mechanism’’ was associated with one’s prior domain knowledge, which resulted in altering the mental causal structure to directly connect the cause and effect. Further, Rottman and Hastie (2014) argue that the exogenous influence of hidden causal

structures with sufficient generative power are being taken into consideration by the decision makers who use causal graphical

models for causal reasoning. Schematically, this mechanism is presented in Figure 1, Panel C where the root cause variable, the

mediating variables, and the effect variable are affected by exogenous influence, or hidden causes, since the presented causal

chain reflects a simplified representation of causal linkages between the variables included in BSC.

Rehder (2014, Experiment 1, Causal chain Condition) observed the VCIA bias when asking participants to make

inferences based on a causal chain structure using a scenario grounded in the economics discipline. This scenario considered

the following economic variables: interest rates (that, according to the scenario, could be low or high), trade deficits (small or

large), and retirement savings (low or high). Participants considered both the value of the initial cause in a chain (interest rates

! trade deficit! retirement savings) and the value of the intermediate variable when predicting the value of the final variable. The results demonstrated that participants’ predictions, informed by their analyses of the suggested causal chains, were subject

to the VCIA because they failed to recognize the causal independence of some variables they faced.

Overall, it is important to note that the VCIA phenomenon has been found to be a persistent factor in human judgment that

is grounded in processing causal information (Rottman and Hastie 2016; Rehder and Waldmann 2017; Rehder 2018). VCIA is

associated with a tendency to attribute a greater weight to the root cause and the ultimate effect than to the mediating variables

when the information being processed by the decision makers is presented as a sequence of causally related factors. Based on

the above, we expect that, due to the simplistic causal nature of the BSC, individuals will likely be influenced by exogenous

factors or hidden causes associated with a given causal chain; thus, the VCIA will hold equally in a BSC context. As such, we

postulate that individuals will be more likely to draw direct inference between the root cause and the ultimate effect of the

causal chain and, hence, will weigh the outer perspectives of the BSC to a greater extent than they will the inner perspectives.

However, whether the bias is relevant to the BSC context is an empirical question.

Stated more formally:

H1: Individuals will attribute greater weight to the outer perspectives than to the inner perspectives of a BSC.

The Interaction Effect from Explicit Links in the Generic Strategy Map

While the BSC has traditionally been based on an underlying causal relationship between the perspectives of the scorecard,

performance measures can be reported in different formats. The BSC performance measurement component is based on

implicit causal relationships between the perspectives, whereas the strategy map explicitly draws causal linkages between

perspectives (Kaplan and Norton 2006). In line with Bayes net theory, the graphical representation of strategy maps, using the

arrows that causally connect strategic objectives, can be considered to be a type of graphical causal model, i.e., the Bayes causal

chain model.4

3 For a comprehensive overview on causal graphical models, see Pearl (1988, 2000) and Sloman (2005). 4 Causal chain models are a type of graphical causal models. Using the terminology from the graph theory, graphical causal models are visualized as

directed (unidirectional) acyclic (without feedbacks) graphs.

Causal Inference in Judgment Using the Balanced Scorecard 205

Journal of Management Accounting Research Volume 32, Number 2, 2020

Kaplan (2009) observed that in practice strategy maps are commonly used in a simplified way, possibly to encourage buy

in by mangers when it comes to the depiction of the causal relationships.5 Such generic strategy maps have been shown to be

effective in reducing some biases such as the common measures bias due to their ability to make the causal relationships more

salient and thus individual cues more relevant (e.g., Humphreys and Trotman 2011).

While generic strategy maps may make individual measures within a particular dimension more relevant, their use in

practice as part of BSC may have the opposite effect when it comes to the relevance of particular dimensions in the chain itself.

We argue that the selective approach of visualizing only some of the cause-and-effect relationship, while leaving others outside

of the modeling scope in generic strategy maps, may lead to biased interpretation of the information communicated via BSC.

In particular, we argue that the strategic objectives and relationships included in generic strategy maps do not necessarily

reflect the complexity of organizational value generating processes. The simplified representation of strategy maps opens the

possibility of individuals inadvertently going beyond the relationships depicted in the strategy map when considering the

simplified links and by doing so inferring other potential generative cause(s) (see Figure 1, Panel C). This phenomenon further

contributes to the degree of the VCIA and leads managers to place even greater emphasis on the beginning and the end of the

causal chain.

Hence, we postulate in our second hypothesis that the generic strategy map may lead to more biased interpretation of the

information communicated via the BSC. Specifically, the simplified depiction of the generic strategy map further increases the

level of the VCIA bias and results in an even greater weighting of the outer perspectives of the BSC by the individual users.

Stated more formally:

H2: The relative difference in weight given to the outer perspectives and that given to the inner perspectives of the BSC

will be greater when a generic strategy map is present than when an ordered list of strategic objectives is present.

III. METHOD

We tested the two hypotheses, H1 and H2, in a laboratory experiment (Experiment 1),6 using a 2 3 2 between-subject factorial design. The factors and the nature of the treatment groups of Experiment 1 are presented in Figure 2. The independent

variables were manipulated at two levels: the performance outcome pattern (positive versus negative performance in the outer

BSC perspective) and the presentation of strategic objectives (in the form of a generic strategy map versus a strategic objective

list). The dependent variable was the individual decision maker’s performance evaluation score. The experimental material was

adapted from Humphreys and Trotman (2011).

