DASHBOARDS IN PERFORMANCE MANAGEMENT: THE FOUNDATIONS AND RESEARCH OPPORTUNITIES FOR THEIR EFFECTIVE USE

~ 1004 ~

DASHBOARDS IN PERFORMANCE MANAGEMENT:

THE FOUNDATIONS AND RESEARCH

OPPORTUNITIES FOR THEIR EFFECTIVE USE

Ogan YIGITBASIOGLU

and

Oana VELCU

ABSTRACT

When designed effectively dashboards are expected to reduce information overload and improve performance management. Hence, interest in dashboards has increased recently, which is also evident from the proliferation of dashboard solution providers in the market. Despite dashboards popularity, little is known about the extent of their effectiveness in organizations. Dashboards draw from multiple disciplines but ultimately use visualization to communicate important information to stakeholders. Thus, a better understanding of visualization can improve the design and use of dashboards. This paper reviews the foundations and roles of dashboards in performance management and proposes a framework for future research, which can enhance dashboard design and perceived usefulness depending on the fit between the features of the dashboard and the characteristics of the users.

KEY WORDS

Dashboards, visualization, presentation formats, performance measurement

INTRODUCTION

A recent development in performance management is the use of dashboards, which are perceived as one of the most useful analysis tools in business intelligence (BI) (Negash and Gray, 2008). Dashboards became popular after the Enron scandal in 2001, which drew the attention upon the need to have managers at all levels to be able to monitor and control the operations of companies (Few, 2006).

The BI vendor market today offers an array of dashboard software that originate from changed names of existing products or from putting together pieces of software that already existed. Dashboards draw from multiple disciplines but ultimately use visualization (i.e. graphs) to communicate important information to stakeholders. Dashboards may collect, summarize, and present information from multiple sources such as legacy-, Enterprise Resource Planning systems, and business intelligence software. As far as data is concerned, a dashboard represents the tip of an iceberg, i.e. what the user sees at first and if needed, analyses further so as to uncover causes for poor performance.

60 _{Correspondence address: Ogan Yigitbasioglu, Post doc Researcher, Department of Accounting,}

~ 1005 ~

Visualization has been studied in a wide range of fields, i.e. history, art and visual studies, philosophy, sociology, architecture, science and technology, but in business, management and organization, it has been rather neglected (Fabbri et al., 2008). The importance of looking into visualization in the business context is more urgent than ever. The economic recession is leading to a growing interest in BI software, as better data analysis can help companies identify cost-cutting and sales opportunities (King, 2009). Under the current economic circumstances, companies are forced to focus on profit by making more accurate forecasts of sales and expenses and by “trying to align the expense stream with projected revenue stream” as John Van Decker says, who is the research vice-president at research firm Gartner (King 2009).

In every day organizational life, employees are faced with processing all sorts of data to accomplish tasks. Managers are overwhelmed with reports that they need to make sense of. Furthermore, reports are the output of information systems implemented in organizations. Most often, the layout and content of reports are designed by software vendors, which may not be optimal for a company in question. Dashboards have the potential to reduce excessive reporting and allow the user to focus on the relevant part of information on which to base important decisions.

The purpose of this paper is to provide a state of the art review of the role of dashboards in performance management. To this end, three specific goals are addressed. First, a review of the theoretical foundations being used to drive dashboards for PM is provided. Second, we briefly address the methodologies and approaches to designing and implementing dashboards. Third, we suggest a framework which could help future research in the area of visualization and dashboards in performance management.

1. PERFORMANCE MEASUREMENT AND VISUALIZATION

Performance measurement is an established concept. Performance measurement systems provide the means of monitoring and maintaining organisational control whereby the achievement of goals and objectives of an organization are ensured (Nani et al., 1990). Thus, “measurement provides the basis for an organisation to assess how well it is progressing towards its predetermined objectives, helps to identify areas of strength and weaknesses, and decides on future initiatives, with the goal of improving organizational performance” (Amaratunga and Baldry 2002, p. 217).

Traditionally, performance measurement was primarily based on accounting information such as summary measures (e.g. net income, operating profit) and ratios (e.g. ROI, ROE) (Merchant and Van der Stede, 2007). A turning point in the area of performance measurement occurred when Kaplan and Norton (1992) published their seminal paper titled “The Balanced Scorecard”. Kaplan and Norton (1992) suggested the use of both financial and non-financial measures in performance measurement. Thus, the Balanced Scorecard concept (BSC) (Kaplan and Norton, 1992, 1996) has become the best-known attempt to broaden performance measurement by including non-financial measures (relating to learning and growth, processes, and customers) aside from conventional accounting information. Typically, a BSC includes more than

~ 1006 ~

20 individual measures (4-7 measures in each dimension), which then make up the holistic performance rating of an organization (Neumann et al. 2008).

However, there is little guidance as to how and to what extent non-financial measures should be used (Corona 2009). According to Ittner and Larcker (2003), executives assign either equal weights to non financial measures, go with fashion or rely on their own assumptions about the measures’ strategic importance. Furthermore, Ittner and Larcker (2003) provided evidence that corporate managers routinely discount or ignore non-financial measures. One explanation for such behaviour could be the subconscious limitation to information processing because decision makers can only process the fraction of the available information (Neumann et al., 2008). Decision-makers are symbol manipulation systems that process information by structuring problem spaces and searching those spaces until a goal is achieved. The space search is limited by the human’s limited span of attention (e.g. the number of symbol structures to which a human can attend simultaneously) (Newell and Simon 1972). In line with the information processing theory, Neumann et al. (2008) suggested that evaluators’ judgements could be traced to less than 4 measures, which agrees with Halford et al. (2005). Hence, given that individuals have limited working memory store, they should rationally focus on information with increased likelihood of diagnostic capability.

