Accounting and Finance 47 (2007) 335–355

Factors influencing organizations to improve data quality

Accounting
S. W.
TheTee
1ORIGINAL
Blackwell
Oxford,
Accounting
ACFI
©
0810-5391
47 and
ARTICLE
etPublishing,
al./
Authors
UK and Finance
Journal 47 (2006) 00–00
Ltd.
compilation © 2006 AFAANZ

in their information systems

Sing What Teea, Paul L. Bowenb, Peta Doylec, Fiona H. Rohdec

a
School of Business, James Cook University, Cairns, 4878, Australia
b
College of Business, Florida State University, Tallahassee, 32306-1110, USA
c
UQ Business School, The University of Queensland, Brisbane, 4072, Australia

Abstract

Although managers consider accurate, timely and relevant information as

critical to the quality of their decisions, evidence of large variations in data quality
abounds. This research examines factors influencing the level of data quality within
a target organization. The results indicate that management’s commitment to
data quality and the presence of data quality champions strongly influence data
quality in the target organization. The results also show that the managers of
the participating organization are committed to achieving and maintaining high
data quality. However, changing work processes and establishing a data quality
awareness culture are required to motivate further improvements to data quality.

Key words: Case study; Champions; Data quality; Management commitment

JEL classification: C52

doi: 10.1111/j.1467-629x.2006.00205.x

1. Introduction

1.1. Background

Almost every activity in which organizations engage involves data. Data

provide the foundation for operational, tactical and strategic decisions. As data
become increasingly important resources in supporting organizational activities,
the quality of the data that managers use becomes critical (Paradice and Fuerst,
1991). Poor-quality data, if not identified and corrected, can have significant
negative economic and social impacts on the health of the organization (Wang
and Strong, 1996; Ballou et al., 2004).

Received 15 September 2005; accepted 30 June 2006 by Robert Faff (Editor)

© The Authors
Journal compilation © 2007 AFAANZ
336 S. W. Tee et al. /Accounting and Finance 47 (2007) 335–355

Anecdotal and empirical evidence of widespread poor data quality exists

(Redman, 1996; Klein et al., 1997; Huang et al., 1999). These impacts range
from operational inconvenience to ill-informed decision-making, to disruption
of business operations, and possibly even to organizational extinction. Some
specific examples of evidence include Hudson Foods who in 1997 lost its largest
customer, Burger King, due to Escherichia coli bacteria contamination that
caused several illnesses. Poor data quality relating to knowledge about which
batches were mixed caused the delivery of contaminated hamburgers to Burger
King. The contamination resulted in 25 million pounds of meat being recalled
– the largest recall in US history. Without their largest customer, Hudson Foods
was not profitable and was acquired by Tyson Foods (Belluck, 1997). In another
case, English (1999) reports that two 20 year old calculation errors in Los Angeles
County’s pension systems resulted in $US1.2bn in unforeseen liabilities. The
county must spend an additional $US25m each year for the next 50 years to
make up for the shortfall.
Data quality researchers recommend that organizations treat data as strategic
corporate resources for competitive advantage (Redman, 1995; Wang, 1998).
Nonetheless, most organizations admit they do not manage data as well as they
manage human and financial resources (Levitin and Redman, 1998). Empirical
evidence also indicates that many information systems contain substantial errors.
Organizational databases with error rates up to 30 per cent are typical in industry
(Redman, 1996) and mission-critical databases generally contain errors ranging
from 1 to 10 per cent (Klein et al., 1997). Poor data quality is estimated to cost US
businesses more than $US600bn a year (TDWI, 2002). Research evidence indicates
that organizations are aware that poor data quality is affecting their business.
Nevertheless, few organizations appear to be actively engaged in systematic
efforts to reduce data problems (see, e.g. TDWI, 2002).
A number of data quality frameworks have been developed to organize and
structure data quality dimensions. Organizations can use data quality frameworks
to understand data quality dimensions; for example, accuracy, timeliness, relevancy,
completeness and reliability (Huh et al., 1990; Ballou and Pazer, 1995; Wang
et al., 1995; Cappiello et al., 2004). They can also use these frameworks to
assist them in developing procedures to measure data quality and to investigate
its relationship to organizational processes. However, these frameworks do little
to increase our understanding of how organizations identify and resolve data
quality problems and, in particular, what factors influence an organization to
improve the quality of its data.
The goal of this research is to identify a number of factors and test whether they
affect the data quality within an organization. This research benefits organizations
in several ways. First, managers will be better able to identify critical factors for
successfully implementing new data quality initiatives and for nurturing existing data
quality activities. Second, managers will be better able to understand the rela-
tionships among these critical success factors. Third, they can use their improved
understanding to develop or improve their organizational data quality policies.

© The Authors
Journal compilation © 2007 AFAANZ
 S. W. Tee et al. /Accounting and Finance 47 (2007) 335–355 337

2. Theoretical foundations and hypothesis development

2.1. Data quality dimensions and definition

Numerous attempts have been made to define data quality and to identify
its dimensions (Fox et al., 1994; Wang and Strong, 1996; Kahn et al., 2002).
Dimensions of data quality typically include accuracy, reliability, importance,
consistency, precision, timeliness, fineness, understandability, conciseness and
usefulness. Unfortunately, a set of data might be completely satisfactory on
most dimensions but inadequate on a critical few. Improving one data quality
dimension can impair another dimension (Ballou et al., 1998). Moreover, different
stakeholders in an organization might have different data quality requirements
and concerns (Giannoccaro et al., 1999; Lee and Strong, 2004).
Within an organizational context, if better data quality is to be achieved, it is
important to understand what data quality means to that organization and also
how data quality is to be measured (Wand and Wang, 1996). Based on data
quality dimensions identified in prior research and on the perceptions of the
participating organization this research focuses on three data quality dimen-
sions: accuracy, timeliness and relevance. Accuracy refers to the degree of cor-
respondence of recorded values to the actual values of the associated real-world
objects. Timeliness refers to the extent to which the data are up-to-date for the
required task. Relevance refers to the extent to which the data are applicable or
appropriate for the required task.
To develop the research hypotheses, several theories, models and frameworks
were reviewed to identify factors that potentially influence an organization to
improve the quality of its data. These include data quality models and frameworks,
information systems implementation and data warehouse success models, total
quality management concepts, and the resource-based view of the firm. Figure 1
shows the hypothesized relations between the research constructs.