Participants

Participants in this experiment were a mix of postgraduate and undergraduate students enrolled at a major university.7 The

appropriateness of using undergraduate and postgraduate students for testing our theory is supported by findings showing that

subject matter experience does not significantly affect individual decision makers’ susceptibility to the VCIA (Rehder 2014).

Thus, for our purpose student judgments can be seen as surrogate for managerial judgments.8 Participants were randomly

allocated to the treatments. The experiment was conducted over 12 sessions in the university’s behavioral laboratory. A total of

114 usable responses were collected. Among the 114 participants, 79 (69.3 percent) were females and 35 (30.7 percent) were

males, with an average age of 21.63 years. The average full-time work experience was 0.54 years.

Experimental Task

In the computerized experimental task, participants were asked to assume the role of a group manager for a fashion apparel

retailer and to conduct an annual performance evaluation of one of the department managers. The information necessary to

perform the task was presented as a BSC. The experiment consisted of four phases.

5 Generic strategy maps are not regarded as best practices or reference models by Kaplan and Norton (2001), who provided examples of much more detailed strategy maps. Kaplan (2009) advocated more detailed mapping in future practice.

6 We also conducted two follow-up experiments (Experiments 2 and 3). They were used to rule out potential alternative interpretations of the findings we obtained in the main experiment. The details of Experiments 2 and 3 are discussed in Section IV. We thank the editor and two anonymous reviewers for suggesting this extension to our study. The suggested extension allowed us to further demonstrate the fit of the suggested theoretical lens in studying causal reasoning with the use of BSC.

7 Approval was granted by the ethics committee of the institution where the experiments took place. 8 Further, as our participants received substantial training related to the use of the BSC as part of the experimental material, we argue that they had

sufficient expertise in the subsequent use of the BSC during performance evaluation.

206 Rotaru, Fehrenbacher, Liang, and Schulz

Journal of Management Accounting Research Volume 32, Number 2, 2020

During the introductory phase (phase one), participants were provided with the background information of the company.

In the second phase of the experiment participants were provided with training material and practice questions. Specifically,

participants were asked to study the company’s managerial structure and their job description as a group manager to reinforce

their role in the experimental task. Then the use of the structure of the BSC was explained to them. After the participants had

carefully studied the training materials, they undertook a series of multiple-choice practice questions, which were designed to

test their awareness of the comparison between actual and target performance, and the achievement toward the strategic

objectives. Participants were not permitted to proceed to the actual business simulation unless they correctly answered the

practice questions. In case of a wrong answer, participants received a message that their answer was incorrect and were given

the training material and then the question again, until the correct answer was obtained. All participants completed the training

section successfully.

In the third phase of the experiment, participants were asked to evaluate the performance of a department manager based

on the detailed description of the department strategy and the BSC provided (see Figure 7 in Appendix A for an example of

treatment Group 1). The BSC integrated the strategic information component and the performance measures component as one

unified report. In line with the outline of experimental treatments presented in Figure 2, participants were provided with one of

the following reports: a BSC that included a generic strategy map and the performance measurement system with the positive

measures (performance above target) associated with the outer perspectives (Group 1), or the negative measures (performance

below target) associated with the outer perspectives (Group 3). Alternatively, the participants were provided with the BSC that

included a strategic objective list and the performance measurement system with the positive measures associated with the outer

perspectives (Group 2), or the negative measures associated with the outer perspectives (Group 4). Other than the names and

the treatments, the departments assigned to each experimental group were equivalent. Participants were instructed to provide a

performance evaluation score on the scale ranging from 1 (Reassign) to 7 (Excellent).

Finally, the post-test experimental questionnaire (phase four) contained demographic questions, including age, gender, and

years of work experience, as well as post-experimental questions.

Independent Variables and Manipulations

The first independent variable in our experiment was the performance outcome pattern. The positive and negative

performance outcomes for the outer perspectives were operationalized as either exceeding performance targets (positive

outcome) or not achieving them (negative outcome). The variable was manipulated at two levels (see Figure 8 in Appendix A):

(1) the actual performance exceeded the targets in FP and LGP (i.e., FP and LGP denote a positive performance), and the actual

performance did not meet the targets in CP or IBP (‘‘positive outer perspectives’’ treatment since FP and LGP are at the outer positions in the original BSC structure of perspectives); or (2) the actual performance measures were below targets in FP and

LGP (i.e., FP and LGP denote a negative performance), and the actual performance measures are above targets in the CP and

IBP (‘‘negative outer perspective’’ treatment). The overall reported performance between these two performance outcome patterns was designed to be equal.

FIGURE 2 2 3 2 Factorial Design and Treatment Groups—Experiment 1

Causal Inference in Judgment Using the Balanced Scorecard 207

Journal of Management Accounting Research Volume 32, Number 2, 2020

In our experiment, all performance measures contained in the BSC were linked to objectives, thereby avoiding potentially

confounding effects associated with participants attributing greater weight to strategically linked measures than to the

strategically non-linked measures (Banker et al. 2004). As shown in Table 4 (see Appendix A), the actual performance was set

either 10 percent above or 10 percent below target performance. In each treatment, two perspectives reflected 10 percent above

target performance and two perspectives reflected 10 percent below target performance. The test of the behavioral outcomes

based on the above two treatments showed no difference in the overall performance between the two performance outcome

patterns (i.e., positive versus negative outer perspective treatments). In addition, the variance of the differences between actual

and target performance were held constant at a very low level of 0.002 across all BSC perspectives in both treatments to avoid

the ambiguity effects from dramatically inconsistent variance between performance outcomes (Liedtka et al. 2008).