Excessive information may not only lead to disregard of information but also to decision inaccuracy. Shields (1983) found an inverted U-relationship between the accuracy of decision-making and the quantity of information supplied. The inverted-U relationship indicates that there is a certain level of accounting information supplied in reports that results in the most accurate decisions. More or less information decreases the decision accuracy.

To overcome the abovementioned problems regarding complex judgement tasks, it has been suggested to supplement or alter the way information is displayed (Silver 1990, 1991). This kind of informative guidance (visualization) can help facilitate better decision making by making it easier to focus on the relevant portion of the information set (Dilla and Steinbart 2005). Thus, Banker et al. (2004) found that BSCs supplemented with strategy maps that show the linkages between performance measures and the different aspects of an organization’s strategy increases attention paid to those measures when evaluating divisional performance.

Although supplementary display tools are available in commercial BSC software (e.g. QPR, IBM Cognos), their effect on decision making quality is little known. According to Dilla and Steinbart (2005), judgement quality is affected by the display format. Judgement quality is a function of consistency (within an individual’s own decisions) and consensus (among individuals’ decisions). The two quality measures are considered important so as to ensure an objective and fair outcome for the performance evaluation (Malina and Selto 2001). The results obtained by Dilla and Steinbart (2005) were mixed as the absolute benefit of supplementary information (tabular versus graphic) to BSC was unclear. However, there was unambiguous evidence that tabular information was superior to the same information displayed in graphical format due to the adverse effect of graphical information on consensus (Dilla and Steinbart 2005). On the other hand, graphical presentations were found to improve the accuracy of bankruptcy forecasts, earnings forecasts and sales forecasts

~ 1007 ~

(Mackay and Villareal, 1987; DeSanctis and Jarvenpaa, 1989; Anderson et al, 1992). Other studies found that the influence of presentation formats interacts with the characteristics of the decision environment and the background of the decision maker. For example, tabular formats were found to be superior to graphical formats as the complexity of the tasks increased (Davis 1989; Liberatore et al. 1988). Furthermore, decision makers with a low level of accounting knowledge made decisions that lead to higher profits when they used customer profitability reports, which were presented in graphical format compared to tabular format (Cardinaels 2008). A surprising result is that the same customer profitability report presented as graphs (versus tables) had a negative effect on profits for users with high levels of cost knowledge. These findings emphasize that in order to enable effective decision-making, the accounting data may need to be presented in different presentation formats according to the level of accounting knowledge of the user. This also agrees with Davis (1989), who concluded that no one form of information presentation is suitable in all situations.

The use of graphs to report accounting information is not free of problems. Beattie and Jones (2000) pointed to a reporting bias in financial graphs in corporate annual reports, which were used to make a favourable impression on the company performance for annual reports readers. However, within the context of management accounting, researchers attempted to show that graphs should lead to superior decision-making performance compared to tables (Vessey, 1991). Vessey and Galletta (1991) concluded that there is no agreement in management accounting over which types of presentation format (i.e. table or graph) is better.

2. PRESENTATION FORMATS AND TASK CHARACTERISTICS: COGNITIVE FIT

Previous research on information display suggests that for example, the use of tables versus graphs depend on the nature of the task (Dilla and Steinbart 2005). Based on human information processing (HIP) theory, Vessey (1991) introduces the cognitive fit theory that explains under what circumstances one mode of information presentation outperforms the other. One underlying principle of this theory is that graphical and tabular representations present the same type of information but in fundamentally different ways: graphical representations are spatial information and tabular representations are symbolic information. Tasks can also be divided into spatial and symbolic tasks. According to cognitive fit theory, graphs are more useful for tasks that require identifying and understanding relationships and for making comparisons (i.e. spatial tasks), while tables are better for tasks that require extracting specific values and combining them into an overall judgement (i.e. symbolic tasks) (Umanath and Vessey, 1994; Vessey, 1991; Vessey and Galletta, 1991).

Huang et al. (2006) tested the cognitive fit theory in the context of expertise management systems by studying the potential of using visualization techniques to support more efficient and effective exploration of the information space. The results showed that visual representations of data generated by self-organizing maps (SOM) and multidimensional scaling (MDS) outperformed tabular representations for four out of the five low-level spatial tasks. A study by O’Donnell and David (2000) revealed that sixteen percent (16 %) of the articles reviewed examined the influence of presentation formats such as hypertext systems, geographical information systems

~ 1008 ~

(GIS), treemaps, and probability maps on decision making. The findings regarding the influence of the presentation format on decision performance were mixed.

Hasbun (2009) found that the group that based their decision about the sales forecasts on the information presented in tabular format made more accurate forecasting decisions compared to the group that based their decision on information presented in animated graphs. However, the group with decisions based on animated graphs felt that the visualization was more useful for selecting among alternative courses of action than the group with tables based decisions.

Cognitive fit theory is not limited to matching information presentation and task. Vessey and Galletta (1991) introduce an extension of the basic concept of cognitive fit, which includes the fit of an individual’s decision-making skills to the information presentation and the task. Effective or efficient decision-making is not only about matching the modes of information presentation to task. The user needs to develop appropriate mental representations for better decision-making performance. The mental processes that decision makers use provide the link between representation and task and are identified as being perceptual and analytical. This extension has implications for designers of decision support systems who should focus on the characteristics of tasks that decision-makers must address, and find appropriate information representations and support tools.

All in all, cognitive fit theory holds that as long as there is a fit between the mode of information presentation, processes and task type, the mode of information presentation will lead to both quicker and more accurate decision-making.