2.2. Hypothesis development

2.2.1. Management commitment

Tubbs (1993) defines commitment as strength of intention. Commitment affects

the persistence of behaviour (Salancik, 1977). In the context of this research,
management commitment is defined as the strength of management intentions to
achieve high data quality.1 Prior research has shown that management commitment
influences the extent to which total quality programmes are successful (Saraph et al.,
1989; Anderson et al., 1995; Flynn et al., 1995; Black and Porter, 1996). Hence,

1
Management here is not confined to top or senior management, but refers to all levels of
management in the organization.

© The Authors
Journal compilation © 2007 AFAANZ
338 S. W. Tee et al. /Accounting and Finance 47 (2007) 335–355

Figure 1 Hypothesized relationships.

H1: Management commitment to data quality is positively associated with the

level of data quality achieved.

2.2.2. Data quality champions

Data quality champions are managers who actively and vigorously promote
their personal vision for using data quality-related technology innovations. They
push projects over approval and implementation hurdles (Beath, 1991). Data
quality champions provide political support, keep participants informed, and
allocate resources to data quality projects (Flynn et al., 1995; Oz and Sosik, 2000).
Data quality champions also exhibit transformational leadership behaviour when

© The Authors
Journal compilation © 2007 AFAANZ
 S. W. Tee et al. /Accounting and Finance 47 (2007) 335–355 339

they strongly support a data quality project (Howell and Higgins, 1990; Heng
et al., 1999; Poon and Wagner, 2001). They possess the skills (e.g. communica-
tion and project management) and power (e.g. reputation and position in the
organization) needed to overcome resistance that might arise when change
occurs within organizations (Guimaraes and Igbaria, 1997; Jiang et al., 2000).
Accordingly,

H2: The presence of data quality champions is positively associated with

management’s commitment to data quality.

2.2.3. Extrinsic rewards

The use of extrinsic rewards (financial and non-financial) as a means of

controlling, managing and enhancing performance has been well established in
marketing, sales force development, and new product development (Ingram and
Bellenger, 1983; Sarin and Mahajan, 2000). Data quality-related extrinsic rewards,
such as recognition for data quality improvement suggestions, increased budgets for
data quality activities, positive feedback, and training (Nambisan et al., 1999), affect
the successful implementation of data quality initiatives. The type and level of
rewards that organizations provide for data quality initiatives reflect management’s
commitment to data quality. Therefore,

H3: Extrinsic rewards are positively associated with management’s commitment

to data quality.

2.2.4. Perceived usefulness of data quality as a strategic resource

An organization’s resources include all assets, capabilities, organizational

processes, attributes, information and knowledge that enable the organization to
conceive and implement strategies that improve its efficiency and effectiveness
(Barney, 1991). Strategic resources are rare, difficult to imitate and non-
substitutable. Use of data quality for competitive advantage means organizations
use high-quality data as strategic resources to earn long-run abnormal returns.
If managers recognize that data quality can provide strategic advantages, they are
more likely to commit to achieving high-quality data within their organizations.
Therefore,

H4: The perceived usefulness of data quality as a strategic resource is positively

associated with management’s commitment to data quality.

2.2.5. Information systems capability

Organizational information systems/information technology capabilities refer

to an organization’s ability to assemble, integrate and deploy information systems/

© The Authors
Journal compilation © 2007 AFAANZ
340 S. W. Tee et al. /Accounting and Finance 47 (2007) 335–355

information technology-based resources, usually in combination with other

resources (Grant, 1991, 1995; Bharadwaj, 2000). An organization’s capability to
use data quality as a source of competitive advantage has two major components
(Grant, 1995). First, a physical infrastructure comprised of computers, com-
munication technologies, sharable technical platforms, and integrated databases
is required. Second, appropriate human resources are required to support use
of data quality as a competitive resource. These include training, experience,
relationships, business skills, technical information technology skills, and
competencies in emerging technologies. They also include managerial skills
and leadership skills (Copeland and McKenney, 1988; Barney, 1991; Grant,
1995). Organizations with strong information systems/information technology
capabilities are better able to recognize and exploit data quality as a strategic
resource. Hence,

H5: Information systems/information technology capabilities are positively

associated with the perceived usefulness of data quality as a strategic
resource.

2.2.6. Perceived need for data quality to support products and services

The value of the products and services organizations offer often depends, in
part, on the quality of the data associated with these products and services. The
quality of data about products and services influences customers’ perceptions
about the quality of products and services organizations offered (Wang and
Strong, 1996). Hence, the impact of data on the value of the products and
services offered by an organization is likely to increase management’s perception
of the need for data quality to support their products and services. Accordingly,

H6: The perceived need for data quality to support products and services is
positively associated with management’s commitment to data quality.

2.2.7. Regulatory requirements

The level of data quality associated with the products and services organizations
offer is often dictated by legal or regulatory constraints. Organizations must
comply with the Privacy Legislation (Gibbs et al., 2002) and the Data Quality
Legislation (Anderson, 2002), which prescribe how organizations should collect,
use, secure and disclose information. Regulatory requirements increase the
organizations’ perceived need for data quality in their products and services
(e.g. to avoid the costs of sanctions or to take advantage of opportunities that
regulations provide to their organizations). Therefore,

H7: The need to comply with regulatory requirements is positively associated

with the perceived need for data quality to support products and services.