The second independent variable in the experiment was the nature of the presentation format, which was manipulated at

two levels (see Figure 9 in Appendix A): (1) generic strategy map presentation (‘‘strategy map’’ treatment), where explicit arrows are outlined to communicate the causal relationship between strategic objectives across four BSC perspectives, and (2)

an ordered list of strategic objectives (‘‘strategic objective list’’ treatment), where strategic objectives are grouped into BSC perspectives in a list format without the explicit arrows between them (see Figure 10 in Appendix A).

Dependent Variable

Participants were instructed to conduct a performance evaluation decision after reviewing the BSC. The dependent variable

was the performance evaluation score assigned by participants. Since the average of the performance measures across the four

perspectives was equal across conditions, differences between evaluations were attributed to differences in weighting the

perspectives. Participants made the performance evaluation decision (see Figure 7 in Appendix A) based on the seven-point

Likert scale ranging from 1 (Reassign) to 7 (Excellent).

Manipulation Check

The first manipulation check was used to test participants’ understanding of the presentation of strategic objectives and the

second one to test the understanding of the manipulation in the performance outcome patterns. The total number of failures in

two manipulation checks was 41. Of the total number, 34 failed the first manipulation check and 16 failed the second one. Since

participants underwent training in the use of BSCs before proceeding to the experiment and were not allowed to proceed if they

answered the understanding check questions (see above) incorrectly, we kept all 114 participants who completed the

experiment for our subsequent analysis.9 Further, we checked whether the random assignment to the manipulations was

successful. Age, gender, and work experience were not significantly associated with the treatment variables.

IV. RESULTS

Experiment One: Main Experiment

Descriptive Statistics

Descriptive statistics, reported in Table 1, Panel A and Figure 3, show a higher evaluation score for treatment groups with

the positive outer perspectives (3.63) as compared to the treatment groups with the negative outer perspectives (2.95). The

presence of a generic strategy map resulted in a similar evaluation score to strategic objects being listed (3.37 versus 3.21). The

difference in the evaluation score between the positive and negative outer perspectives treatments was greater when a generic

strategy map was administered (3.83 – 2.89¼ 0.94) than when the perspectives were represented as a strategic objective list (3.43 – 3.00 ¼ 0.43).

Hypothesis Testing

H1 predicts that individuals will place a greater weight on the outer perspectives (FP and LGP) than on the inner

perspectives (CP and IBP). Since the averages of the performance measures between the positive outer perspective condition

and negative outer perspective condition are the same, differences in performance evaluation are attributed to differences in the

weighting of the outer versus the inner perspectives. Our results, reported in Table 1, Panel B, show that the groups with

positive outer perspectives arrive at a significantly higher performance evaluation score (3.63) than those with negative outer

perspectives (2.95, F ¼ 24.40, p , 0.01). We thus find support for H1.

9 Removing those participants who failed the manipulation checks and conducting the analysis reported in Table 2 on the reduced sample lead to similar findings and did not change our results.

208 Rotaru, Fehrenbacher, Liang, and Schulz

Journal of Management Accounting Research Volume 32, Number 2, 2020

H2 predicts that the relative difference in weight given to the outer perspective compared to that given to the inner

perspective of the BSC is greater when a generic strategy map is present than when an ordered list of strategic objectives is

present. Our results, reported in Table 1, Panel A, show that the difference in evaluation score between the positive and

negative outer perspectives cases was greater when a generic strategy map was used (0.94) than when a strategic objective list

was used (0.43). The ANOVA reported in Table 1, Panel B shows a significant interaction effect between the presentation of

strategic objectives and performance outcome pattern (F ¼ 3.36, p , 0.05). Thus, we find support for H2. Testing the mean differences (not tabulated) between Group 1 and Group 2 and between Group 3 and Group 4 show that

there is a significant difference between the positive outer perspective conditions (t ¼ 2.08, p , 0.05), but not between the negative outer perspective conditions (t ¼ 0.54, p ¼ 0.59), indicating that the interaction is driven by the positive outer perspective conditions. While our theory is not specific enough to predict the differential role of the positive performance in the

outer perspectives of BSC, it may be an interesting aspect to explore in future research.

Further Analyses

A potential alternative explanation for our Experiment 1 findings is that managers deliberately weight the financial

perspective to a greater extent than the other three perspectives. As part of the experimental design, we asked the participants to

what extent the performance in each perspective influenced their evaluation in the post-experimental questionnaire. Participants

rated each of the four perspectives on a scale from 1 (very low influence) to 7 (very high influence). The means of the FP¼5.45 (SD¼ 1.30), CP¼ 5.68 (SD¼ 1.16), IBP¼ 5.15 (SD¼ 1.01), and LGP¼ 5.18 (SD¼ 1.17) did not vary significantly. Thus, individuals did not perceive the financial perspective as more important than the other perspectives for their performance

evaluation; rather, they indicated that they weighted the perspectives equally on average. As shown in the main analysis above,

however, they applied some weights leading to the VCIA, suggesting that the bias is unconscious.