3. MYERS-BRIGGS TYPE INDICATOR IN ACCOUNTING

Personality theories have been greatly applied in different research fields, and to a lesser extent in accounting. Wheeler (2001) examines Jungian personality theory and Myers-Briggs Type Indicator (MBTI) to illustrate those aspects of personality that are more relevant to accounting research and education. MBTI instrument is built on Jungian theory of personality. Jung’s theory states that a person’s personality consists of the interaction between the way of perceiving (sensing or intuition) and the way of judging (thinking or feeling). The Jungian psychology focuses on conscious aspects of personality, decision making, and the effect of personality on understanding. Intuitive persons are more creative and insightful during decision-making processes whereas sensing individuals prefer to gather data from facts and observations. Thinking individuals are logical and rational whereas feeling individuals are more idealistic. According to Myers-Briggs Type Indicator (MBTI), there are 16 personality types that can be characterized by occupational and organizational traits, educational traits and learning styles, and decision-making traits and cognitive styles. In each person, there is an innate inclination towards one of the four traits, but the non-preferred traits are present too and the individual can be competent using them.

The accounting research that used MBTI covered mainly two topics: (1) personality types of accounting professionals and (2) correlations between the personality types of accounting professionals and performance. The results point to differences in the

~ 1009 ~

personality types of accountants within the various specialities of the profession. For example, an extrovert attitude is more likely to be found in national companies than in local companies. In addition, Chenhall and Morris (1991) and Vassen et al. (1993) found significant relationships between the MBTI preferences and decision making and cognitive style in practising accountants. Middle and senior level managers with an intuition preference included opportunity cost information in resource allocation decisions, whereas managers with a sensing preference ignored opportunity cost (Chenhall and Morris , 1991). According to Vassen et al. (1993), auditors have a significantly higher preference for the sensing type of information acquisition. Furthermore, the subjects with thinking preferences accessed more information and took more time to process the information accessed.

Findings on personality type of accountants are consistent with the sensing and thinking preference. This personality preference applies to individuals working in management and administration whose general personal characteristics are practical, sensible, decisive, logical, detached, observant, active, and rational problem solver (Myers, 1998 as quoted in Wheeler, 2001). There is therefore a relationship between the personality type and the role of individuals in organizations.

Depending on their role, decision-makers need to interact with data reflecting day-to-day operations or strategic choices for the company. A study by IBM defines the new roles of CFOs to go beyond reporting financial performance metrics to what activities to outsource and what markets to focus on (IBM 2010). CFOs are expected to analyze, report and advise CEOs to catch the competition off guard. Dashboards possess drill-down features that enable active interaction of decision-makers with the data at different levels of granularity and for different dimensions (e.g. product family, geographical areas). In order to obtain an effective and efficient use of dashboards, it may be worthwhile to design dashboards centred around the traditional and new roles of decision-makers.

4. PERSONALITY TYPE, PROBLEM-SOLVING SKILLS AND INFORMATION SYSTEMS FEATURES

Over the last ten years technology became pervasive in accounting profession and the use of information systems has become a must for the achievement of accounting tasks. In the competencies framework of The American Institute of Certified Public Accountants (AICPA), one of the functional competencies for entry into accounting profession is the necessary skill to use technology tools effectively and efficiently. O’Donnell and David (2000) proposed a unified framework from the accounting information systems (AIS) literature and the human information processing (HIP) literature to analyze how different information systems features can influence user decision making. This framework described judgement and decision making (JDM) as a function of one endogenous variable, processing strategy and three exogenous variables: (1) information system features, (2) the decision-making environment, and (3) problem-solving skills. These variables interact through cognition to produce decisions.

Information systems (IS) researchers used Myers-Briggs Type Indicator to measure the relationship between personality and information systems use (Bowen at al.,

~ 1010 ~

2003). Boon and Tak (1991) suggested that the customization of interfaces to match personality types may lead to successful use of information systems. Kostov and Fukuda (2001) found that users performed better when they handled an interface that matched their personality type. We argue that personality type is a relevant factor to investigate in relation with the design of efficient dashboards. There may be a mismatch between the personality characteristics of users of dashboards and designers of dashboards, hence the importance of including different cognitive elements into the design of dashboards.

Bowen et al. (2003) investigated how Myers-Briggs personality type can influence managerial end-user performance in database querying tasks. They found that perceiving managers who relied more on intuition composed queries more accurately than perceiving managers who relied more on sensing. Furthermore, perceiving managers composed more accurate queries as opposed to judging managers. However, Liberatore et al. (1988) and Carpenter (1993) found that personality type had no relationship to a person’s perception of the value of different data presentation formats.

Considering the contradictory findings regarding the fit between the features of an information system with the personality of the decision-makers on the performance of decision-making, we suggest that dashboards might have to be designed not with the personality type of the user in mind, but with the role of user in mind. According to previous research, accountants fall under the same personality type, with differences with the same type depending on the particular accounting profession. The different roles of accounting profession may serve as a proxy for the personality type.

O’Donnell and David (2000) reviewed papers that investigated other ways in which information systems may influence decision making: the use of a decision support systems, the effect of information aggregation and load, and the system’s facility to provide interaction and feedback. However, Vessey and Galletta (1991) concluded that providing decision support systems to satisfy individual managers’ desires will not have an efficient or effective effect on the performance of decision-making. Dashboards enable different information displays of performance metrics at different levels of aggregation through a dynamic user interaction. Hence, they provide the best of both worlds: graphs for comparing indicator performances at high level and the possibility to analyze data further at a detailed level, often presented in tables. Thus, dashboards can be used for both spatial and symbolic tasks. Latest generations of dashboards also enable “what-if” scenarios for selecting the course of action that leads to the most desired outcome. The novelty of dashboards is the dynamism of the information displayed. Therefore, dashboards possess all the capabilities to influence decision-makers involved in accounting processes. However, we perceive a gap between the promise of dashboards and the reality of accounting profession regarding the use of dashboards in the decision-making process. A more customized design may fill this gap.