© The Authors
Journal compilation © 2007 AFAANZ
 S. W. Tee et al. /Accounting and Finance 47 (2007) 335–355 341

2.2.8. Contractual requirements

Organizations often need high-quality data because of contractual obligations

they have to their customers.2 Increasing the requirements for data quality to
support contractual obligations is likely to increase management’s commitment
to attain high levels of data quality. Therefore,

H8: The need to comply with contractual requirements is positively associated

with the perceived need for data quality to support products and services.

2.2.9. Competitive pressures

Competitive pressures drive organizations to improve the quality of the products

and services they provide to customers. Customers are likely to be dissatisfied
if they are wrongly billed.3 Competitive pressures increase the need to improve
the quality of data associated with an organization’s products and services
(Redman, 1995, 1996). Hence,

H9: Competitive pressures are positively associated with the perceived need for
data quality to support products and services.

3. Research method

3.1. Research setting and design

This research investigated a single case study organization (Zelda)4 using a

combination of data collection methods. The study consisted of a data-quality
survey of senior managers and general users as well as interviews with senior
managers. Zelda is a government-funded service organization in Australia.
Zelda employs just over 300 staff of which approximately 50 per cent are
tertiary-qualified professionals. Zelda provides three types of specialized
services: information, advisory (approximately 63 700 per year provided by

2
For example, Ford Motor Company terminated its century-old business relationship with
Bridgestone/Firestone. Failure to comply with the safety regulatory requirements in relation
to safety-test information cost Ford millions of dollars. In addition, the publicity negatively
affected customers’ perceptions about the quality of both Ford’s and Firestone’s products
(Whalen, 2001).
3
For example, AT&T, a long-distance call service provider, used local telephone companies
for access service to connect calls to its customers. The local telephone companies billed
AT&T for these access services, typically $US15bn per year. AT&T often found itself over-
charged for the access service (Redman, 1996, p. 17).
4
Zelda is a fictitious name. The potential limitation regarding the generalizability of results
is examined in Section 5.

© The Authors
Journal compilation © 2007 AFAANZ
342 S. W. Tee et al. /Accounting and Finance 47 (2007) 335–355

Zelda’s professional staff), and practical assistance (approximately 30 500 per

year: 30 per cent provided by Zelda’s professional staff and 70 per cent by
external professional staff). These services are provided through a head office,
a network of regional offices, a panel of several hundred professional service
suppliers, and a client information service accessible from anywhere in the state.
In 1994, Zelda adopted total quality service as its business philosophy. By
embracing the total quality service concept, Zelda is committed to deliver
high-quality, effective and efficient services through technology and innovation.
In 1998, the senior management of Zelda perceived a need for high-quality data
to support their operations and client services. They designed and implemented
Vision to improve operational efficiency and to further improve the quality of
their services. Zelda uses this mission-critical information system to create, store,
and maintain its clients’ information, record services provided to the public,
and report to stakeholders on its performance. After gaining an understanding
of the business, business processes, and the software associated with these
activities a data quality survey was prepared.

3.2. Survey

The first technique used to gather data was a survey of general users and of
senior managers. The survey responses were used to test the research hypotheses
and as the basis of formulating questions for follow-up interviews.
Measures for seven constructs were adapted from existing instruments. The
remaining measures were developed by the researchers and went through
extensive pretesting to ensure construct validity. Table 1 contains the constructs,
descriptions and source for the questions making up the survey instrument.
All constructs except information systems/information technology capabilities
were measured using multiple items. Information systems/information technology
capabilities were measured by the participants’ self-reported information systems/
information technology experience. To obtain measures for each construct, parti-
cipants were asked to mark their perceptions on a continuous scale of 0 (strongly
disagree) to 1 (strongly agree).5 The participant’s score on each question was
the ratio of the marked distance (from 0 to x) to the total distance (from 0 to 1).
Only members of senior management were able to answer questions related
to the perceived use of high-quality data as a strategic resource for competitive
funding and the perceived need for high-quality data to support operations and
client services. As a result, two sets of questionnaires were prepared: one set
for general users and the other set for senior managers.6 The questionnaire for
general users contained the first four constructs whereas the questionnaire for

5
Throughout this research it has been assumed that respondents would have selected the
mid-point on the scale as the neutral point.
6
The final survey instruments are available from the researchers.

© The Authors
Journal compilation © 2007 AFAANZ
 S. W. Tee et al./Accounting and Finance 47 (2007) 335–355 343

Table 1
Instrument development

Description in model Source of questions Construct name a

Data quality Wang and Strong (1996), Data quality (3 questions)

Kahn et al. (2002), as
modified by researchers
Management commitment to Saraph et al. (1989), Flynn et al. Data quality committment
data quality (1995), Wixom and Watson (5 questions)
(2001), as modified by
researchers
Data quality champions Wixom et al. (2001), Beath (1991), Data quality champions
Reich and Benbasat (2000), (3 questions)
as modified by researchers
Data quality-related extrinsic Saraph et al. (1989), Flynn Data quality REWARDS
rewards (extrinsic rewards) et al. (1995), as modified (3 questions)
by researchers
Perceived usefulness of data quality Powell (1995), Douglas and Competitive funding
as a strategic resource (perceived Judge (2001), as modified (4 questions)
usefulness of data quality for by researchers
competitive funding submissions)
Information systems/information Bharadwaj (2000), Reich and Information systems/
technology capabilities Benbasat (2000), as modified information technology
by researchers capability
Perceived need for data quality to SERVQUAL (Parasuraman et al., Operations and client
support products and services 1988, 1991), as modified by services (5 questions)
(perceived need for data quality researchers to take into account
to support operations and client just the perceived need
services)
Regulatory requirements Developed by researchers Regulatory requirements
(3 questions)
Contractual requirements Developed by researchers Funding agreement
(funding agreements) (4 questions)
Competitive pressures Developed by researchers Government priorities
(government priorities) (3 questions)

a
To obtain measures for each construct (except information systems/information technology capability),
participants were asked to mark their perceptions on a continuous scale of 0 (strongly disagree) to 1
(strongly agree).

senior managers contained all constructs. A total of 67 surveys were distributed

(Table 2 shows the total population, sample and responses). Of these, 14 were
sent to members of senior management and 53 surveys were sent to general users.
In total, 51 usable responses were received. Of surveys, 13 were from senior
management and 38 were from general users (76.1 per cent response rate).7

7
Of the 51 respondents, 20 are data producers, and the remaining 31 are data consumers.
Data producers capture, enter and process data. Data consumers use the data entered by the
data producers.