TABLE 1

Results of Experiment 1

Panel A: Means (Standard Deviations) for Performance Evaluation Score—Experiment 1

Performance Outcome Patterns

Presentation of Strategic Objectives

Row Average

Generic Strategy Map

Strategic Objective List

Group 1 Group 2

Positive Outer Perspectives 3.83 3.43 3.63

(0.80) (0.63) (0.75)

n ¼ 29 n ¼ 28 n ¼ 57 Group 3 Group 4

Negative Outer Perspectives 2.89 3.00 2.95

(0.79) (0.71) (0.74)

n ¼ 28 n ¼ 29 n ¼ 57 Column Average 3.37 3.21 3.29

(0.92) (0.70) (0.82)

n ¼ 57 n ¼ 57 n ¼ 114

Panel B: Two-Way ANOVA Model for Mean Performance Evaluation Score—Experiment 1

SS df MS F p-value

(two-tailed)

Performance Outcome Pattern (Positive Outer Perspectives versus Negative

Outer Perspectives)

13.24 1 13.24 24.40 0.00

Presentation of Strategic Objectives (Generic Strategy Map versus Strategic

Objective List)

0.61 1 0.61 1.12 0.15

Presentation of Strategic Objectives 3 Performance Outcome Pattern 1.83 1 1.83 3.36 0.04 Error 59.67 110 0.54

Total 1309.00 114

Causal Inference in Judgment Using the Balanced Scorecard 209

Journal of Management Accounting Research Volume 32, Number 2, 2020

We also tested whether participants in the generic strategy map treatments, as opposed to those in the strategic objective

list treatments, perceived particular perspectives as more influential. Four separate ANOVAs with the influence of the

perspective (either FP, CP, IBP or LGP) as the dependent variable and the presentation of strategic objectives (strategy map

versus strategic objective list) as the independent variable did not provide evidence that the presentation significantly influences

how participants rate the perceived importance of the perspectives.

Follow-Up Experiments

The above reported analyses reveal that performance evaluation outcomes were not driven by participants’ conscious

preference for any of the BSC perspective(s) over the remaining ones. Yet, the BSC literature (Kaplan and Norton 1992;

Kaplan 2009) postulates the role of FP as the ultimate output in the value creation process reflected through the BSC. As such,

the performance level reported through FP may have more influence on performance evaluation outcomes than other BSC

perspectives have (Cardinaels and van Veen-Dirks 2010). To rule out such a bias we conducted two follow-up experiments.

We conducted the follow-up experiments using two separate participant cohorts to isolate the effect of the FP in contrast to

all three of the non-FPs (Experiment 2) as well as the effect of the FP in contrast to the LGP, while keeping the IBP and the CP

neutral (Experiment 3). The 2 3 2 factorial design of the follow-up experiments and their procedures were identical to that in the main experiment, with the exception of changing the manipulation of the performance patterns. Figure 4 provides an

overview of the performance patterns used across all three experiments.

In Experiment 2, we used a different outcome pattern in order to test whether differences in weightings are attributed

specifically to the weighting of the FP. As shown in Figure 4, the pattern described was either a positive outcome for the FP and

negative outcomes for the other three perspectives (Experiment 2—Pattern 1) or a negative outcome for the FP and positive

outcomes for the other three perspectives (Experiment 2—Pattern 2). In Experiment 3, we used a further variation of the

outcome pattern to directly test for differences in weightings between the two outer perspectives by keeping the inner

perspective outcomes neutral at ‘‘equal to target.’’ As shown in Figure 4, we administered an outcome pattern where the outcome was positive for the FP and negative for the LGP (Experiment 3—Pattern 1) or where the outcome was negative for

the FP and positive for the LGP (Experiment 3—Pattern 2). The BSC outcome patterns tested in Experiments 2 and 3 (see

Figure 4) were specifically designed to rule out a potential alternative explanation that a relatively greater weighting of the FP

alone, rather than both outer perspectives (FP and LGP), explain the results for our Experiment 1.

Experiment 2

In Experiment 2, we designed conditions in which all three of the non-FP were positive (negative) and the FP was negative

(positive), as shown in Figure 5 and Figure 10 (Appendix A). Following the design of Experiment 1 the mean performance

difference of the measures in the FP was 10 percent above target (10 percent below target) in the positive (negative) FP

FIGURE 3 Mean Performance Evaluation Score—Experiment 1

210 Rotaru, Fehrenbacher, Liang, and Schulz

Journal of Management Accounting Research Volume 32, Number 2, 2020

condition. In contrast, the mean performance difference of the measures in the three non-FP was 10 percent below target (10

percent above target). The manipulation of the strategy map component was identical to that administered in Experiment 1.

A further cohort of 116 participants with similar demographics and backgrounds took part in the second experiment: 71

(61.2 percent) were females and 45 (38.8 percent) were males. The mean age of the participants was 20.6 years, and the average

full-time work experience was 0.64 years. Consistent with the choice of analysis adopted in Experiment 1 and motivated by the

similarity of reported outcomes in the reduced and the full samples, we used the full sample for our analysis. See Table 2, Panel

A for an overview of descriptive statistics.