~ 1011 ~

5. DASHBOARD DESIGN AND ROLE IN PERFORMANCE MEASUREMENT

There is not a clear definition of dashboards, neither given by software vendors nor by academics. The dashboard vendors define the dashboards from the perspective of characteristics that their products have. Researchers talk about different types of applications of the dashboard concept (Adam and Pomerol, 2008). A generic description of dashboards may be that of graphical user interface (GUI) that contains measures of business performance to enable managers’ decision making. This definition comprises the visual display of the dashboard concept, the content and the purpose for which dashboards are used. Few (2006) defined dashboards in terms of common characteristics of every example of dashboard that he could find on the web.

“A dashboard is a visual display of the most important information needed to achieve one of more objectives; consolidated and arranged on a single screen so the information can be monitored at a glance.” (Few, 2006, p. 34)

Dashboards that make use of visualization are cognitive tools that improve the control over business data (Brath and Peters, 2004).

The terminology originates from the vehicle dashboard, which reports the few metrics that the driver needs to know. Dashboards help managers to visually identify trends, patterns and anomalies about business (see Figure 1), which makes the issue of visual information design very important. Dashboards do not only help to monitor data but they enforce consistency across measures, help with the planning of strategy, and communicate to stakeholders the values of the company. According to Pauwels et al. (2009), dashboards and BSC’s have different perspectives. Whereas a BSC shows how a firm is currently performing, a dashboard provides the way forward. Hence they do not compete and are ideally used in combination, drawing from both of their strengths.

~ 1012 ~

Figure 1. Dashboard Example

(Source: http://www.smallbusinesscomputing.com)

Dashboard design, as many other visualization tools draw on the principles of visual perception. Visual perception can be explained through the application of Gestalt psychology to visualization. Gestalt psychology was born in reaction to atomism at the end of the 19th _{century with the view of things “as more than the sum of their}

parts”. The Gestalt psychologists were intrigued by the way our minds perceive wholes out of incomplete elements (Behrens 1984, Mullet and Sano 1995). Among the Gestalt principles that dashboards use are proximity, similarity, continuity, figure-ground, symmetry, and the closure of objects (Moore and Fitz 1993).

Dashboards often make use of colours to discriminate objects from one another or to recognize and identify them (Goldstein 2007). Although the use of colours may improve the process of visualization, excessive use of colours can distract the user and therefore have an adverse effect. Similarly, redundant visual information in graphs such as decorated frames and non value adding 3D objects could limit attention paid to important information. To remedy this problem, it has been suggested to maximize the data-ink ratio, which measures the proportion of ink used to represent data to the total ink used to print the graph (Tufte 2001).

Although it is strongly suggested that a dashboard fit on a single computer screen, the information displayed on dashboard should open the door to additional information (Few 2006). Thus new generation dashboards require point and click interactivity that allow users to drill down information so as to obtain further details on various performance measures. Furthermore, dashboards can also help users to identify

~ 1013 ~

measures that need immediate attention by visually alerting the user when performance indicators are out of range.

Overall, the requirements of new generation dashboards refer to (1) aligning business process with latest information to provide business intelligence at all levels of the company, (2) using intuitive and easy to digest visuals for delivering information to busy executives, and (3) visualize and navigate (Ziff Davis Enterprise 2009). The last two requirements highlight the importance of dashboards to provide data visualization in a way that makes sense to the individual. Thus, the essence of dashboards consists in the communication of rich and dense information to decision makers. If designed properly, dashboards can offer the solution for information overload regarding reporting in companies. Therefore, the complexity of dashboards comes from choosing the appropriate design that enables the managers to easily and efficiently monitor the data.

Dashboards in various formats have been in use by organizations. For example, a case study by Schulte (2006) evaluated the software BusinessObjects Dashboard Manager, which generates a dashboard, which makes it easier and faster for the patient accounts team to manage account receivables (A/R) of Edward hospital. The evaluation showed that A/R dashboard saved significant time each month by providing users with immediate access to current and accurate information. This helped the hospital improve its cash flow.

6. INFORMATION VISUALIZATION

Dashboards convey information through a process referred to as visualization. Information visualization refers to the “use of interactive visual representations of abstract, nonphysically based data to amplify cognition” (Card et al. 1999). The process of visualization is illustrated in Figure 2. Encoding and decoding is facilitated through the use of visual attributes such as shape, position and colour, and textual attributes such as text and symbols, which themselves are represented with simple visual attributes (Wunsche, 2004). Visualization is effective if the decoding is done “correctly”, where perceived data quantities and relationships between data reflect the actual data. Visualization is efficient if the maximum amount of data is perceived in a minimum amount of time. According to Friedman (1979) and Oliva (2005), visual perception involves two elements: the perceptual and conceptual gist. The perceptual gist refers to the process of the brain when it determines the image properties that provide the structural representation of a scene, like colour and texture. The conceptual gist refers to the meaning of the scene (what the viewer sees), which is improved after the perceptual information.

~ 1014 ~

Figure2.Information Visualization Process

(Source: Wunsche, 2004)

Dashboards can be evaluated according to how well they facilitate the encoding and decoding of information. Furthermore, a good balance between visual complexity and information utility is required. Visual complexity refers to “the degree of difficulty in providing a verbal description of an image” (Heaps and Handel 1999, Oliva et al. 2004). Visual complexity might increase with the quantity and range of objects as well as with varying material and surface styles (Heylighen 1997). Conversely, repetitive and uniform patterns and existing knowledge of the objects in the scene reduce visual complexity (Oliva et al. 2004).