© The Authors
Journal compilation © 2007 AFAANZ
344 S. W. Tee et al./Accounting and Finance 47 (2007) 335–355

Table 2
Population, sample and responses

General users

Senior managers Support staff Professionals Total

Head office 15 90 150 255

Regional offices 0 44 38 82
Total population 15 134 188 337
Sample 14 41 12 67
Responses 13 33 5 51

3.3. Interviews with senior managers

After conducting the survey, seven interviews, ranging in length from 60 to

90 min, were conducted. The interviewees included the chief executive officer,
the managers of business units one and three, the second officer in charge of
business unit two, the business analyst, the database administrator, and the
senior administrative officer of business unit one. Before the interviews, the
researcher reviewed the interviewees’ demographic data collected during
the data quality survey to obtain greater knowledge about the experience and skills
of each interviewee. A set of open-ended questions were developed to assist in
the interviews. Interviewees were asked about issues ranging from data quality
awareness to benefits of data quality programmes. Questions were introduced
to elicit information to address the hypothesized relations and to gain more
insights about interesting issues. Interview transcripts were analysed using the
deductive analysis approach (Patton, 2002). The interviews provided insights
into factors that affect Zelda’s senior management’s commitment to improve the
quality of their data. The interviews with senior managers provided insights
about the nature and causes of the data quality issues experienced by Zelda.
Most errors currently experienced by Zelda were caused by staff not following
procedures correctly when processing applications. Zelda attempts to reduce
these errors by conducting data quality awareness programmes with all levels of
staff in the organization.

4. Results

4.1. Descriptive statistics

Of the 51 respondents, approximately 59 per cent of the general user

respondents have worked in the organization for more than 5 years. Responses
also indicated that general users rarely attend information systems/information
technology/data quality conferences, seminars, workshops or exhibitions

© The Authors
Journal compilation © 2007 AFAANZ
 S. W. Tee et al./Accounting and Finance 47 (2007) 335–355 345

Table 3
Construct descriptive statistics

Panel A: Descriptive statistics general users

Constructs (N = 38) Minimum Maximum Meana SD

Data quality 0.3918 0.9931 0.6828 0.1349

Data quality commitment 0.2742 0.9794 0.6107 0.1634
Data quality champions 0.4794 0.9639 0.7188 0.1307
Data quality rewards 0.0241 0.8179 0.3386 0.1833

Panel B: Descriptive statistics senior managers

Constructs (N = 13) Minimum Maximum Meana SD

Data quality 0.4089 0.9519 0.6964 0.1691

Data quality commitment 0.2948 0.9526 0.7156 0.2119
Data quality champions 0.5200 0.9725 0.7354 0.1517
Data quality rewards 0.1924 0.6900 0.3984 0.1251
Competitive funding 0.4900 0.9948 0.7078 0.1452
Operations and client services 0.3979 0.8680 0.6948 0.1380
Regulatory requirements 0.2749 0.7835 0.5507 0.1554
Funding agreement 0.0900 0.6500 0.4493 0.1741
Government priorities 0.4467 1.0000 0.7105 0.1889

a
On a 0 –1 point scale, participants were asked to mark their perceptions on a continuous scale of 0
(strongly disagree) to 1 (strongly agree). The participant’s score on each question was the ratio of the
marked distance (from 0 to x) to the total distance (from 0 to 1). SD, standard deviation.

(approximately one every 3 years). Approximately 50 per cent of the senior

manager respondents have worked in the organization for more than 10 years.
Members of senior management attend approximately two information systems/
information technology/data quality-related conferences, exhibitions, seminars
and workshops each year.
For each questionnaire, the scores for all the questions related to each construct
were averaged8 to compute the value of the construct.9 The questionnaires for
general users and senior managements were analysed separately. Panel A of
Table 3 presents descriptive statistics for the constructs measured through the
general users’ questionnaires. Panel B of Table 3 presents descriptive statistics
for the constructs measured through the senior managers’ questionnaires.

8
Because the number of responses was well below the desirable level, factor analysis was
not performed.
9
Taking into account the fact the some of the constructs were being measured by new
instruments and also the small number of observations, the Cronbach alpha scores indicate
that the constructs appear reasonable.

© The Authors
Journal compilation © 2007 AFAANZ
346 S. W. Tee et al./Accounting and Finance 47 (2007) 335–355

Table 4
Perceptions of relative importance of data quality dimension (100 points in total)

By staff level By data user type

Mean p-value Mean p-value

SM GU DP DC
Data quality dimension (N = 13) (N = 38) SM vs GU (N = 20) (N = 31) DP vs DC

Accuracy 49.85 43.53 0.242 36.89 49.41 0.011

Relevance 24.45 29.90 0.115 32.52 26.38 0.058
Timeliness 25.68 26.53 0.805 30.58 24.16 0.042

SM, senior management; GU, general user; DP, data producer; DC, data consumer.