A two-factor ANOVA was used to investigate the effects of the role of financial performance in the BSC (positive FP

versus negative FP) crossed with the presentation of objectives (strategic objective list versus strategy map). In the positive FP

condition, the performance evaluation score was 3.64, whereas in the negative FP condition, the performance evaluation score

was 5.12 (see Table 2, Panel A). As reported in Table 2, Panel B, the ANOVA showed a statistically significant main effect of

the performance patterns (F ¼ 76.56, p , 0.001). The observed significant main effect indicates that the three non-FP perspectives influenced the performance evaluation score to a greater extent than the FP perspective; hence, our results in the

main experiment were not solely driven by participants’ focus on the FP perspective. While this finding provides additional

support for our theory, the fact that the manager in the positive FP condition receives an average evaluation of 3.64 (lower than

that of 5.12 in the negative FP condition) may also indicate that participants interpret the pattern in a way that the strategy may

have worked in the past (as indicated by the positive financial outcome), but may not continue to work in the future given the

leading indicators are negative (see also Campbell et al. 2015).10 We conducted Experiment 3 to address this alternative

explanation, holding the mediating perspectives of IBP and CP constant at the target level.

TABLE 2

Results of Experiment 2

Panel A: Means (Standard Deviations) for Performance Evaluation Score—Experiment 2

Performance Outcome Patterns

Presentation of Strategic Objectives

Row Average

Generic Strategy Map

Strategic Objective List

Group 1 Group 2

Positive FP 3.87 3.39 3.64

(0.957) (1.100) (1.047)

n ¼ 31 n ¼ 28 n ¼ 59 Group 3 Group 4

Negative FP 5.00 5.27 5.12

(0.894) (0.667) (0.803)

n ¼ 31 n ¼ 26 n ¼ 57 Column Average 4.44 4.30 4.38

(1.081) (1.312) (1.191)

n ¼ 62 n ¼ 54 n ¼ 116

Panel B: Two-Way ANOVA Model for Mean Performance Evaluation Score—Experiment 2

SS df MS F p-value

(two-tailed)

Performance Outcome Pattern (Positive FP versus Negative FP) 65.13 1 65.13 76.56 0.00

Presentation of Strategic Objectives (Generic Strategy Map versus Strategic Objective List) 0.32 1 0.32 0.37 0.54

Presentation of Strategic Objectives 3 Performance Outcome Pattern 4.03 1 4.03 4.73 0.03 Error 95.28 112 0.85

Total 2379.00 116

10 We thank the anonymous reviewer for this suggestion.

Causal Inference in Judgment Using the Balanced Scorecard 211

Journal of Management Accounting Research Volume 32, Number 2, 2020

Experiment 3

In Experiment 3, we designed conditions in which the FP (positive/negative) was contrasted with the LGP (negative/

positive), while the IBP and CP remained at the target level across both treatment conditions. As per Figure 11 (Appendix A),

the mean performance difference of the measures in the FP was 10 percent above target (10 percent below target), while the

mean performance score on the LGP was 10 percent below target (10 percent above target). The performance of the IBP and CP

measures corresponded to the target levels across both conditions. The manipulation of the strategy map component was

identical to that administered in Experiments 1 and 2 (see Figure 6).

A further cohort of 123 participants with similar demographics and backgrounds took part in the third experiment: 75

(60.98 percent) were females and 48 (39.02 percent) were males. The mean age of the participants was 20.7 years, and the

average full-time work experience was 0.69 years. Consistent with the choice of analysis adopted in Experiment 1 and 2, we

used the full sample for our analysis. See Table 3, Panel A for an overview of descriptive statistics.

A two-factor ANOVA was used to investigate the effects of the role of financial performance in the BSC (positive FP/

negative LGP versus negative FP/positive LGP), crossed with the presentation of objectives (strategic objective list versus

strategy map). In the positive FP/negative LGP condition the performance evaluation score was 3.79, whereas in the negative

FP/positive LGP condition, the performance evaluation score was 3.72 (see Table 3, Panel A). As reported in Table 3, Panel B,

the ANOVA showed a statistically non-significant main effect of the performance patterns (F ¼ 0.23, p ¼ 0.64). This result indicates that there is no differential effect of the beginning (LPG) or the end (FP) of the causal chain, thus providing further

support for our theoretical prediction that individuals place greater emphasis on both outer perspectives rather than solely on the

FP perspective.

FIGURE 4 Experimental Design Summary for Experiments 1 to 3

FIGURE 5 2 3 2 Factorial Design and Treatment Groups—Experiment 2

212 Rotaru, Fehrenbacher, Liang, and Schulz

Journal of Management Accounting Research Volume 32, Number 2, 2020

V. DISCUSSION AND CONCLUSION

In our study, we examine how a causal inference judgment bias affects performance evaluation judgment based on a

performance measurement system structured as a Balanced Scorecard (BSC). In particular, we find support for the violation of

the causal independence assumption (Rehder 2014; Rehder and Burnett 2005; Rottman and Hastie 2014; Walsh and Sloman

TABLE 3

Results of Experiment 3

Panel A: Means (Standard Deviations) for Performance Evaluation Score—Experiment 3

