Content uploaded by Davood Askarany
Author content
All content in this area was uploaded by Davood Askarany on Jun 13, 2019
Content may be subject to copyright.
Int. J. Managerial and Financial Accounting, Vol. 4, No. 3, 2012 291
Copyright © 2012 Inderscience Enterprises Ltd.
The effect of innovation characteristics on
activity-based costing adoption
Davood Askarany
Business School,
University of Auckland,
Private Bag 92019, Auckland, New Zealand
E-mail: d.askarany@auckland.ac.nz
John A. Brierley*
Management School,
University of Sheffield,
9 Mappin Street, Sheffield, S1 4DT, UK
Fax: 0114 (0)114-222-3348
E-mail: j.a.brierley@shef.ac.uk
*Corresponding author
Hassan Yazdifar
Accounting and Finance,
Business School,
University of Glasgow,
Glasgow, G12 8QQ, UK
E-mail: hassan.yazdifar@glasgow.ac.uk
Abstract: This paper contributes to the analysis of the factors influencing the
adoption of ABC by assessing the contribution of the characteristics of an
innovation on adoption, which has not been carried out in prior research.
Specifically, the paper applies innovation diffusion theory to examine the
impact of five characteristics of an innovation, and organisation size, industry
and location on the decision to adopt activity-based costing (ABC). The data
for the study was obtained from a questionnaire survey of management
accountants in Australia, New Zealand and the UK. The relationships with
ABC adopted were tested using logistic regression analysis. The best model
specification arises when organisations that have adopted ABC are compared
with those that have rejected it. The results reveal that organisations are more
likely to adopt ABC when they attach a high level of importance to the relative
advantages offered by innovations, are large and located in Australasia.
Keywords: innovation diffusion theory; IDT; characteristics of an innovation;
activity-based costing; ABC; questionnaire survey.
Reference to this paper should be made as follows: Askarany, D.,
Brierley, J.A. and Yazdifar, H. (2012) ‘The effect of innovation characteristics
on activity-based costing adoption’, Int. J. Managerial and Financial
Accounting, Vol. 4, No. 3, pp.291–313.
292 D. Askarany et al.
Biographical notes: Davood Askarany is a Senior Lecturer in Accounting and
Finance at the University of Auckland Business School. His research interest is
in the area of the adoption of accounting innovations.
John A. Brierley is a Senior Lecturer in Accounting and Finance at the
University of Sheffield Management School. His research interest is in the
calculation of products costs and their use in decision-making.
Hassan Yazdifar is a Senior Lecturer in Accounting and Finance at the
University of Glasgow Business School. His research interest is in the area of
management accounting and changing management accounting practises.
1 Introduction
It has been suggested that cost accounting information is too late, too aggregated, and too
distorted to be relevant for management’s needs (e.g., Kaplan, 1985; Johnson and
Kaplan, 1987; Kaplan and Cooper, 1998). This is because the system uses a small
number of volume-based cost drivers to assign overhead costs to product costs, rather
than using non-volume-based cost drivers to assign overhead costs based on the activities
that drive those costs (Kaplan and Cooper, 1998). One management accounting
innovation that has been proposed to overcome this problem is activity-based costing
(ABC), whereby overhead costs are assigned to products using non-volume-based cost
drivers (e.g., Cooper and Kaplan, 1987, 1988, 1991, 1999; Cooper, 1988a, 1988b, 1990,
1991, 1997; Cooper et al., 1992; Kaplan and Cooper, 1998). Notwithstanding ABC’s
purported benefits relative to traditional cost accounting techniques, its adoption rates
have been low (Gosselin, 1997). Research has found that ABC adoption rates as a
percentage of organisations actually using ABC relative to all organisations have often
been less than 30% (e.g., Drury et al., 1993; Innes and Mitchell, 1995; Lukka and
Granlund, 1996; Booth and Giacobbe, 1998; Clarke et al., 1999; Groot, 1999; Drury and
Tayles, 2000; Innes et al., 2000; Lamminmaki and Drury, 2001; Cotton et al., 2003;
Brierley, 2011). A problem when reviewing prior ABC research, however, is that
researchers use different definitions of the adoption of ABC. Although some researchers
have defined adoption in terms of those that are actually using ABC (see above), others
have defined it as using or implementing ABC (Armitage and Nicholson, 1993; Cohen
et al., 2005), implemented, implementing or planning to implement ABC (Bjørnenak,
1997), using, implementing, approved for implementing or implemented and abandoned
(Krumwiede, 1998) and using, implementing or intending to implement ABC (Al-Omiri
and Drury, 2007).1 However, from the point of view of a supplier of an innovation, an
innovation can only be considered to be successful when it has been accepted and
integrated into an organisation and an adopter has shown a commitment to use an
innovation in the future (Bhattacherjee, 1998). In other words, an innovation, like
ABC [for examples of other management accounting innovations, see Pollanen and
Abdel-Maksoud (2010)], can only be considered to have been adopted when an
organisation is using it permanently. Consequently, some of the other definitions of ABC
adoption mentioned above may not be appropriate. In other words, an innovation would
not be deemed to have been adopted permanently when organisations are, for example,
implementing, intending to implement, or implemented and abandoned an innovation.
The effect of innovation characteristics on activity-based costing adoption 293
The appropriateness of this definition of ABC adoption has been confirmed in recent
research (Brierley, 2011).
Given that an innovation like ABC can only be regarded as being adopted when it is
being used, some studies have investigated the impact of a number of influences upon the
adoption of ABC (e.g., Brown et al., 2004) and others have identified the differences
between organisations that have adopted and not adopted ABC (e.g., Clarke et al., 1999;
Groot, 1999; Malmi, 1999). These researchers have compared the percentage of
organisations sampled that have adopted and are using ABC with other organisations in
two distinct ways. These are the percentage of organisations that are using ABC relative
to all other organisations that are not using ABC (Clarke et al., 1999; Malmi, 1999),
referred to, hereafter, as the version 1 definition of ABC adoption, and organisations that
are using ABC relative to those organisations that have rejected ABC (Groot, 1999;
Brown et al., 2004), referred to hereafter as the version 2 definition of ABC adoption. A
possible problem with the version 1 definition is that it assumes that those organisations
that have not adopted ABC are a homogeneous group. However, these non-adopters are
made up of organisations that have considered and rejected ABC, are currently
considering ABC (which can range from those that are conducting preliminary
investigations into ABC to those that have implemented ABC on a pilot basis) and those
that have not considered ABC. Consequently, these organisations may not be a
homogeneous group. For example, Krumwiede (1998) suggests that organisations that
reject ABC after having considered it usually have better quality information systems (IS)
than organisations that have not considered it, and this has been confirmed by Brierley
(2011). Given this, it may be more appropriate to use the version 2 definition of ABC
adoption, namely whether organisations that are using ABC compared with those that
have rejected it, and this is because they both have in common the fact that they have
considered ABC.