7. DASHBOARD IMPLEMENTATION

The implementation of dashboards can be divided in to five distinct stages (Pauwels et al. 2009). In Stage 1, the key metrics for the company are selected. In Stage 2, the dashboard is populated with data. Stage 3 involves the establishment of cause and effect relationships between dashboard items, which adds a deeper understanding of the business. Stage 4 enables what if analysis for forecasting and budgeting. Finally, stage 5 connects all dashboard items to financial consequences (Pauwels et al. 2009). Hence, the maturity and success of a dashboard implementation will depend on how far the implementation is within its development cycle. Dashboard implementations should start from the end user, whose daily tasks include the monitoring of the selected metrics (Chiang 2009).

According to Kawamoto and Mathers (2007), one of the key success factors (CSF) in dashboard implementations is the identification of metrics. The metrics are desired to be relevant and meaningful for stakeholders. They should include a mix of operational, financial and project-specific information. This agrees with DeBusk et al. (2003), where non-financial measures are found to have a more significant role in PM so as to be able to influence financial measures that report on outcomes. Furthermore, dashboards should generate actions toward specific goals, be interactive and change over time depending on the new business conditions (Kawamoto and Mathers, 2007). The process of metric selection for the use in dashboards is outside the scope of this study.

~ 1015 ~

O’Donnell and David’s (2000) literature review on the information presentation variables that have an effect on decision-making shows that only three information systems features have been investigated to a great extent: (1) using versus not using decision support systems (DSS), (2) tabular versus graphical presentation format and (3) differences in the amount of information load available. To some extent dashboards as a decision support system avoid the problem of presentation formats and information load. The new generation dashboards often present information in both formats and at various levels of detail (through point and click) thereby reducing information overload.

Although dashboards seem to offer a panacea for the problems of presentation formats and information overload, we argue that the efficient and effective use of dashboards in performance measurement might depend on a number of factors. These factors are illustrated in Figure 2. The model is based on the cognitive fit theory according to which there should be a fit between the role of dashboards, the information presented on dashboards and the user background, which influences the usability of dashboards. The model presents potential research for the advancement regarding visualization in performance management by introducing an extension to the cognitive fit theory, the link between the usability of dashboards and the quality of decision making. According to the model, there needs to be a fit between the purpose of the dashboard and its features. For example, there will be a weak fit when the dashboard is intended to be used as a planning tool but when it lacks the features to carry out what if analysis. The dashboard may also be used as a tool to communicate strategy (as in the BSC) but it may not reflect this in terms of how the performance measures are displayed (e.g. the order of measures, display size, etc.). Even if there is a fit (i.e. all the required information and features are available to the user), a poor visual design (e.g. excessive use of colours, low data-to-ink ration, etc.) may confuse and distract the user.

Furthermore, user background in terms of education and experience and skills (e.g. IT skills) and the role of user in organization might have an effect on the usability of dashboards. If so, the knowledge obtained from such an investigation could be used to improve the design of dashboards or to advocate user training. For example, is it only the younger, less tenured and more technology oriented accountants who are more likely to use dashboards? On the other hand, if there is a weak or no link between user background/skills and usability, this might suggest that dashboard need not to be designed according to user background and skills. Hence, we suggest that both, the fit between its purpose and features as well as the visual design with the role will have an impact on to what extent users find the dashboard useful. Paying attention to these factors could improve the process of visualization and decision making quality.

~ 1016 ~

Figure 3.A Research Framework for Effective Dashboard Design and Use

Especially when companies go through recessions, the ability to access and quickly evaluate different aspects of a company’s performance is essential (Hanoa, 2009). In the current economic climate, Business Intelligence (BI) software may help companies to stay competitive by giving a complete overview of critical information at all times. In essence, BI software provides managers with logical connections between cause and effects within a company’s figures so that the issues can be proactively tackled. For example, BI software can show if the sales growth is caused by marketing expenditure or cash income. The biggest advantage brought by latest BI solutions is that it integrates external sources of information with intra-organizational information. Thus, they benchmark the company’s performance against external market metrics. Dashboards are such a BI solution. We argue that the information presented by dashboards can be turned into knowledge provided the dashboards are designed according to the groups of users’ responsibilities and requirements. Dashboards would present graphical or tabular representations, functions and capabilities that would be useful for users in different roles.

If dashboards are the technology that makes explicit the connections between causes and effects in company’s figure, we can assume that the management is more aware of what the drivers of profit growth are. Hence, another research stream based on

Dashboard purpose: - Monitoring - Communication - Planning Dashboard features: - Automated Alerts - Limited/no customisability - What-if analysis Dashboard design: - On single page - Use of colours - Data-ink ratio User background: - MSc (Econ) - MBA - Lon-term tenures - IT training User role: - CFO - Business controller Dashboard usability: - Perceived Usefulness - Perceived Ease of Use - Satisfaction

Decision Making Quality:

- Speed

- Accuracy

- Consistency

Fit

~ 1017 ~

empirical research is to investigate if companies that use this type of innovative management accounting systems are more likely to perform better compared to companies that don’t implement dashboard solutions.

We suggest design science research and behavioural research methodologies be applied intertwined to study the relationships between the constructs suggested in Figure 3. For example, design research can be applied to model and deploy innovative dashboard designs to address different business performance management (BPM) questions at different levels in organization (top level, middle level or operational level). The focus could be on finding dashboard design techniques that enable the process of dashboard design. The main goal could be to build dashboard designs that can be easy to maintain based on the latest changes in the competitive context.

An example of design science research that can be borrowed to the dashboard design context is the study of Helfman and Goldberg (2007) who suggested a decision support tool for selecting appropriate graph visualizations based on characteristics of data, task and end user. More specifically, they suggested the selection of appropriate graphs for users of business applications be based on a table with questions such as whether the data are categorical or quantitative, whether the tasks are about comparison or identification of trends or totals, and whether the end users are experts or casual users of graphs.