Performance Outcome Patterns

Presentation of Strategic Objectives

Row Average

Generic Strategy Map

Strategic Objective List

Group 1 Group 2

Positive FP/Neutral IBP, CP/Negative LGP 3.77 3.81 3.79

(1.055) (0.998) (1.027)

n ¼ 31 n ¼ 32 n ¼ 63 Group 3 Group 4

Negative FP/Neutral IBP, CP/Positive LGP 3.83 3.60 3.72

(0.699) (0.770) (0.803)

n ¼ 30 n ¼ 30 n ¼ 60 Column Average 3.80 3.71 3.76

(0.877) (0.884) (0.898)

n ¼ 61 n ¼ 62 n ¼ 123

Panel B: Two-Way ANOVA Model for Mean Performance Evaluation Score—Experiment 3

SS df MS F p-value

(two-tailed)

Performance Outcome Pattern (Positive FP/Negative LGP versus Negative FP/Positive LGP) 0.18 1 0.18 0.23 0.64

Presentation of Strategic Objectives (Generic Strategy Map versus Strategic Objective List) 0.29 1 0.29 0.36 0.55

Presentation of Strategic Objectives 3 Performance Outcome Pattern 0.57 1 0.57 0.71 0.40 Error 95.66 119 0.80

Total 1832.00 123

FIGURE 6 2 3 2 Factorial Design and Treatment Groups—Experiment 3

Causal Inference in Judgment Using the Balanced Scorecard 213

Journal of Management Accounting Research Volume 32, Number 2, 2020

2007), which manifests itself in greater weighting attributed to measures associated with the outer perspectives of the BSC than

to those associated with the inner perspectives.

We also show that the use of a generic strategy map to visually represent and emphasize causal relationships between the

measures further exacerbates this bias, resulting in increased weighting attributed to the outer perspectives of the BSC. This

observed interaction between the strategy map representation of strategic objectives and the performance pattern contained in

the BSC is of particular importance as the use of generic strategy maps has been found to be an effective remedy for other

cognitive biases associated with the BSC, such as the common measures bias (Libby et al. 2004; Dilla and Steinbart 2005).

Results from our follow-up experiments allow us to rule out a possible alternative explanation based on the importance

individuals may place (implicitly or explicitly) on the financial perspective versus other BSC perspectives.

The findings reported in our study offer a number of practical implications. A balanced consideration of all BSC

perspectives is a non-biased way to evaluate the organizational strategic performance (Banker et al. 2004; Lipe and Salterio

2000). Our findings provide empirical evidence that performance measures in the outer perspectives may receive a greater

weighting by decision makers, particularly when a generic strategy map is used. Such differential weighting can lead to

potentially inequitable performance evaluation. Further, results of our follow-up experiments support the notion that the FP

perspective is not automatically weighted more than the other perspectives by users of the BSC.

Our study has a number of limitations. As it is based on a laboratory experiment, our task uses a simplified BSC, hence, the

underlying causal relationships are highly stylized and simplified. Future research might investigate whether substituting the

generic strategy map with a more detailed strategy map as part of the experimental treatment could lead to a reduction of the

VCIA reported in our study. Future research could also consider examining a more complex context, including the use of

additional categories or more diverse performance patterns. Further, our experiment was limited to a single period context;

hence, we were not in a position to examine potential longitudinal learning effect that may be associated with BSC use.

REFERENCES

2GC Active Management. 2018. Balanced scorecard usage survey: Summary of findings. Available at: https://2gc.eu/media/resource_ files_survey_reports/2018_Survey_Document_10_Final-compressed.pdf

Banker, R. D., H. Chang, and M. Pizzini. 2004. The balanced scorecard: Judgmental effects of performance measures linked to strategy.

The Accounting Review 79 (1): 1–23. https://doi.org/10.2308/accr.2004.79.1.1

Campbell, D., S. M. Datar, S. L. Kulp, and V. G. Narayanan. 2015. Testing strategy with multiple performance measures: Evidence from

a balanced scorecard at Store24. Journal of Management Accounting Research 27 (2): 39–65. https://doi.org/10.2308/jmar-51209

Cardinaels, E., and P. M. 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 (6): 565–578. https://doi.org/10.1016/j.aos.2010. 05.003

Cheng, P. W. 1997. From covariation to causation: A causal power theory. Psychological Review 104 (2): 367–405. https://doi.org/10. 1037/0033-295X.104.2.367

Cheng, M. M., and K. A. Humphreys. 2012. The differential improvement effects of the strategy map and scorecard perspectives on

managers’ strategic judgments. The Accounting Review 87 (3): 899–924. https://doi.org/10.2308/accr-10212

Cohen, S., D. Thiraios, and M. Kandilorou. 2008. Performance parameters interrelations from a balanced scorecard perspective: An

analysis of Greek companies. Managerial Auditing Journal 23 (5): 485–503. https://doi.org/10.1108/02686900810875307

Dilla, W. N., and P. J. Steinbart. 2005. Relative weighting of common and unique balanced scorecard measures by knowledgeable

decision makers. Behavioral Research in Accounting 17 (1): 43–53. https://doi.org/10.2308/bria.2005.17.1.43

Glymour, C. 1998. Learning causes: Psychological explanations of causal explanation. Minds and Machines 8 (1): 39–60. https://doi.org/ 10.1023/A:1008234330618

Hagmayer, Y. 2016. Causal Bayes nets as psychological theories of causal reasoning: Evidence from psychological research. Synthese 193 (4): 1107–1126. https://doi.org/10.1007/s11229-015-0734-0