When examining the factors influencing the adoption of innovations, innovation
diffusion theory (IDT) suggests that the characteristics of an innovation, such as relative
advantage, compatibility, complexity, observability and trialability, are crucial to their
adoption (Rogers, 2003). To the authors’ knowledge, prior research has not examined the
association between the perceived characteristics of an innovation and the actual adoption
of ABC. This is an important omission from prior ABC research and needs to be included
for a number of reasons. First, these characteristics have been included extensively in
models in the IS literature (Agarwal and Prasad, 1997), but have not been included in
prior ABC research. Second, Rogers’ (2003) states that between 49% and 87% of the
variance in the rate of adoption of innovations is accounted for by the perceived
characteristics of an innovation. Given this, there is a need to conduct research in this
area to investigate if a large proportion of the variance in the decision to adopt ABC is
accounted for by these characteristics. Third, although the perceptions of the relative
advantage of ABC has been included in prior ABC research, it is important to test the
impact of other characteristics of an innovation postulated in IDT. Fourth, the
identification of these characteristics in the adoption decision is important because it
means that it is possible to target adoption efforts appropriately and from a managerial
perspective, management can practicably implement these efforts (Wynekoop et al.,
1992). Fifth, understanding the determinants of adoption behaviour is important because
the success of the impact of the innovation is dependent on the innovation being adopted
(Agarwal and Prasad, 1997).
294 D. Askarany et al.
Given the research issues discussed above, the objective of this paper is to use the
results of a questionnaire survey of management accountants in Australia, New Zealand
and the UK to investigate the extent to which the adoption rate for ABC, when defined
using both the version 1 and version 2 definitions of ABC adoption, is influenced by the
characteristics of an innovation. In addition, the impact on adoption of the size and
industry of the organisation, are also included in the research model because they have
been included in prior ABC adoption research. In addition, as the research is undertaken
in three countries, the country of the respondent is included as a control variable.
The remainder of the paper is organised in the following way. Section 2 discusses
IDT in more detail. Section 3 develops a series of hypotheses for testing and incorporates
them into the research model. Section 4 discusses the research method. Section 5 presents
and discusses the research results. Section 6 concludes the research, identifies its
limitations and offers some opportunities for future research.
2 Innovation diffusion theory
According to IDT, an innovation can be an idea, object or practise that is regarded as new
by an individual or other body (Rogers, 2003). The diffusion of an innovation can be
explained as a process by which an innovation is communicated through certain channels
over time among the members of a social system, and its adoption refers to the decision
of an organisation or individual to use an innovation (Rogers, 2003). Rogers (1983)
identified the characteristics of an innovation as having a major role in influencing the
adoption, which he suggests can be described by five characteristics. These are its
relative advantage, compatibility, complexity, observability and trialability. Moore and
Benbasat (1991) expanded these five characteristics into eight characteristics made up of
relative advantage, compatibility, ease of use, result demonstrability, trailability,
visibility, image and voluntariness of use. Each of these eight characteristics is discussed
below in more detail.
The relative advantage of an innovation can be explained as the degree to which an
innovation seems to be better than the method it is supposed to replace. Compatibility
refers to the degree of consistency of an innovation with the needs, expected values and
the norms of potential adopters and their social systems (Rogers, 2003). Ease of use is the
extent to which an innovation can be used with little effort (Davis et al., 1989). Result
demonstrability relates to the tangibility of using the innovation, and includes the
observability and communicability of the results of using an innovation. Trialability is the
degree to which an innovation can be tried on a limited basis before full implementation.
This reduces the risk and uncertainty of the consequences of using the innovation
(Karahanna et al., 1999). The visibility of an innovation relates to the extent to which one
can physically see an innovation. Image is the degree to which using an innovation
enhances the user’s image or status and voluntariness of use is the extent to which the use
of an innovation is regarded as voluntary.
The effect of innovation characteristics on activity-based costing adoption 295
3 Hypothesis development and research model development
3.1 Characteristics of innovations
In prior non-ABC research, the decision to adopt/use an innovation has been found to be
related positively to the perception of the level of:
• The relative advantage of an innovation relative to the method it is intended to
replace (e.g., Lai, 1997; Jebeile and Reeve, 2003). Although a non-significant effect
has been observed by Grover (1993), it is expected that an innovation is more likely
to be adopted if it has a large number of advantages relative to the current and other
methods.
• The compatibility of an innovation with the needs, expected values and norms of its
potential adopters because for an innovation to be adopted it would be expected to
meet their expectations (e.g., Grover, 1993; Lai, 1997; Jebeile and Reeve, 2003).
• The ease of use of an innovation for its adopters because if an innovation is to be
adopted it should be easy to use (e.g., Grover, 1993; Lai, 1997; Jebeile and
Reeve, 2003).2
• The result demonstrability of an innovation, in terms of the observability and
communicability of using it. In other words for an innovation to be adopted the
advantages from using it should be apparent (e.g., Jebeile and Reeve, 2003).
• The trialability of an innovation prior to its full adoption, which means that it should
be possible to undertake a pilot of the innovation prior to making the final decision to
adopt it (e.g., Jebeile and Reeve, 2003).
Although prior ABC research has not considered the collective influence of these five
characteristics together on the adoption of ABC, Anderson’s (1995) case study of
General Motors found a positive relationship between the relative advantage and the ease
of use of ABC and its adoption. While Brown et al. (2004) did not find a significant
association between relative advantage and the version 2 definition of ABC adoption. In
addition, the measures of the perceptions of visibility, image and voluntariness of use
have not been tested in prior ABC research and would not be expected to influence the
adoption of an innovation like ABC, which is regarded as having the intangible-like
quality of an administrative process innovation (Swanson, 1994).
Given the results of prior non-ABC research mentioned above, it is possible to test
the following hypotheses relating to the influence of each of the five characteristics of an
innovation using Moore and Benbasat’s (1991) measure of the adoption of an innovation.
This is important given that prior ABC research has not considered the impact of the
compatibility, result demonstrability and trialability, which is expected to be a positive
impact, on ABC adoption.
H1a The perception of the relative advantage of an innovation is related positively to
the adoption of ABC.