These results show that general users perceive the data quality and data
quality commitment as moderately high and that Zelda has an effective data quality
champion. The low result for data quality rewards is primarily attributable to
the organizational setting (i.e. government-funded agencies can seldom provide
direct performance-based payments to employees). The results show that senior
managers rated all constructs except data quality rewards, regulatory requirements
and funding agreement as moderately high (> 0.6). The mediocre ratings for
regulatory requirements and funding agreement suggest that senior management
perceives that the benefits associated with complying with these requirements
provide little motivation for Zelda to improve data quality. Similar to the results
for general users, the results for senior management also indicated that data
quality rewards provide few incentives for data quality improvements.
The participants were asked to indicate their perceptions of the relative
importance of the accuracy, timeliness and relevance of the data they entered or
used. Table 4 summarizes the perceptions of the relative importance of each of
the data quality dimensions. There was no significant difference between senior
management and general users’ perceptions of the relative importance of the
three data quality dimensions. Both senior management and general users
perceived accuracy as approximately twice as important as relevance or timeliness.
When a similar analysis was undertaken from the perspective of data producers
versus data consumers, significant differences existed between the relative
importance of the three data quality dimensions. Data producers perceived
accuracy, timeliness and relevance as approximately equally important. Data
consumers perceived accuracy as twice as important as relevance or timeliness.
Data consumers also perceived accuracy as significantly more important than data
producers. However, data producers perceived both relevance and timeliness as
more important than data consumers. Data producers are likely to consider
timeliness, in particular, as more important than data consumers because the
organization constantly encourages data producers to enter their data on a

© The Authors
Journal compilation © 2007 AFAANZ
 S. W. Tee et al./Accounting and Finance 47 (2007) 335–355 347

Table 5
Pearson correlations of factors affecting data quality

Panel A: Pearson correlations of factors affecting data quality (general users)

N = 38 Constructs Correlations coefficients ( p-value)

Data quality Data quality commitment Data quality rewards

Data quality commitment 0.487 (0.001)

Data quality rewards 0.184 (0.141) 0.251 (0.073)
Data quality champions 0.387 (0.013) 0.480 (0.003) 0.151 (0.189)

Panel B: Pearson correlations of factors affecting data quality (senior management)

N = 13 Correlations coefficients ( p-value)

1 2 3 4 5 6 7 8 9

Data quality 1
Data quality 2 0.687
commitment (0.005)
Data quality 3 0.522 0.328
rewards (0.034) (0.137)
Data quality 4 0.119 0.581 –0.172
champions (0.350) (0.019) (0.287)
Competitive 5 –0.411 –0.078 –0.429 0.646
funding (0.081) (0.400) (0.072) (0.009)
IS/IT/DQ 6 0.000 –0.382 0.043 –0.052 0.255
experience (0.500) (0.099) (0.444) (0.335) (0.200)
Operations 7 0.507 0.516 0.077 0.419 0.138 0.136
and client (0.039) (0.036) (0.401) (0.077) (0.372) (0.329)
services
Regulatory 8 0.362 0.322 0.560 0.355 0.122 0.178 0.525
requirement (0.112) (0.141) (0.023) (0.117) (0.346) (0.281) (0.033)
Funding 9 0.211 0.076 0.523 0.034 0.154 – 0.032 0.261 0.639
agreement (0.244) (0.402) (0.033) (0.457) (0.308) (0.459) (0.195) (0.009)
Government 10 0.179 0.336 –0.444 0.576 0.589 0.341 0.630 0.040 – 0.169
priority (0.279) (0.131) (0.064) (0.020) (0.017) (0.127) (0.010) (0.448) (0.291)

IS/IT/DQ, information systems/information technology/data quality.

timely basis. This emphasis on timeliness occurs because the organization might
not receive recognition from funding agencies for activities that occurred before the
end of a reporting period but that were entered after the end of the reporting period.

4.2. Tests of hypotheses

The research hypotheses were tested using Pearson correlations. Panel A of

Table 5 reports the Pearson correlations between the constructs and the levels of

© The Authors
Journal compilation © 2007 AFAANZ
348 S. W. Tee et al./Accounting and Finance 47 (2007) 335–355

significance for general users. Panel B of Table 5 reports the Pearson correlations
between the constructs and the levels of significance for senior managers. Where
the responses provide support for a hypothesis that is not supported statistically,
responses from the interviewees that do support the hypothesis are discussed below.
Hypothesis 1 asserts that management commitment is positively associated
with data quality. The correlation coefficient between the two measures is
highly significant for both the general users (correlation coefficient of 0.487,
p = 0.001) and senior management (correlation coefficient of 0.687, p = 0.005).
Therefore, Hypothesis 1 is supported.
Hypothesis 2 states that the presence of one or more data quality champions
is positively associated with management commitment to data quality. The
correlation coefficient between the two measures is highly significant for the
general users (correlation coefficient of 0.480, p = 0.003) and significant for
senior management (correlation coefficient of 0.581, p = 0.019). Hence,
Hypothesis 2 is supported.
Hypothesis 3 maintains that the presence of extrinsic rewards is positively
associated with management commitment to data quality. The correlation
coefficient between the two measures for the general users was moderately
significant (correlation coefficient of 0.251, p = 0.073).10 The correlation coeffi-
cient between the two measures in the senior management was not statistically
significant (correlation coefficient of 0.328, p = 0.137). Given the low number
of senior management observations, the lack of results might simply be because
of the small sample size. Therefore, Hypothesis 3 is partially supported.
Hypothesis 4, which asserts that management’s perceptions of the usefulness
of data quality for competitive funding is positively associated with management
commitment to data quality, was not supported (correlation coefficient of
−0.078, p = 0.400). One possible reason for the non-significant relationship was
the low number of responses (13). Another possible reason is that Zelda is a
government-funded agency that operates in a non-competitive environment. From
the interviews, Zelda’s management perceives that high-quality data enhance
Zelda’s reputation and improve its relationships with its funding agencies. They
perceive that high-quality data help Zelda to negotiate better future funding
arrangements. The following statement by the chief executive officer provides
qualitative support for Hypothesis 4:

Both Commonwealth and the State Government officials have high regard for the quality
of the information they are getting from Zelda. Having a reputation for accurate data has
positive effects and helps Zelda obtain more of the funding we request. (SM1)

Hypothesis 5, which asserts that the organization’s information systems/

information technology capabilities are positively associated with management’s

10
Although this result is not significant at p ≤ 0.05, it can be considered moderately signi-
ficant in light of the small sample size of 38.