Creating dashboard prototypes is not an end in itself. Concurrently, usability evaluation methods should be conducted to verify the impact on users. We agree that behavioural research, with its theoretical underpinnings embedded in psychological theory, can provide the background for the usability evaluation. Enterprise software developers embed dashboards into the Business Intelligence software packages offered to companies. For what type of decisions are the dashboards used in companies? Is the information provided by dashboards perceived to be useful by makers? Is the use of dashboards bringing any value-added to the decision-making process? Management accounting researchers can have a valuable contribution here, by identifying what performance management issues can be addressed in an efficient manner by different data visualization techniques embedded in dashboards for different users. Can the information presented on dashboards be a complement to the cognitive capacity of decision makers to consider all environmental events and possible actions? Can dashboard designs help decision makers generate creative ideas? Can dashboards be used to boost the confidence and justify the decisions made?

Dashboards can be very important communication tools of organizational knowledge. Based on how accurate and informative they are, the decision-makers can rely more on them to make educated decisions. The fast speed of technological development makes it difficult if not impossible for business managers to adopt these developments into the business practices. Business Intelligence solutions can improve the business, one of the crucial questions is “how can these solutions be designed to match the users requirements”. Academic research can find answers to this question. Research on dashboard design is one track where researchers can make a difference related to the role that dashboards can play in the decision making related to performance management.

~ 1018 ~

The aim of this paper was to provide an overview of the current state-of-art regarding the role of dashboards in business performance management. Dashboards are often an integral part of Business Intelligence (BI) software, which have recently received widespread attention from the business community. Dashboards are the graphical user interfaces between the BI technologies and the users, which communicate information through a process referred as information visualization To be effective, dashboards need to adhere to “good design principles” such as refraining from the use of excessive colours and maximising the data-ink ratio.

A full dashboard implementation consist of five steps: 1) selection of key metrics; 2) data population; 3) identifying relationships between dashboard items; 4) adding new features, i.e. what if analysis; and 5) linking dashboard items with their financial consequences. Dashboards in the final stage can provide their full potential benefits to companies that adopt them. Given the right design, dashboards can provide various advantages to decision makers. Dashboards can balance the attention paid to financial and non-financial ratios when making strategic decisions. Also, the process of designing dashboards may help to communicate strategy and clarify priorities in terms of which performance measures to monitor more closely.

Individuals have limited working memory store which may often lead to some of the valuable information to be disregarded during decisions making. Dashboards may reduce this effect by enabling the user to focus on the most important/urgent part of the information. Thus, through selective and layered information presentation, the decision accuracy can be improved by providing the right amount of information to the user. Otherwise, more or less accounting information may lead to lower decisions accuracy (Shields 1983).

The cognitive fit theory holds that as long as there is a fit between the mode of information presentation, processes and task type, the mode of information presentation will lead to both quicker and more accurate decision-making (Vessey 1991). Dashboards offer an elegant solution to the problem of presentation: they provide best of both worlds: information in the format of graphs and tables using point-and-click. Personality theories state that individuals can be categorized into personality types based on how perceiving or judging they are in their mental processes. Research shows that accountants are individuals who prefer to gather data from facts and observations, logical and rational. This might have percussion on how dashboards are to be designed.

Over the last years, the use of information systems has become a must for the completion of accounting tasks. The use of efficient and effective technological tools is a necessary functional skill for entry into accounting profession. Furthermore, the quality of decision-making depends on the features of information systems, the decision-making environment and the problem-solving skills (O’Donnell and David 2000). The literature is also in disagreement over whether information system interfaces should be designed according to the personality type of the user or not. This might also be relevant to dashboards although dashboards can be to some extent customized. Nevertheless, more research into dashboard use is needed so as to

~ 1019 ~

maximise their benefits to organizations that use them. We have suggested a model, which identifies important factors from a design point of view and that could affect the usability of dashboards. The factors relate to the fit between the purpose and features of the dashboard, design quality, user background, as well as user role. These links offer new research possibilities, which could help designers and users of dashboards in the future.

REFERENCES

Adam, F. and Pomerol, J-C. (2008) “Developing Practical Decision Support Tools Using Dashboards of Information”, International Handbooks on Information Systems VII Amaratunga, D. and Baldry, D. (2002) “Moving from performance measurement to

performance management”, Facilities, Vol. 20, no. 5/6:217-223

Anderson, J. C. (1989) “A theory of the origins of human knowledge”, Artificial Intelligence, Vol. 40, no. 1-3:313-351

Anderson, J. C., Kaplan, S. E., and Reckers, P.M.J (1992) “The Effects of Output Interference on Analytical Procedures Judgements”, Auditing, Vol. 11, no. 2:1

Banker, R.D., Chang, H., and Pizzini M.J (2004) “The balanced scorecard: judgemental effects of performance measures linked to strategy”, Accounting Review, Vol. 79, no. 1:1--23

Beattie, V. A., and Jones, M. J. (2000) “Changing Graph Use in Corporate Annual Reports”, Contemporary Accounting Research, Vol. 17, no. 2:213-226

Behrens, R. (1984) Design in the visual arts. Englewood Cliffs, NJ: Prentice-Hall, Inc, 1984. Bowen, P. L., Ferguson, C. B., Lehmann, T. H., and Rohde, F. H. (2003) “Cognitive style

factors affecting database query performance”, International Journal of Accounting Information Systems, no. 4:251-273

Boon, W.T. and Tak, W.L. (1991) “The impact of interface customization on the effect of cognitive style on information system success”, Behaviour & information technology, vol. 10, no. 4:297-310

Brath, R. and Peters, M. (2004) “Dashboard Design: Why Design is Important”, DM Direct, 2004