H1b The perception of the compatibility of an innovation is related positively to the
adoption of ABC.
296 D. Askarany et al.
H1c The perception of the ease of use of an innovation is related positively to the
adoption of ABC.
H1d The perception of the result demonstrability of an innovation is related positively
to the adoption of ABC.
H1e The perception of the trailability of an innovation is related positively to the
adoption of ABC.
3.2 Organisation size
Organisation size has been found to influence the likelihood of adopting an innovation
(Frambach and Schillewaert, 2002; Abdel-Maksoud, 2011). This is because larger
organisations have the infrastructure, contacts and communication channels needed to
adopt innovations (Bjørnenak, 1997) and have a higher need to adopt innovations in
order to maintain or improve their performance (Frambach and Schillewaert, 2002).
In prior ABC research, larger organisations have been found to use ABC, when using
the version 1 definition of adoption, when size has been measured by the number of
employees (Groot, 1999) and annual sales revenue (Innes and Mitchell, 1995; Clarke
et al., 1999; Innes et al., 2000). In contrast, other studies have found no association
between size and the version 1 definition of adoption, when size has been measured by
the number of employees (Drury et al., 1993; Brown et al., 2004).3 Hence, there is a need
for further research to attempt to reconcile these conflicting results. However, to the
authors’ knowledge, researchers have not tested the relationship between size and the
version 2 definition of ABC adoption; hence, this provides a reason for testing for the
existence of such a relationship in this research. It has been suggested that organisational
size is related positively to ABC adoption because larger organisations have the resources
necessary to invest in ABC systems (Booth and Giacobbe, 1998; Drury and Tayles, 2000;
Al-Omiri and Drury, 2007). Given the above, the following hypothesis is tested.
H2 The size of the organisation is related positively to the adoption of ABC.
3.3 Manufacturing and non-manufacturing industry
Although ABC was developed initially in the context of manufacturing organisations, has
been suggested subsequently that ABC systems are suitable in non-manufacturing
organisations because of their higher level of fixed and indirect overhead costs relative to
total costs that need to be assigned to cost objects (Kaplan and Cooper, 1998). In contrast,
in manufacturing organisations these overhead costs make up a smaller proportion of
total costs because of the high level of direct material costs. Consequently, the use of
traditional costing systems in manufacturing industry may report quite accurate costs
(Al-Omiri and Drury, 2007). To the authors’ knowledge, prior research has not tested
whether ABC adoption is higher in non-manufacturing industry for the version 2
definition of ABC adoption. Hence, this research corrects for this omission from prior
research. For the version 1 definition, Innes and Mitchell (1995) and Innes et al. (2000)
have found no significant differences between adopters and non-adopters. While Cagwin
and Bouwman (2002) found that manufacturing organisations were more likely to adopt
ABC.4 Given the conflicting evidence, it is necessary to conduct further research in this
area to test the following hypothesis that is based upon Kaplan and Cooper (1998):
The effect of innovation characteristics on activity-based costing adoption 297
H3 Organisations in non-manufacturing industry rather than manufacturing industry are
more likely to adopt ABC.
3.4 ABC in different countries
Although research has examined the adoption rates for ABC in a variety of countries,
little research has compared the adoption rates for ABC between countries in a single
study. Some exceptions include Yakhou and Dorweiler (1995) and Bhimani et al. (2007).
Notwithstanding their small sample sizes, Yakhou and Dorweiler (1995) found high ABC
adoption levels in England, France and the USA. Bhimani et al. (2007) report that the
adoption levels for ABC are higher in Canada, France, the UK and the USA, than in
Germany, Italy and Japan. They suggest that this is because ABC has received
considerable publicity in the late 1980s in Canada, the USA and the UK, and
organisations in France had already been using a method similar to ABC. In addition,
there have been replications of research undertaken in the UK by Drury et al. (1993) and
Innes et al. (2000) in New Zealand by Lamminmaki and Drury (2001) and Cotton et al.
(2003) respectively. The results of this research show that the rates of adoption for ABC
using the version 1 definition are similar between New Zealand and the UK. This
research examines the adoption of ABC in three Western and English-speaking counties
(namely Australia, New Zealand and the UK) as a control variable in the research and,
based on the results of prior research discussed above, it is anticipated that these
countries will have similar adoption levels for ABC. Hence:
H4 The location of the adopting organisation, namely Australia, New Zealand or the
UK, is not related to the adoption of ABC.
3.5 The research model
Given that the dependent construct is whether ABC has been adopted, the research model
is in the form of a binary logistic regression (hereafter logistic regression) model. Given
the hypotheses discussed above, the model can be written in the form of the log of the
odds of ABC being adopted, that is:
()
01 2
34
56
78
9
()
log ( ) ( )
()
()( )
()()
()( 1)
(2)
p ABC adopted BBRELADBCOMPAT
p ABC not adopted
B EASE B RESDEM
B TRIAL B SIZE
BNONMAN BCOUNTRY
BCOUNTRY
=+ +
++
++
++
+
(1)
or as the odds of ABC being adopted, that is:
0 1( ) 9( 2)
()
()
B
B RELAD B COUNTRY
p ABC adopted e
p ABC not adopted
+++
=… (2)
01( ) 9( 2)B B RELAD B COUNTRY
ee e= (3)
where
RELAD relative advantage of an innovation
298 D. Askarany et al.
COMPAT compatibility of an innovation
EASE ease of use of an innovation
RESDEM result demonstrability of an innovation
TRIAL trialability of an innovation
SIZE the natural logarithm of the approximate number of employees employed
by an organisation
NONMAN Dummy construct, whether or not the organisation is in
non-manufacturing or manufacturing industry,
coded 1 = non-manufacturing industry, 0 = manufacturing industry.
COUNTRY1 Dummy construct, coded 1/2 = Australia, 1/2 = New Zealand, –1 = UK.
COUNTRY2 Dummy construct, coded 1 = Australia, –1 = New Zealand, 0 = UK.
The dependent construct is defined in two ways, which means that the research model is
defined in two ways, referred to as Model 1 and Model 2. In Model 1, the dependent
construct is defined, in terms of the version 1 definition of ABC adoption, as:
()
()
log or
()
()
().
p ABC adopted
p ABC not adopted
p ABC adopted
p ABC not adopted
(4)
In this case, organisations that have adopted ABC are those that have implemented ABC
on a permanent basis, and non-adopters are all other organisations.