© The Authors
Journal compilation © 2007 AFAANZ
 S. W. Tee et al./Accounting and Finance 47 (2007) 335–355 349

perceptions about the usefulness of data quality for competitive funding, was
not supported (correlation coefficient of 0.255, p = 0.200). Two possible
explanations for the non-significant relationship are a weakness in the survey
questions intended to measure the construct11 and the limited number of obser-
vations available for analysis. Based on interview responses, Zelda’s ability to
recognize the potential benefits of high-quality data helped motivate improve-
ments in the organization’s information systems/information technology
capabilities. The increased use of technology innovations allowed Zelda to
implement online electronic lodgement of applications. Improvements in
technology have facilitated the implementation of a sophisticated accounting
system that more accurately tracks Zelda’s budget. The new accounting system
enables the staff to conduct more sophisticated data analyses. One senior
manager asserted that better information systems/information technology
capabilities help link business and data together:

Information technology enables the organization to allow people to play with the data and
to understand it. They begin to realize that it is not just a piece of paper with graphs on
it, but something that when you look at it, you can feel what is right and what is wrong.
More importantly, people now know how to use it and benefit from it. (SM2)

This response provides qualitative support for Hypothesis 5.

Hypothesis 6 asserts that management’s perception of the need for high-
quality data to support the organization’s operations and clients’ services is
positively associated with management commitment to data quality. The
analysis supported this hypothesis (correlation coefficient of 0.516, p = 0.036).
Hypothesis 7 asserts that regulatory requirements influence organizations’
perceptions of the need for high-quality data to support their operations and
client services. The analysis supported this hypothesis (correlation coefficient
of 0.525, p = 0.033).
Hypothesis 8, which asserts that funding agreement requirements influence
management’s perception of the need for data quality to support operations and
client services, was not supported (correlation coefficient of 0.261, p = 0.195).
One possible explanation for the lack of support was that Zelda currently has a
4 year funding agreement with the funding agencies. Therefore, senior manage-
ment might not perceive an immediate need for high-quality data to support
operations and client services and funding agreements. From the interview data,
the overwhelming majority of Zelda’s operating funds come from the Common-
wealth and State Governments. These funds are provided to Zelda through
funding agreements that normally last 3–4 years. The funding agreements
require Zelda to provide accurate data on the services it provides and the clients

11
Information systems/information technology/data quality experience was used as proxy
for information systems/information technology capabilities. This proxy construct might
not have measured information systems/information technology capabilities effectively.

© The Authors
Journal compilation © 2007 AFAANZ
350 S. W. Tee et al./Accounting and Finance 47 (2007) 335–355

it serves. The frequency of submissions and the types of data Zelda needs to
provide are specified in the agreement. As the database administrator recalled:
We (Zelda) developed Vision to satisfy the requirements agreed upon between Zelda and
the Commonwealth. The agreement specifies what data Zelda (and other similar service
providers) should collect and submit. (SM6)

These statements provide qualitative support for Hypothesis 8.

Hypothesis 9 asserts that government priorities influence management’s
perception of the need for data quality to support products and services. The
analysis supported this hypothesis (correlation coefficient of 0.630, p = 0.010).

4.3. Ex post analyses

Further analyses were conducted to examine other relationships between

the various constructs. Results indicated a moderate correlation between data
quality rewards and data quality (correlation coefficient of 0.522, p = 0.034).
This relationship suggested that senior management perceived data quality rewards
to directly influence data quality.
Results also show moderate correlations between operations and client services
and data quality (correlation coefficient of 0.507, p = 0.039). Zelda is committed
to deliver high-quality, effective and efficient services through technology and
innovation. More recently, senior management perceived a need for high-
quality data to support their operations and client services. This relationship
suggested a link between operations and client services and data quality.
Ex post analysis was also conducted to examine the relationships within the
research model from the perspectives of data consumers versus data producers.
The majority of the senior managers were, as expected, data consumers (10 out
of 13). Approximately half the general users were data consumers (21 out of
38: 55 per cent). Analyses were conducted to examine correlations between the
constructs for data producers and data consumers. The most interesting finding
was that extrinsic rewards are associated with management’s commitment to
data quality for data consumers (correlation coefficient of 0.682, p = 0.001)
but not data producers (correlation coefficient of −0.278, p = 0.118). When
compared with the results for both senior managers and general users this result
suggests that the relationship between extrinsic rewards and management’s
commitment to data quality for data consumers is important for general users
who are data consumers.

5. Conclusions, limitations and future research

This research was motivated by empirical and anecdotal evidence about

the impacts of poor data quality on organizations’ information systems. Data
quality researchers have developed data quality frameworks to organize and
structure data quality dimensions. Organizations can use these frameworks to

© The Authors
Journal compilation © 2007 AFAANZ
 S. W. Tee et al./Accounting and Finance 47 (2007) 335–355 351

Table 6
Summary results of the testing of hypotheses in the model

Statistical Interview
Hypothesis Description test results responses

H1 Management commitment to data quality is positively Supported Supported

associated with the level of data quality achieved.
H2 The presence of a champion is positively associated Supported Supported
with management’s commitment to data quality.
H3 Extrinsic rewards are positively associated with Partially Insufficient
management’s commitment to data quality. supported evidence
H4 The perceived usefulness of data quality for competitive Not supported Supported
funding is positively associated with management’s
commitment to data quality.
H5 Information systems/information technology capabilities Not supported Supported
are positively associated with the perceived usefulness
of data quality for competitive funding.
H6 The perceived need for data quality to support operations Supported Supported
and client services is positively associated with
management’s commitment to data quality.
H7 The need to comply with regulatory requirements is Supported Insufficient
positively associated with the perceived need for data evidence
quality support operations and client services
H8 The need to comply with funding agreements is Not supported Supported
positively associated with the perceived need for data
quality support operations and client services
H9 Meeting government priorities is positively associated Supported Supported
with the perceived need for data quality to support
operations and client services

understand data quality. Nevertheless, organizations continue to experience

problems with data quality. The purpose of this research was to identify a number
of factors and test whether they affect the data quality within an organization.
The relations were tested using data collected from a data quality survey and
interviews with senior managers at a government-funded service organization.
Table 6 summarizes the results relative to each of the nine hypotheses. It
indicates that Hypotheses 1 and 2 were supported by both data collection
methods. Hypothesis 3 was supported by the survey results, but not by the inter-
views. Hypotheses 4 and 8 were supported only by the interviews. Hypotheses 6
and 9 were supported by both data collection methods. Hypothesis 7 was
supported only by the survey results.
This research validated the assertions that management responsibilities,
including commitment to continuously improving data quality, effective
communication among stakeholders, and data quality awareness are important
organizational elements that influence data quality. Data management researchers
can use this research to refine existing data quality theory and models. They can