Hagmayer, Y., S. A. Sloman, and A. Steven. 2009. Decision makers conceive of their choices as interventions. Journal of Experimental Psychology. General 138 (1): 22–38. https://doi.org/10.1037/a0014585

Hausman, D. M., and J. Woodward. 2004. Modularity and the causal Markov condition: A restatement. The British Journal for the Philosophy of Science 55 (1): 147–161. https://doi.org/10.1093/bjps/55.1.147

Humphreys, K. A., and K. T. Trotman. 2011. The balanced scorecard: The effect of strategy information on performance evaluation

judgments. Journal of Management Accounting Research 23 (1): 81–98. https://doi.org/10.2308/jmar-10085

Humphreys, K. A., M. S. Gary, and K. T. Trotman. 2016. Dynamic decision making using the balanced scorecard framework. The Accounting Review 91 (5): 1441–1465. https://doi.org/10.2308/accr-51364

Ittner, C., D. Larcker, and T. Randall. 2003. Performance implications of strategic performance measurement in financial services firms.

Accounting, Organizations and Society 28 (7-8): 715–741. https://doi.org/10.1016/S0361-3682(03)00033-3

214 Rotaru, Fehrenbacher, Liang, and Schulz

Journal of Management Accounting Research Volume 32, Number 2, 2020

Kaplan, R. S. 2009. Conceptual foundations of the balanced scorecard. Handbooks of Management Accounting Research 3: 1253–1269. https://doi.org/10.1016/S1751-3243(07)03003-9

Kaplan, R. S., and D. P. Norton. 1992. The balanced scorecard: Measures that drive performance. Harvard Business Review 70 (1): 71– 79.

Kaplan, R. S., and D. P. Norton. 1996a. The Balanced Scorecard: Translating Strategy into Action. Boston, MA: Harvard Business School Press.

Kaplan, R. S., and D. P. Norton. 1996b. Using the balanced scorecard as a strategic management system. Harvard Business Review 74 (1): 75–85.

Kaplan, R. S., and D. P. Norton. 2000. Having trouble with your strategy? Then map it. Harvard Business Review 78 (5): 167–176. Kaplan, R. S., and D. P. Norton. 2001. The Strategy-Focused Organization: How Balanced Scorecard Companies Thrive in the New

Business Environment. Boston, MA: Harvard Business School Press. Kaplan, R. S., and D. P. Norton. 2004. Strategy Maps: Converting Intangible Assets into Tangible Outcomes. Boston, MA: Harvard

Business School Press.

Kaplan, R. S., and D. P. Norton. 2006. Alignment: Using the Balanced Scorecard to Create Corporate Synergies. Boston, MA: Harvard Business School Press.

Libby, T., S. E. Salterio, and A. Webb. 2004. The balanced scorecard: The effects of assurance and process accountability on managerial

judgment. The Accounting Review 79 (4): 1075–1094. https://doi.org/10.2308/accr.2004.79.4.1075 Liedtka, S. L., B. K. Church, and M. R. Ray. 2008. Performance variability, ambiguity intolerance, and balanced scorecard-based

performance assessments. Behavioral Research in Accounting 20 (2): 73–88. https://doi.org/10.2308/bria.2008.20.2.73 Lipe, M. G., and S. E. Salterio. 2000. The balanced scorecard: Judgmental effects of common and unique performance measures. The

Accounting Review 75 (3): 283–298. https://doi.org/10.2308/accr.2000.75.3.283 Lipe, M. G., and S. E. Salterio. 2002. A note on the judgmental effects of the balanced scorecard’s information organization. Accounting,

Organizations and Society 27 (6): 531–540. https://doi.org/10.1016/S0361-3682(01)00059-9 Lueg, R. 2015. Strategy maps: The essential link between the balanced scorecard and action. The Journal of Business Strategy 36 (2): 34–

40. https://doi.org/10.1108/JBS-10-2013-0101

Nørreklit, H. 2000. The balance on the balanced scorecard a critical analysis of some of its assumptions. Management Accounting Research 11 (1): 65–88. https://doi.org/10.1006/mare.1999.0121

Pearl, J. 1988. Probabilistic Reasoning in Intelligent Systems: Networks of Plausible Inference. San Mateo, CA: Morgan Kaufman. Pearl, J. 2000. Causality. Cambridge, UK: Cambridge University Press. Rehder, B. 2006. When causality and similarity compete in category based property induction. Memory & Cognition 34 (1): 3–16. https://

doi.org/10.3758/BF03193382

Rehder, B. 2014. Independence and dependence in human causal reasoning. Cognitive Psychology 72: 54–107. https://doi.org/10.1016/j. cogpsych.2014.02.002

Rehder, B. 2018. Beyond Markov: Accounting for independence violations in causal reasoning. Cognitive Psychology 103: 42–84. https://doi.org/10.1016/j.cogpsych.2018.01.003

Rehder, B., and R. C. Burnett. 2005. Feature inference and the causal structure of categories. Cognitive Psychology 50 (3): 264–314. https://doi.org/10.1016/j.cogpsych.2004.09.002