In Model 2, the dependent construct is defined, in terms of the version 2 definition of
ABC adoption, as:
()
()
log or
()
()
().
p ABC adopted
p ABC rejected
p ABC adopted
p ABC rejected
(5)
In this case, adoption is defined in the same way as in Model 1, but non-adopters are
defined as those that have rejected ABC.
4 Research method
Hard copies of the survey questionnaire were posted to 2,041 members of the Chartered
Institute of Management Accountants (CIMA) who worked in the management
accounting departments of a variety of organisations. The distribution of the
questionnaires was divided between 1,175 management accountants in Australia, 366 in
New Zealand and 500 in the UK. The first author administered the distribution and
collection of the questionnaires in Australia and New Zealand, and the third author
administered it in the UK.5 Three weeks after the hard copies of the questionnaire had
been posted a general announcement about the questionnaire survey was made on
CIMA’s website. This general announcement encouraged CIMA members who had
The effect of innovation characteristics on activity-based costing adoption 299
received the hard copy of the questionnaire, but had not returned it, to complete an
online version of the questionnaire. A total of 584 useable hard copy and online
questionnaire responses were received (response rate = 28.6%). This was made up of
310 questionnaires from Australia (response rate = 26.4%), 142 questionnaires from
New Zealand (response rate = 38.8%) and 132 questionnaires from the UK (response
rate = 26.4%).
The possibility of non-response bias was tested using data provided by CIMA about
the total number of CIMA members working in manufacturing and non-manufacturing
industry and the length of time they had been CIMA members. This information was
used to compare the characteristics of respondents and non-respondents, and early and
late respondents (defined as respondents who returned the questionnaires within three
weeks or three weeks after the questionnaires being sent out initially). Chi-square tests
did not reveal any significant differences between the total number of respondents and
non-respondents, and early and late respondents between manufacturing and non-
manufacturing industry. In addition, t-tests did not reveal significant differences between
these two types of respondents over the length of time they had been CIMA members.
This suggests that non-response bias may not be a problem.
Information was obtained about the adoption of ABC, by asking respondents to
indicate the extent to which ABC was used in their organisation, with responses on a
five-point nominal scale, coded:
1 discussions have not taken place regarding the introduction of this practise (which, in
this case is ABC)
2 a decision has been taken not to introduce this practise
3 some consideration is being given to the introduction of this practise
4 this practise has been introduced on a trial basis
5 this practise has been implemented and accepted.
For the purpose of this research, the version 1 measure of the adoption of ABC was
coded: 1 = this practise has been implemented and accepted and 0 = all other experiences
of ABC, and the version 2 measure was coded: 1 =this practise has been implemented
and accepted and 0 = a decision has been taken not to introduce this practise.
The characteristics of an innovation were measured by Moore and Benbasat’s (1991)
18-item short-form measure (see Appendix). The characteristics measure the perceptions
of the relative advantage, compatibility, ease of use, result demonstrability and trialability
of innovations. For each of the 18 items, respondents were asked to determine the
level of importance of the influence of each of the characteristic of an innovation on
their decision to implement a new technique (or innovation) in their organisation
with responses on a five-point Likert-type scale ranging from 5 = very important to
1 = irrelevant.
The discriminant validity of the 18-item instrument was assessed by a confirmatory
factor analysis using a principal components analysis with varimax rotation.6 Factor
loadings in excess of 0.50 [as recommended by Hair et al. (1987)] and with
Eigenvalues greater than one were regarded as significant. The five items measuring
relative advantage, three items measuring compatibility, four items measuring result
demonstrability and the two items measuring trialability each loaded on to a single
factor.7 Only two of the four items measuring ease of use loaded on to a single factor.
300 D. Askarany et al.
Table 1 Cramer’s coefficient C and Spearman rank correlation coefficients for Model 1
(N = 482)
1 2 3 4 5 6 7 8 9 10
1 Using v not using ABC
a
1.000
2 Relative advantage 0.158 1.000
3 Compatibility 0.220* 0.281** 1.000
4 Ease of use 0.131 0.293** 0.324** 1.000
5 Result demonstrability 0.282*** 0.196** 0.136** 0.283** 1.000
6 Trialability 0.204* 0.171** 0.199** 0.349** 0.283** 1.000
7 Size 0.769 0.085 0.093* –0.093* 0.068 0.088 1.000
8 Non-manufacturing industry
a
0.071 0.284*** 0.260** 0.171* 0.238** 0.254*** 0.727 1.000
9 Country1 (Australasia v UK)
a
0.108* 0.191 0.309*** 0.094 0.202 0.147 0.761 0.213*** 1.000
10 Country2 (Australia v N Zealand)
a
0.109 0.183 0.250*** 0.163* 0.197* 0.156 0.739 0.259*** 1.000*** 1.000
Notes: * p < 0.05, ** p < 0.01, *** p < 0.001
a
As these constructs are coded on nominal scales all of the correlation coefficients involving these are calculated using Cramer’s coefficient C.
All of the other correlation coefficients between the remainin
g
constructs are S
p
earman rank correlations.
The effect of innovation characteristics on activity-based costing adoption 301
These were ‘being easy to learn how to operate’ and ‘being easy to use/implement’
with loadings of 0.740 and 0.711 respectively, and these two items formed the ease of
use measure.8 The other two items had loadings of less than 0.50 and these were
omitted from the subsequent data analysis. The reliability of the five factors was
confirmed by Cronbach’s (1951) coefficient alpha, with alphas for the measures
of relative advantage = 0.684, compatibility = 0.739, ease of use = 0.688, result
demonstrability = 0.791 and trialability = 0.840. Although, only one of the reliability
scores exceeded the recommended level for alpha of 0.80 (Carmines and Zeller, 1979),
the other four alphas exceeded the minimum acceptable level of 0.60 (Price and Mueller,
1986). In addition, the discriminant validity of these five factors was confirmed by
reviewing the Spearman rank correlation coefficients between each of the factors. Table 1
shows that although all the correlations are significant (p < 0.05, 2-tailed test), the highest
correlation coefficient between any pair of the five characteristics is 0.349, which
provides further evidence that the measures of the five characteristics are measuring
different constructs. For the purpose of subsequent data analysis, the values of each
measure of the five characteristics are summed and divided by the number of items
measuring each attribute to give a score ranging from 1 to 5.
The size of an organisation was measured by respondents indicating the approximate
number of employees employed by their organisation. As the distribution of the
number of employees was positively skewed, the construct was transformed using
a logn transformation in subsequent data analysis. Whether an organisation is in
non-manufacturing or manufacturing industry was obtained by respondents indicating
their organisation’s industry and this was coded as 1 = non-manufacturing industry and
0 = manufacturing industry.