Card, S. K., Mackinlay, J.D., and Schneiderman, B. (1999) “Readings in Information Visualization: Using Vision to Think”, Morgan Kaufmann; 1st edition, 1999

Cardinaels, E. (2008) “The interplay between cost accounting knowledge and presentation formats in cost-based decision-making”, Accounting, Organizations and Society, no. 33:582-602

Carpenter, D.A., Anders, J., and Anderson, A. (1993) “Influence of Myers-Briggs type on preference for data presentation format”, The Journal of Computer Information Systems, vol. 33, no. 4:85-90

Chenhall, R. and Morris, D. (1991) “The effect of cognitive style and sponsorship bias on the treatment of opportunity costs in resource allocation decisions”, Accounting, Organizations and Society, vol. 16, no. 1:27-46

Chiang, A. (2009) “Creating Dashboards: The Players and Collaboration You Need for a Successful Project”, Business Intelligence Journal, vol. 14, no. 1:59-63

Corona, C. (2009) “Dynamic Performance Measurement with Intangible Assets”, Rev Account Stud, no. 14:314-348

David, J. S. and Steinbart, P. J. (1999) “Drowning in data. How you can turn oceans of data into useful information”, Strategic Finance, vol. 81, no. 6:30-34

Davis, L. R. (1989) “Report format and the decision maker’s task: An experimental investigation”, Accounting, Organizations and Society, vol. 14, no. 5-6:495-508

DeBusk, G. K., Brown, R. M., Killough, L. N. (2003) “Components and relative weights in utilization of dashboard measurement systems like the Balanced Scorecard”, The British Accounting Review, no. 35:214-231

~ 1020 ~

DeSanctis, G. and Jarvenpaa, S.L. (1989) “Graphical presentation of accounting data for financial forecasting: an experimental investigation”, Accounting, Organizations and Society, vol. 14, no. 5-6:509—525

Dilla, W.N. and Steinbart, P.J. (2005) “The effect of alternative supplementary display formats on balanced scorecard judgements”, International Journal of Accounting Information Systems, no. 6:159--176

Fabbri, P., Mackenzie, D., and Thrift, N. (2008) “Imagining Business”, Accounting, Organizations and Society, no. 33:103--104

Few, S. (2006) “Information Dashboard Design, The Effective Visual Communication of Data”, O’Reilly Media, Inc, First Edition 2006

Friedman, A. (1979). “Framing pictures: the role of knowledge in automatized encoding and memory of gist”, Journal of Experimental Psychology: General, vol. 108, 316-355 Gilad, T. and Gilad, B. (1986) “Business intelligence: The quiet revolution”, Sloan

Management Review, vol. 27, no. 4:53-61

Goldstein, E.B. (2007) “Sensation and Perception”. Thomson Wadsworth, 7th _{Edition 2007.} Halford, G.S.R., Baker, R., McGredden, J.E. and Bain J.D. (2005) “How many variables can

human process?”, Psychological Science, vol. 16, no. 1:70-76 Hanoa, E. (2009 ) “Brain Gain”, Accountancy Age, 2009:21

Hasbun, A. (2009) “An Empirical Investigation: Do Animated Graphs Improve the Quality of Sales Forecasting Decisions in Comparison to Tables?”, Master Thesis, 2009, Hanken School of Economics

Heaps, C., & Handel, C.H. (1999) “Similarity and features of natural textures”, Journal of Experimental Psychology: Human Perception and Performance, no. 25:299-320 Helfman, J. I. and Goldberg, J. H. (2007) “Selecting the Best graph based on Data, Tasks, and

User Roles”, UPA 2007 Conference Patterns: Blueprints for Usability, June 11-15,

2007, Austin, Texas, USA,

http://www.upassoc.org/usability_resources/conference/2007/prp_049.pdf, Last accessed on 25.3.2010

Heylighen, F. (1997) “The Growth of Structural and Functional Complexity during Evolution”, F. Heylighen & D. Aerts (eds.) 1997

Huang, Z., Chen, H., Chen, G., Fei, X., Jennifer, J., Wu, S. and Chen, W-H. (2006) “Expertise visualization: An implementation and study based on cognitive fit theory”, Decision Support Systems, no. 4:1539--1557

IBM (2010), The New CFO: Chief Focus Officer: http://www-03.ibm.com/innovation/ca/en/operations/cfo.shtml) Last accessed on 22.3.2010

Ittner, C.D. and Larcker, D.F. (2003) “Coming up short on nonfinancial performance measurement”, Harvard Business Review, vol. 81, no. 11:88-95

Ittner, C.D., Larcker, D.F., Meyer, M.W. (2003) “Subjectivity and the weighing of performance measures: Evidence e from a Balanced Scorecard”, The Accounting Review, no. 78:725-758

Kaplan, R.S. and Norton, D.P. (1992) “The balanced scorecard: Measures that drive performance”, Harvard Business Review, vol. 70, no. 1:71-79

Kaplan, R.S., Norton, D.P. (1996) “The balanced scorecard”, Boston MA, Harvard Business School Press.1996

Kawamoto, T. and Mathers, B. (2007) “Key Success Factors for a Performance Dashboard”, DM Review, 2007:20-21.