Rehder, B., and M. R. Waldmann. 2017. Failures of explaining away and screening off in described versus experienced causal learning

scenarios. Memory & Cognition 45 (2): 245–260. https://doi.org/10.3758/s13421-016-0662-3 Rigby, D., and B. Bilodeau. 2011. Management Tools and Trends 2011. Boston, MA: Bain & Company. Rottman, B. M., and R. Hastie. 2014. Reasoning about causal relationships: Inferences on causal networks. Psychological Bulletin 140

(1): 109–139. https://doi.org/10.1037/a0031903

Rottman, B. M., and R. Hastie. 2016. Do people reason rationally about causally related events? Markov violations, weak inferences, and

failures of explaining away. Cognitive Psychology 87: 88–134. https://doi.org/10.1016/j.cogpsych.2016.05.002 Scholey, C. 2005. Strategy maps: A step-by-step guide to measuring, managing and communicating the plan. The Journal of Business

Strategy 26 (3): 12–19. https://doi.org/10.1108/02756660510597065 Sloman, S. A. 2005. Causal Models: How We Think About The World And Its Alternatives. Cambridge, MA: Oxford University Press. Tayler, W. B. 2010. The balanced scorecard as a strategy-evaluation tool: The effects of implementation involvement and a causal-chain

focus. The Accounting Review 85 (3): 1095–1117. https://doi.org/10.2308/accr.2010.85.3.1095 Walsh, C., and S. Sloman. 2007. Updating beliefs with causal models: Violations of screening off. In Memory and Mind: A Festschrift for

Gordon H. Bower, edited by M. A. Gluck, J. R. Anderson, and S. M. Kosslyn, 345–357. Wong-On-Wing, B., L. Guo, W. Li, and D. Yang. 2007. Reducing conflict in balanced scorecard evaluations. Accounting, Organizations

and Society 32 (4): 363–377. https://doi.org/10.1016/j.aos.2006.05.001

APPENDIX A

Appendix A presents Table 4 and Figures 7, 8, 9, 10, and 11.

Causal Inference in Judgment Using the Balanced Scorecard 215

Journal of Management Accounting Research Volume 32, Number 2, 2020

TABLE 4

Performance Difference between Actual and Target Performance—Experiment 1

Perspectives

Positive Outer Perspective Treatment

Negative Outer Perspective Treatment

Mean Difference

Variance of Difference

Mean Difference

Variance of Difference

Financial þ10% 0.002 �10% 0.002 Customer �10% 0.002 þ10% 0.002 Internal Business Process �10% 0.002 þ10% 0.002 Learning and Growth þ10% 0.002 �10% 0.002 Average 0% 0.002 0% 0.002

A positive sign of the mean difference indicates the actual performance exceeded the target performance. A negative sign of the mean difference indicates the actual performance did not reach the target performance.

216 Rotaru, Fehrenbacher, Liang, and Schulz

Journal of Management Accounting Research Volume 32, Number 2, 2020

FIGURE 7 BSC and Evaluation Scale for Treatment Group 1—Experiment 1

Panel A: Strategic Performance Report

Panel B: Evaluation (based on the above report, please provide your evaluation of the performance of the Department Manager on the scale below ranging from 1 [Reassign] to 7 [Excellent])

This is the example of the Strategic Performance Report and the seven-point scale for the performance evaluation provided in Treatment Group 1.

Causal Inference in Judgment Using the Balanced Scorecard 217

Journal of Management Accounting Research Volume 32, Number 2, 2020

FIGURE 8 Manipulations of Performance Outcome Pattern—Experiment 1

Panel A: Positive Outer Perspectives Treatment

(continued on next page)

218 Rotaru, Fehrenbacher, Liang, and Schulz

Journal of Management Accounting Research Volume 32, Number 2, 2020

FIGURE 8 (continued)

Panel B: Negative Outer Perspectives Treatment

Performance measures are equivalent in the treatments including the strategy map.

Causal Inference in Judgment Using the Balanced Scorecard 219

Journal of Management Accounting Research Volume 32, Number 2, 2020

FIGURE 9 Manipulations of Presentation of Strategic Objectives

Both of the treatments present the same strategy, but in a different presentation format.

220 Rotaru, Fehrenbacher, Liang, and Schulz

Journal of Management Accounting Research Volume 32, Number 2, 2020

FIGURE 10 Manipulations of FP Performance Outcome—Experiment 2

Panel A: Positive FP/Negative LGP, IBP, CP Treatment

(continued on next page)

Causal Inference in Judgment Using the Balanced Scorecard 221

Journal of Management Accounting Research Volume 32, Number 2, 2020

FIGURE 10 (continued)

Panel B: Negative FP/Positive LGP, IBP, CP Treatment

222 Rotaru, Fehrenbacher, Liang, and Schulz

Journal of Management Accounting Research Volume 32, Number 2, 2020

FIGURE 11 Manipulations of FP Performance Outcome—Experiment 3

Panel A: Positive FP/Neutral IBP, CP/Negative LGP Treatment

(continued on next page)

Causal Inference in Judgment Using the Balanced Scorecard 223

Journal of Management Accounting Research Volume 32, Number 2, 2020

FIGURE 11 (continued)

Panel B: Negative FP/Neutral IBP, CP/Positive LGP Treatment

224 Rotaru, Fehrenbacher, Liang, and Schulz

Journal of Management Accounting Research Volume 32, Number 2, 2020

Copyright of Journal of Management Accounting Research is the property of American Accounting Association and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.