The country of a responding organisation was readily obtained for the paper-based
questionnaire respondents from the country in which the questionnaire was posted and
returned. For the online version, the respondents had to indicate the country of location
of their organisation. For the purpose of the research model, the country of the
respondents was coded into two constructs using contrast coding. The first construct
compared the two Australasian countries with the UK and was coded 1/2 = Australia,
1/2 = New Zealand and –1 = the UK, and the second construct compared Australia with
New Zealand and was coded 1 = Australia, –1 = New Zealand and 0 = the UK.
5 Results and discussion
Using listwise deletion, there were 482 useable respondents included in the logistic
regression analysis for Model 1 and 168 for Model 2. Table 2 shows the descriptive
statistics for the constructs included in the two research models. This shows that the
scores for each of the five characteristics of an innovation are higher than the midpoint of
value of 3 on the 5-point scale, the organisations vary in size from very small to very
large, the majority are from non-manufacturing industry, and they are divided
approximately between Australia, New Zealand and the UK in the ratio 2:1:1.
The remainder of the section presents the results of Model 1 and Model 2. From
Table 1, the significant Cramér’s Coefficient C with the version 1 measure of the
adoption of ABC in Model 1 are with compatibility, result demonstrability, trailability
and country1.9 The results of the logistic regression analysis for Model 1 are not reported,
however, because although 99.5% of the non-adopters of ABC are predicted correctly by
302 D. Askarany et al.
the model, only 2.1% of the adopters are predicted correctly. This is probably because the
non-adopters are a heterogeneous group (Brown et al., 2004), consisting of those who
have not considered, are considering, have rejected or have implemented ABC on a trial
basis. This is confirmed by comparing the scores on the independent constructs in
Model 1 between these respondents. The results of Kruskal Wallis statistical tests for the
five characteristics of an innovation and the number of employees, and chi-square
statistical tests for the three nominally coded independent constructs reveal significant
differences across the four non-adopter categories for the number of employees
(X2 = 12.818, df = 3, p = 0.005) and country 1 construct comparing Australasian
respondents with the UK (X2 = 11.214, df = 3, p = 0.011).10
There is a much better fit for Model 2. The model predicts 54.2% of the non-adopters
correctly and 82.3% of the adopters are predicted correctly. This results in 70.2% of the
total adopters and non-adopters being predicted correctly, which is higher than the 50%
indicated by the naïve model. The Homer and Lemeshow goodness-of-fit statistic is not
significant (p = 0.09), but this is low, however, and may indicate that there are omitted
constructs from the model.
Table 2 Descriptive statistics
Median Mean Standard deviation Minimum Maximum
Model 1 (N = 482)
Relative advantagea 4.200 4.244 0.483 2.600 5.000
Compatibilitya 3.333 3.468 0.764 1.333 5.000
Ease of usea 4.000 3.770 0.737 1.000 5.000
Result demonstrabilitya 4.000 3.939 0.592 1.500 5.000
Trialabilitya 4.000 3.708 0.778 1.000 5.000
Size (number of employees)b 298.509 321.244 11.536 1 300,000
N (%)
Adopted ABC 96 (19.9)
Not adopted ABC 386 (80.1)
482 (100.0)
Manufacturing industry 184 (38.2)
Non-manufacturing industry 298 (61.8)
482 (100.0)
Australia 244 (50.6)
New Zealand 112 (23.2)
UK 126 (26.2)
482 (100.0)
Notes: aScored from lowest possible score = 1 to highest possible score = 5.
bValues are the antilog of the logn(number of employees) values.
The effect of innovation characteristics on activity-based costing adoption 303
Table 2 Descriptive statistics (continued)
Median Mean Standard deviation Minimum Maximum
Model 2 (N = 168)
Relative advantagea 4.400 4.238 0.515 3.000 5.000
Compatibilitya 3.333 3.591 0.785 2.000 5.000
Ease of usea 4.000 3.804 0.766 2.000 5.000
Result demonstrabilitya 4.000 3.991 0.640 1.500 5.000
Trialabilitya 4.000 3.685 0.868 1.000 5.000
Size (number of employees)b 454.956 426.111 16.844 1 300,000
N (%)
Adopted ABC 96 (57.1)
Not adopted ABC 72 (42.9)
168 (100.0)
Manufacturing industry 64 (38.1)
Non-manufacturing industry 104 (61.9)
168 (100.0)
Australia 86 (51.2)
New Zealand 42 (25.0)
UK 40 (23.8)
168 (100.0)
Notes: aScored from lowest possible score = 1 to highest possible score = 5.
bValues are the antilog of the logn(number of employees) values.
Table 3 shows the Cramér’s coefficient C and Spearman rank correlation coefficients for
the 168 respondents included in Model 2. There are significant correlations with the
adoption of ABC for relative advantage, result demonstrability, size, non-manufacturing
industry, country 1 and country 2. Of the 36 correlations between the nine independent
constructs in the model, 22 (61%) are significant. Two of these correlations/coefficients
are greater than 0.50 (between ease of use and result demonstrability, and country1
and country2). In the former case, the correlation is 0.531, which indicates that
multicollinearity in the logistic regression analysis is unlikely to be a problem. In the
latter case, Cramér’s coefficient C is 1.000, which is because the country construct is
measured in two different ways. Although the latter correlation may indicate problems
with multicollinearity, the logistic regression results for Model 2 were rerun including
only one of the two dummy country constructs. This did not lead to different results from
those displayed in Table 4, when both of the country constructs are included.
304 D. Askarany et al.
Table 3 Spearman rank correlation coefficients for Model 2 (N = 168)
1 2 3 4 5 6 7 8 9 10
1 Using v not using ABC
a
1.000
2 Relative advantage 0.429** 1.000
3 Compatibility 0.292 0.381*** 1.000
4 Ease of use 0.224 0.301*** 0.425*** 1.000
5 Result demonstrability 0.358** 0.161* 0.379*** 0.531*** 1.000
6 Trialability 0.208 0.055 0.245** 0.349*** 0.313*** 1.000
7 Size 0.888 0.009 0.138 0.372*** 0.070 0.193* 1.000
8 Non-manufacturing industry
a
0.163* 0.380** 0.335* 0.411*** 0.344* 0.358** 0.898 1.000
9 Country1 (Australasia v UK)
a
0.194* 0.334* 0.315 0.233 0.293 0.206 0.882 0.310*** 1.000
10 Country2 (Australia v N Zealand)
a
0.205* 0.284 0.289 0.278* 0.334* 0.262 0.864 0.337*** 1.000*** 1.000
Notes: *p < 0.05, ** p < 0.01, *** p < 0.001
a
As these constructs are coded on nominal scales all of the correlation coefficients involving these are calculated using Cramer’s coefficient C.