© The Authors
Journal compilation © 2007 AFAANZ
352 S. W. Tee et al./Accounting and Finance 47 (2007) 335–355

also use the results of this research to refine existing data management policies
or to develop new policies.
The research benefits organizations in several ways. First, managers will be
better able to identify critical factors for successfully implementing new data
quality initiatives and for nurturing existing data quality activities. Second,
managers will be better able to understand the relationships among these critical
success factors. Third, they can use their improved understanding to develop or
enhance their organizational data quality policies and initiatives.
The usual caveats associated with surveys and interview-based research
apply here. Within these caveats the most significant limitation is the small
sample size upon which to base the statistical conclusions. The small sample
did not allow more advanced techniques such as factor analysis or structural
equation modelling to be used. With data being collected from only one case
study firm, the additional limitation of generalizability of results also applies.
Furthermore, the researchers relied upon the business analyst within the case
study firm to help with the identification of the groups for receiving the
questionnaire and also for the interviews. Finally, the possibility of measure-
ment issues needs to be raised. For example, the constructs were modified from
a number of existing instruments and the measurement scales were treated as
continuous rather than ordinal.
The results of this study suggest several areas for future research. First, this
study should be replicated with organizations from different industry sectors
and with organizations that operate in environments exhibiting different levels of
competitiveness. Such replication would allow for firm specific attributes to be
incorporated into the research. Furthermore, an expanded dataset would offer the
opportunity to examine the model using additional statistical techniques such as
structural equation modelling. Second, future studies could examine the impact
of product types (information products vs traditional products) and business envir-
onments to further develop the framework and to obtain a deeper understanding
of the need for high-quality data. Third, future research can be undertaken to refine
the researcher developed constructs and to improve the theoretical basis for
examining the association between these constructs and the usefulness of data
quality for competitive advantage. This research could also take into consider-
ation issues in relation to the actual level of data quality, ideal levels of data
quality, and acceptable levels of data quality. Finally, future research could
refine the proposed set of relations into a model that can be tested empirically.

References

Anderson, J. C., M. Rungtusanthanam, R. Schroeder, and S. Devaraj, 1995, A path analytic

model of a theory of quality management underlying the Deming management method:
preliminary empirical findings, Decision Sciences 19, 473–509.
Anderson, F. R., 2002, Data Quality Act, National Law Journal 25, B9.
Ballou, D. P., and H. L. Pazer, 1995, Designing information systems to optimize the
accuracy-timeliness tradeoff, Information Systems Research 6, 51–72.

© The Authors
Journal compilation © 2007 AFAANZ
 S. W. Tee et al./Accounting and Finance 47 (2007) 335–355 353

Ballou, D. P., R. Y. Wang, H. Pazer, and G. K. Tayi, 1998, Modelling information manu-
facturing systems to determine information product quality, Management Science 44,
463–485.
Ballou, D. P., S. Madnick, and R. Wang, 2004, Assuring information quality, Journal of
Management Information Systems 20, 9–11.
Barney, J. B., 1991, Firm resource and sustained competitive advantage, Journal of Man-
agement 17, 99 –120.
Beath, C. M., 1991, Supporting the information technology champion, MIS Quarterly 15,
355–371.
Belluck, P., 1997, The tangled trail that led to a beef recall, New York Times, 24 August, p. 1.
Bharadwaj, A., 2000, A resource-based perspective on information technology capability
and firm performance: an empirical investigation, MIS Quarterly 24, 169–196.
Black, S. A., and L. J. Porter, 1996, Identification of critical success factors of TQM, Deci-
sion Sciences 27, 1–21.
Cappiello, C., C. Francalanci, and B. Pernici, 2004, Time-related factors of data quality in
multichannel information systems, Journal of Management Information Systems 20, 71–
92.
Copeland, D. G., and J. L. McKenney, 1988, Airline reservation systems: lessons from his-
tory, MIS Quarterly 12, 353–370.
Douglas, T. J., and W. Q. Judge, 2001, Total quality management implementation and com-
petitive advantage: the role of structural control and exploration, Academy of Manage-
ment Journal 44, 158–170.
English, L. P., 1999, Improving Data Warehouse and Business Information Quality: Methods
for Reducing Costs and Increasing Profits (Wiley, New York).
Flynn, B. B., R. G. Schroeder, and S. Sakakibara, 1995, The impact of quality management
practices on performance and competitive advantage, Decision Sciences 26, 659–692.
Fox, C., A. V. Levitin, and T. C. Redman, 1994, The notion of data and its quality dimen-
sions, Information Processing and Management 30, 9 –19.
Giannoccaro, A., G. Shanks, and P. Darke, 1999, Stakeholder perceptions of data quality in
a data warehouse environment, Australian Computer Journal 31, 110 –117.
Gibbs, M. R., G. Shanks, R. Lederman, and R. De Silva, 2002, Privacy and customer data
quality: exploring the issues, paper presented at the 13th Australasian Conference on Infor-
mation Systems, Melbourne, Vic., 4–6 December.
Grant, R. M., 1991, The resource-based theory of competitive advantage, California Man-
agement Review 33, 114 –135.
Grant, R. M., 1995, Contemporary Strategy Analysis (Blackwell Publishers, Oxford, UK).
Guimaraes, T., and M. Igbaria, 1997, Client/server system success: exploring the human
side, Decision Sciences 28, 851–876.
Heng, M. S., E. M. Trauth, and S. J. Fisher, 1999, Organisational champions of IT innova-
tion, Accounting Management and Information Technology 9, 193–222.
Howell, J. M., and C. A. Higgins, 1990, Champions of technological innovations, Adminis-
trative Science Quarterly 35, 317–341.
Huang, K. T., T. W. Lee, and R. Y. Wang, 1999, Quality Information and Knowledge
(Prentice Hall, Upper Saddle River, NJ).
Huh, Y. U., F. R. Keller, T. C. Redman, and A. R. Watkins, 1990, Data quality, Information
and Software Technology 32, 559–565.
Ingram, T. N., and D. N. Bellenger, 1983, Personal and organizational variables: their rela-
tive effects on reward valences of industrial salespeople, Journal of Marketing Research
20, 198–205.
Jiang, J. J., W. A. Muhanna, and G. Klein, 2000, User resistance and strategies for promot-
ing acceptance across system types, Information and Management 37, 25 –36.
Kahn, B. K., D. M. Strong, and R. Y. Wang, 2002, Information quality benchmarks: prod-
uct and services performance, Communications of the ACM 45, 184 –192.