Keen, P. G. W. (1978) “Decision support systems: an organizational perspective”, Reading, Mass., Addison-Wesley Pub. Co. ISBN 0-201-03667-3 1978

King, R. (2009) “Business Intelligence Software’s Time is Now”, Business Week, Special

Report, 2009:

http://www.businessweek.com/technology/content/mar2009/tc2009032_101762.htm Last accessed on 25.3.2010

Kostov, V. and Fukuda, S. (2001) “Development of Man-Machine Interfaces Based on User Preferences”, Proceedings of the 2001 IEEE International Conference on Control Applications, 5-7 July, Mexico City, Mexico, 2001:1124-1128

~ 1021 ~

Liberatore, MJ., Titus, GJ., and Dixon, PW. (1988) “The Effects of Display Formats on Information Systems Design”, Journal of Management Information Systems, vol. 5, no. 3:85-99

MacKay, D. B. and Villarreal A. (1987) “Performance differences in the use of graphic and tabular displays of multivariate data”, Decision Sciences, vol. 18, no. 4:535-546

Malina, M.A. and Selto, F.H. (2001) “Communicating and controlling strategy: an empirical study of the effectiveness of the balanced scorecard”, J Management Accounting Research, no.13:47-90

Merchant, K.A., Van der Stede, W.A. (2007) “Management Control Systems”, Prentice Hall (Second Edition) 2007

Mullet, K. & Sano, D. (1995) “Designing visual interfaces: Communication oriented techniques”, Englewood Cliffs, NJ: Prentice Hall 1995

Moore, P. & Fitz, C. (1993) “Gestalt theory and instructional design”, Journal of Technical Writing and Communication, vol. 23, no. 2:137-157

Nani, A.J., Dixon, J.R., Vollman T.E. (1990) “Strategic Control and Performance Measurement”, Journal of Cost Management, 33-42

Negash, S. and Gray, P. (2008) “Business Intelligence”, International Handbook on Information Systems, Handbook on Decision Support Systems, Springer Berlin Heidelberg, 2008:175-193

Neumann, B.R., Roberts, M.L., Cauvin E. (2008) “Financial and Nonfinancial Performance Measures”, Cost Management, vol. 22, no. 6:5-14

Newell, A., and Simon H.A. (1972) “Human problem solving”. Englewood Cliffs, NJ: Prentice Hall 1972

O’Donnell, E. and David, J. S. (2000) “How information systems influence user decisions: a research framework and literature review”, International Journal of Accounting Information Systems, no. 1:178-203

Oliva, A. (2005) “Gist of the scene”. In the Encyclopedia of Neurobiology of Attention. L. Itti, G. Rees, and J.K. Tsotsos (Eds.), Elsevier, San Diego, CA, 2005:251-256

Oliva, A., Mack M.L., Shrestha M., and Peeper A. (2004) “Identifying the perceptual dimensions of visual complexity of scenes.”, In Proceedings of the 26th Annual Meeting of the Cognitive Science Society, 2004

Pauwels, K., Ambler, T., Bruce, H. C., LaPointe, P., Reibstein D., Skiera, B., Wierenga, B. and Wiesel, T. (2009) “Dashboards as a Service: Why, What, How, and What Research Is Needed?”, Journal of Service Research, vol. 12:175-189

Peters, J. M. (1993) “Decision making, cognitive science and accounting: an overview of the intersection”, Accounting, Organizations and Society; vol. 18, no. 5:383-405

Quattrone, P. (2009) “Books to be practiced: Memory, the power of the visual, and the success of accounting”, Accounting, Organizations and Society, no. 34:85-118

Rasmussen, N., Chen C. Y., and Bansal M. (2009) “Business Dashboards, A Visual Catalogue for Design and Deployment”, John Wiley & Sons, Inc, Hoboke, New Jersey, 2009

Shields, M. D. (1983) “Effects of Information Supply and Demand on Judgment Accuracy: Evidence from Corporate Managers”, The Accounting Review, vol. 58, no. 2:284-303 Schulte, M. F. (2006) “BusinessObjects Dashboard Manager”, DM Review, vol. 16, no. 2:49-- Silver, M.S. (1990) “Decision support systems: directed and nondirected change”,

Information Systems Research, vol. 1, no. 1:47-70

Silver, M.S. (1991) “Decisional guidance for computer-based decision support”, MIS Quarterly, vol. 15, no. 1:105-122

Teague, J. (1998) “Personality type, career preference and implications for computer science recruitment and teaching”, ACM International Conference Proceeding Series; Proceedings of the 3rd Australasian conference on Computer Science Education, The University of Queensland Australia, vol. 3:155--163

Tufte, E. (2006) “Visual Display of Quantitative Information”. Graphics Press, 2006

Umanath, N.S. and Vessey, I. (2004) “Multiattribute data presentation and human judgment: a cognitive fit perspective”, Decision Sciences, no. 25:5/6:795–-824

~ 1022 ~

Vaassen, E., Baker, C., and Hayes, R. (1993) “Cognitive styles of experienced auditors in Netherlands”, British Accounting Review, no. 25:367-382

Vessey, I. and Weber, R. (1986) “Structured Tools and Conditional Logic: An Empirical Investigation”, Communications of the ACM, vol. 29, no. 1:48--57

Vessey, I. (1991) “Cognitive Fit: A Theory-Based Analysis of the Graphs Versus Tables Literature”, Decision Sciences, vol. 22, no. 2:219--240

Vessey, I. and Galletta, D. (1991) “Cognitive Fit: An Empirical Study of Information Acquisition”, Information Systems Research, vol. 2, no. 1:63-84

Ziff Davis Enterprise (2008) “Reaping the benefits of next generation dashboards”, White Paper, available at http://portals.tdwi.org/whitepapers/2009/02/the-next-generation-of-enterprise-reporting-business-intelligence-reporting-tools-birt.aspx, pp. 1-4, Last Accessed on 7.5.2009

Wheeler, P., (2001) “The Myers-Briggs Type Indicator and Applications to Accounting Education and Research”, Issues in Accounting Education 2001

Wunsche, B. (2004), “A survey, classification and analysis of perceptual concepts and their application for the effective visualization of complex information”, Proceedings of the 2004 Australasian symposium on Information Visualisation, Darlinghurst, Australia, Australian Computer Society, Inc 2004:17--24