All of the other correlation coefficients between the remainin
g
constructs are S
p
earman rank correlations.
The effect of innovation characteristics on activity-based costing adoption 305
Table 4 shows the results of the logistic regression analysis for Model 2 and that three of
the nine constructs are significant. These are the higher the relative advantage, the higher
the organisational size and if the operating unit is located in Australasia, the more likely it
is that operating units will adopt ABC. Hence, Hypotheses 1a and 2 are accepted.
Although it was expected that the country of origin of an organisation would not affect
ABC adoption, the results indicate that organisations in Australasia are more likely to
adopt ABC than organisations in the UK. This is shown by the version 2 adoption rates in
Australia and New Zealand being higher than in the UK at 65.1%, 57.1% and 40.0%
respectively.11 In addition, although there is an Australasian effect, there is no effect for
Australia compared with New Zealand on ABC adoption. All of the other hypotheses are
rejected. The compatibility, ease of use and result demonstrability and trialability
characteristics of innovations and whether an organisation is in non-manufacturing
industry are not related to whether an organisation adopts ABC.
Table 4 Logistic regression analysis for Model 2 (N = 168)
Regression coefficient Standard error Wald statistic p Exp(B)
Constant –3.290 1.907 2.975 0.085 0.037
Relative advantage 1.025 0.391 6.863 0.009 2.788
Compatibility 0.047 0.276 0.029 0.864 1.048
Ease of use 0.156 0.305 0.260 0.610 1.168
Result demonstrability –0.405 0.351 1.331 0.249 0.667
Trialability –0.420 0.247 2.887 0.089 0.657
Size 0.229 0.081 7.898 0.005 1.257
Non-manufacturing
industry
0.296 0.400 0.548 0.459 1.345
Country1
(Australasia v UK)
0.695 0.311 5.009 0.025 2.004
Country2
(Australia v New Zealand)
0.098 0.222 0.196 0.658 1.103
Notes: Hosmer and Lemeshow test: X2 = 13.687, df = 8, p = 0.090
–2 Log Likelihood = 197.468
Cox and Snell R2 = 0.173
Nagelkerke R2 = 0.233
6 Conclusions
This paper has developed and tested a model of the extent to which the importance of the
characteristics of an innovation, the size, industry and country of location of
organisations influences the extent to which ABC has been adopted, when the adoption of
ABC has been defined in two distinct ways. The results of the research show that when
organisations that adopt ABC are compared with all other organisations the model cannot
be operationalised because it cannot predict organisations that have adopted ABC. This is
likely to be because the non-adopters are a heterogeneous group made up of those that are
considering, have rejected and have not considered ABC. When the organisations that
adopt ABC are compared with those that reject it, the model can be operationalised
because adopters and rejecters have in common the fact that they have both considered
306 D. Askarany et al.
ABC. Given that only one of the characteristics of an innovation is related positively to
those that adopt ABC relative to those that reject it, Rogers’ (2003) assertion that
between 49% and 87% of the variance in the rate of adoption of innovations is accounted
for by the characteristics of innovations is clearly not the case for the adoption of ABC in
the organisations studied in this research. The positive association between relative
advantage and ABC adoption indicates that the decision to adopt is dependent partly on
the rational consideration that the advantages offered by an innovation to potential
adopters should exceed its disadvantages. Thus, the adopters of ABC systems regard
them as superior to traditional costing systems that they are likely to have replaced
(Brown et al., 2004).
Although the other four characteristics have not been tested in prior ABC adoption
research, the results of this exploratory research into their influence indicate that they are
not related to ABC adoption. Although relative advantage has a significant effect on
adoption, the non-significant effect of result demonstrability indicates that adopters of
ABC are not concerned with rationalising their decisions in order to highlight the
purported relative advantages of ABC. This indicates that management do not find it
worthwhile to direct organisational resources to identify the advantages of ABC in terms
of extensive training programs and the provision of information about ABC. Thus,
although the ABC system should possess the appropriate relative advantages that
overcome the deficiencies with the existing cost accounting system, the results of this
research indicate that the users of the ABC system do not need to be provided with
detailed information that ABC has the appropriate relative advantages to overcome
deficiencies with the current system. This is a counter-intuitive result that needs to be
confirmed in future research, because the most effective way of indicating the relative
advantage of an innovation, like ABC, lies with demonstrating its results.
The non-significant effect of the importance of compatibility on ABC adoption
indicates that it is not important for work patterns and work flows to be compatible
throughout the system. This result may be a consequence of the measure not
distinguishing between organisational compatibility and technical compatibility (Beatty
et al., 2001). In the former case, organisations are assumed more likely to adopt an
innovation if it is perceived, as being consistent with organisational culture, values, work
practises and the existing IS infrastructure (Beatty et al., 2001). In the case of technical
compatibility, this is the extent to which the innovation can be integrated into the IS
environment (Beatty et al., 2001). The non-significance of compatibility needs to be
confirmed in future research by utilising a measure of compatibility that distinguishes
between these two types of compatibility.
Ease of use’s non-significance indicates that the respondents anticipate being able to
overcome any problems with trying to use an innovation like ABC. This may be because
the respondents to the survey are all qualified management accountants who are well
educated and may be able to cope with the introduction of new techniques into their
organisations. However, lower-ranked clerical staff, who may be involved in inputting
data into the accounting system, may attach a great deal of importance to new techniques
being easy to use. Hence, future research should replicate this research with samples of
unqualified management accounting staff that would be likely to have day-to-day contact
with the ABC system and compare the results with those obtained in this research.
Notwithstanding the fact that 32 respondents to the survey had accepted ABC on a
trial basis, the trialability does not have a significant effect on ABC adoption. This
indicates that the decision to adopt ABC does not depend on whether an innovation is
The effect of innovation characteristics on activity-based costing adoption 307
subject to some sort of trial prior to the decision to adopt it permanently. However, this
result needs to be interpreted in the context that is not known to what extent the adopters
and rejecters of ABC undertook a trial of ABC, and, if they did, what form this trial took.
The extent to which adopters and rejecters undertake a trial of ABC prior to making a
final decision regarding its adoption needs to be confirmed in future research.