© The Authors
Journal compilation © 2007 AFAANZ
354 S. W. Tee et al./Accounting and Finance 47 (2007) 335–355

Klein, B. D., D. L. Goodhue, and G. B. Davis, 1997, Can humans detect errors in data?
Impact of base rates, incentives, and goals, MIS Quarterly 21, 169 –195.
Lee, Y. W., and D. M. Strong, 2004, Knowing-why about data processes and data quality,
Journal of Management Information Systems 20, 13– 40.
Levitin, A. V., and T. C. Redman, 1998, Data as a resources: properties, implications, and
prescription, Sloan Management Review 40, 89–101.
Nambisan, S., R. Agarwal, and M. Tanniru, 1999, Organizational mechanisms for enhanc-
ing user innovation in information technology, MIS Quarterly 23, 365–395.
Oz, E., and J. J. Sosik, 2000, Why information systems projects are abandoned: a leadership
and communication theory and exploratory study, Journal of Computer Information Sys-
tems 41, 66–78.
Paradice, D. B., and W. L. Fuerst, 1991, An MIS data quality methodology based on opti-
mal error detection, Journal of Information Systems 5, 48–66.
Parasuraman, A., L. L. Berry, and V. A. Zeithaml, 1988, SERVQUAL: a multiple-item
scale for measuring consumer perceptions of service quality, Journal of Retailing 64, 12–
40.
Parasuraman, A., L. L. Berry, and V. A. Zeithaml, 1991, Refinement and reassessment of
the SERVQUAL scale, Journal of Retailing 67, 420 – 450.
Patton, M. Q., 2002, Qualitative Research and Evaluation Methods, 3rd edn (Sage Publica-
tions, Thousand Oaks, CA).
Poon, P., and C. Wagner, 2001, Critical success factors revisited: success and failure cases
of information systems for senior executives, Decision Support Systems 30, 393–
418.
Powell, T., 1995, Total quality management as competitive advantage: a review and empir-
ical study, Strategic Management Journal 16, 15–37.
Redman, T. C., 1995, Improve data quality for competitive advantage, Sloan Management
Review 36, 99–107.
Redman, T. C., 1996, Data Quality for the Information Age (Artech House, Norwood,
MA).
Reich, B. H., and I. Benbasat, 2000, Factors that influence the social dimension of
alignment between business and information technology objectives, MIS Quarterly 24,
81–113.
Salancik, G. R., 1997, Commitment and control of organizational behavior and belief, in: B.
Staw and G. R. Salancik, eds, New Directions in Organizational Behavior (St Clair Press,
Chicago, IL), 1–54.
Saraph, J. V., P. G. Bensen, and R. G. Schroeder, 1989, An instrument for measuring the
critical factors of quality management, Decision Sciences 20, 810 –829.
Sarin, S., and V. Mahajan, 2000, The effects of reward structures on the performance of
cross-functional products development team, Journal of Marketing 65, 35–53.
TDWI, 2002, Data quality and the bottom line: achieving business success through a com-
mitment to high quality data, TDWI Report Series [online]. Available: http://www.dw-
institute.com/research/display.asp?id=6064
Tubbs, M. E., 1993, Commitment as a moderator of the goal-performance relation: a case
for clearer construct definition, Journal of Applied Psychology 78, 86 –97.
Wand, Y., and R. Y. Wang, 1996, Anchoring data quality dimensions in ontological founda-
tions, Communications of the ACM 39, 86–95.
Wang, R. Y., 1998, A product perspective on total data quality management, Communica-
tions of the ACM 40, 58–65.
Wang, R. Y., and H. B. Kon, 1993, Toward total data quality management (TDQM), in:
R. Y. Wang, ed., Information Technology in Action: Trends and Perspectives (Prentice
Hall, Englewood Cliffs, NJ), xx–xx.
Wang, R. Y., V. C. Storey, and C. P. Firth, 1995, A framework for analysis of quality
research, IEEE Transactions on Knowledge and Data Engineering 7, 623–640.

© The Authors
Journal compilation © 2007 AFAANZ
 S. W. Tee et al./Accounting and Finance 47 (2007) 335–355 355

Wang, R. Y., and D. M. Strong, 1996, Beyond accuracy: what data quality means to data
consumers, Journal of Management Information Systems 12, 5–33.
Whalen, C., 2001, An explorer may lead Ford to rain, Insights on the News, 20 August, vol. 17,
issue 31, 16 – 20.
Wixom, B. H., and H. J. Watson, 2001, An empirical investigation of the factors affecting
data warehousing success, MIS Quarterly 25, 16–41.

© The Authors
Journal compilation © 2007 AFAANZ