The results confirm, as in prior research, that larger firms are more likely to
adopt ABC. Hence, organisations must have the appropriate level of resources and
infrastructure to adopt ABC. In subsequent research, there is a need to measure these
resources and infrastructure more precisely than a proxy construct based on size. This
could involve the development of psychometric constructs to measure the amount of
funds available for investment in accounting innovations and the amount of support
available both within and outside the organisation to support and maintain accounting
innovations.
The results indicate that the country of adoption is a significant factor influencing
adoption. There is a need to undertake further cross-country studies to confirm this result
and compare ABC adoption between other countries. There is a need for interpretive
studies to understand the historical and cultural factors in the development of ABC in
these countries that have led to the greater likelihood of ABC being adopted in
Australasia than in the UK.
Although it was hypothesised that non-manufacturing organisations would be more
likely to adopt ABC, the results did not reveal any relationship between industry and
ABC adoption. Thus, although Kaplan and Cooper (1998) suggest that organisations in
non-manufacturing industry may be more likely to adopt ABC, the results of this research
suggest that this is not the case.
The main limitation of this research stems from the cross-sectional survey
questionnaire approach adopted. Cross-sectional models, such as this one, may not
identify fully the complexity of the adoption decision. The data used in Model 2 was
collected after the organisations had accepted or rejected ABC rather than during the
decision-making process that led to the adoption/rejection decision. Future research
should investigate how the various possible influences discussed above influence the
adoption decision at the various stages in the adoption process and how these influences
change over time. This could be carried out by undertaking a longitudinal research design
using quantitative or qualitative methods, to see how these influences change over the
adoption decision.
Related to the above, there is a lack of contextual information and in depth
information obtained from the questionnaire approach that can be obtained instead from
interviews and case studies. Furthermore, the respondents are qualified management
accountants, so care should be taken when trying to generalise the results of this study to
organisations in Australia, New Zealand and the UK that do not employ qualified
management accountants. Furthermore, although tests for non-response bias were
undertaken, given that the overall response rate to the survey is 28.6%, non-response bias
may be a problem. Another limitation is that respondents are required to self-specify
whether they are using ABC. Given the problems of defining what is meant by ABC
(Dugdale and Jones, 1997), it is possible that the reported adoption rates for ABC may be
under or overstated.
A further limitation of the research relates to the measurement of the importance of an
innovation is that the questionnaire assesses as the impact of the perception, in general,
on the adoption of ABC. It is possible that the respondents’ perceptions of the importance
308 D. Askarany et al.
of the characteristics of the innovation called ABC are different from those for some
general innovation. To test for this effect, future research should replicate this study to
examine if the results are different when the characteristics of an innovation are measured
in terms of the importance of the ABC innovation.
A possible reason for the non-significant effects in Model 2 and the problems with
Model 1 arises from model misspecification such as from the omission of an intervening
construct in the model that results in the characteristics of an innovation having an
indirect effect on the intention to adopt. For example, using the Theory of Reasoned
Action (Fishbein and Ajzen, 1975), Karahanna et al. (1999) model and confirm that the
effect of the characteristics of an innovation on the adoption of Windows software is an
indirect effect through the attitude to the innovation. This modification to IDT research
could be incorporated into future models of ABC adoption.
This paper contributes to our understanding of the factors influencing the decision to
adopt ABC because it is the first study to test the influence of a variety of the
characteristics of an innovation on the adoption decision. The results indicate that only
the perceived relative advantage of an innovation has a significant effect on adoption.
Organisations when considering whether to adopt an innovation like ABC are concerned
with the worthwhileness of the outputs from the ABC system. In addition, the results
show that the adoption varies across countries. Given this, there is a need to conduct
further multinational studies to examine whether adoption rates vary between other
countries and, if so, why they vary.
Acknowledgements
The authors are grateful to the Research Foundation of the Chartered Institute of
Management Accountants in the UK, the University of Sheffield and the University of
Auckland for funding and facilitating this research.
References
Abdel-Maksoud, A.B. (2011) ‘Management accounting practices and managerial techniques and
practices in manufacturing firms: Egyptian evidence’, International Journal of Managerial
and Financial Accounting, Vol. 3, No. 3, pp.237–254.
Agarwal, R. and Prasad, J. (1997) ‘The role of innovation characteristics and perceived
voluntariness in the acceptance of information technologies’, Decision Sciences, Vol. 28,
No. 3, pp.557–582.
Al-Omiri, M. and Drury, C. (2007) ‘A survey of factors influencing the choice of product
costing systems in UK organizations’, Management Accounting Research, Vol. 18, No. 4,
pp.399–424.
Anderson, S.W. (1995) ‘A framework for assessing cost management system changes: the case of
activity-based costing implementation at General Motors, 1986–1993’, Journal of
Management Accounting Research, Vol. 7, pp.1–51.
Armitage, H.M. and Nicholson, R. (1993) Activity-Based Costing – Management Accounting Issues
Paper 3, The Society of Management Accountants, Hamilton, Canada.
Bartlett, M.S. (1950) ‘Test of significance in factor analysis’, British Journal of Statistical
Psychology, Vol. 3, No. 2, pp.77–85.
Bartlett, M.S. (1951) ‘A further note on tests of significance in factor analysis’, British Journal of
Statistical Psychology, Vol. 4, No. 1, pp.1–2.
The effect of innovation characteristics on activity-based costing adoption 309
Beatty, R.C., Shim, J.P. and Jones, M.C. (2001) ‘Factors influencing corporate web site adoption: a
time-based assessment’, Information & Management, Vol. 38, No. 6, pp.337–354.
Bhattacherjee, A. (1998) ‘Managerial influences on intra-organizational information technology
use: a principal-agent model’, Decision Sciences, Vol. 29, No. 1, pp.139–162.
Bhimani, A., Gosselin, M., Ncube, M. and Okana, H. (2007) ‘Activity-based costing: how far have
we come internationally?’, Cost Management, May–June, No. 3, pp.12–17.
Bjørnenak, T. (1997) ‘Diffusion and accounting: the case of ABC in Norway’, Management
Accounting Research, Vol. 8, No. 1, pp.3–17.
Booth, P. and Giacobbe, F. (1998) ‘The impact of demand and supply factors in the diffusion of
accounting innovations: the adoption of activity-based costing in Australian Manufacturing
Firms’, paper presented at the Management Accounting Research Conference, University of
NSW, Sydney.
Brierley, J.A. (2011) ‘Why the proper definition of the ABC matters: a note’, Advances in
Management Accounting, Vol. 19, pp.225–249.
Brown, D.A., Booth, P. and Giacobbe, F. (2004) ‘Technological and organizational influences on
the adoption of activity-based costing in Australia’, Accounting and Finance, Vol. 44, No. 3,
pp.329–356.
Cagwin, D. and Bouwman, M.J. (2002) ‘The association between activity based costing and
improvement in financial performance’, Management Accounting Research, Vol. 13, No. 1,
pp.1–39.
Carmines, E.G. and Zeller, R. (1979) Reliability and Validity Assessment, Sage, Beverley Hills,
CA.
Clarke, P.J., Hill, N.T. and Stevens, K. (1999) ‘Activity-based costing in Ireland: barriers to, and
opportunities for, change’, Critical Perspectives on Accounting, Vol. 10, No. 4, pp.443–468.
Cohen, S., Venieris, G. and Kaimenaki, E. (2005) ‘ABC: adopters, supporters, deniers and
unawares’, Managerial Auditing Journal, Vol. 20, No. 9, pp.981–1000.
Cooper, R. (1988a) ‘The rise of activity-based costing – part one: what is an activity-based cost
system?’, Journal of Cost Management for the Manufacturing Industry, Vol. 2, No. 2,
pp.45–54.
Cooper, R. (1988b) ‘The rise of activity-based costing – part two: when do I need an activity-based
cost system?’, Journal of Cost Management for the Manufacturing Industry, Vol. 2, No. 3,
pp.41–48.
Cooper, R. (1990) ‘Explicating the logic of ABC’, Management Accounting, November, Vol. 58,
No. 10, pp.58–60, UK.
Cooper, R. (1991) ‘ABC: the right approach for you?’ Management Accounting, January, Vol. 59,
No. 1, pp.70–72, UK.
Cooper, R. (1997) ‘Activity-based costing: theory and practice’, in Brinker, B.J. (Ed.): Handbook
of Cost Management, Warren Gorham and Lamont, New York, NY.
Cooper, R. and Kaplan, R.S. (1987) ‘How cost accounting systematically distorts product costs’, in
Burns, W.J. and Kaplan, R.S. (Eds.): Accounting and Management: Field Study Perspectives,
Harvard Business School, Boston, MA.
Cooper, R. and Kaplan, R.S. (1988) ‘Measure costs right: make the right decisions’, Harvard
Business Review, September–October, Vol. 66, No. 5, pp.96–103.
Cooper, R. and Kaplan, R.S. (1991) The Design of Cost Management Systems: Text, Cases and
Readings, 1st ed., Prentice Hall, Englewood Cliffs, NJ.
Cooper, R. and Kaplan, R.S. (1999) The Design of Cost Management Systems: Text, Cases and
Readings, 2nd ed., Prentice Hall, Upper Saddle River, NJ.
Cooper, R., Kaplan, R., Maisel, L.S., Morrissey, E. and Oehm, R.M. (1992) Implementing
Activity-Based Cost Management: Moving from Analysis to Action, Institute of Management
Accountants, Montvale, NJ.
310 D. Askarany et al.
Cotton, D.J., Jackman, S.M. and Brown, R.A. (2003) ‘Note on a New Zealand replication of the
Innes et al. UK activity-based costing survey’, Management Accounting Research, Vol. 14,
No. 1, pp.67–72.
Cronbach, L.J. (1951) ‘Coefficient alpha and the internal structure of tests’, Psychometrika,
Vol. 16, No. 3, pp.297–334.
Davis, F.D., Bagozzi, R.P. and Warshaw, P.R. (1989) ‘User acceptance of computer technology: a
comparison of two theoretical models’, Management Science, Vol. 35, No. 8, pp.982–1003.
Drury, C. and Tayles, M. (2000) Cost System Design and Profitability Analysis in UK Companies,
Chartered Institute of Management Accountants, London.
Drury, C., Braund, S., Osborne, P. and Tayles, M. (1993) A Survey of Management Accounting
Practices in UK Manufacturing Companies, Chartered Association of Certified Accountants,
London.
Dugdale, D. and Jones, T.C. (1997) ‘How many companies use ABC for stock valuation? A
comment on Innes & Mitchell’s questionnaire findings’, Management Accounting Research,
Vol. 8, No. 2, pp.233–240.
Dziuban, C.D. and Shirkey, E.C. (1974) ‘When is a correlation matrix appropriate for factor
analysis’, Psychological Bulletin, Vol. 81, No. 6, pp.358–361.
Fishbein, M. and Ajzen, I. (1975) Belief, Attitude, Intention and Behavior: An Introduction to
Theory and Research, Addison-Wesley, Reading, MA.
Frambach, R.T. and Schillewaert, N. (2002) ‘Organizational innovation adoption. A multi-level
framework of determinants and opportunities for future research’, Journal of Business
Research, Vol. 55, No. 2, pp.163–176.
Gosselin, M. (1997) ‘The effect of strategy and organizational structure on the adoption and
implementation of activity-based costing’, Accounting, Organizations and Society, Vol. 22,
No. 2, pp.105–122.
Groot, T.L.C.M. (1999) ‘Activity-based costing in the U.S. and Dutch food companies’, Advances
in Management Accounting, Vol. 7, pp.47–63.
Grover, V. (1993) ‘An empirically derived model for the adoption of customer-based
interorganizational systems’, Decision Sciences, Vol. 24, No. 3, pp.603–640.
Hair, J.F., Anderson, R.E. and Tatham, R.L. (1987) Multivariate Data Analysis, 2nd ed.,
Macmillan, New York, NY.
Innes, J. and Mitchell, F. (1995) ‘A survey of activity-based costing in the UK’s largest
companies’, Management Accounting Research, Vol. 6, No. 2, pp.137–153.
Innes, J., Mitchell, F. and Sinclair, D. (2000) ‘Activity-based costing in the UK’s largest
companies: a comparison of 1994 and 1999 survey results’, Management Accounting
Research, Vol. 11, No. 3, pp.349–362.
Jebeile, S. and Reeve, R. (2003) ‘The diffusion of e-learning in an Australian secondary college:
strategies and tactics for educational leaders’, The Innovation Journal, Vol. 8, No. 4, pp.1–21.
Johnson, H.T. and Kaplan, R.S. (1987) Relevance Lost: The Rise and Fall of Management
Accounting, Harvard Business School Press, Boston, MA.
Kaiser, H.F. (1963) ‘Image analysis’, in Harris, C.W. (Ed.): Problems in Measuring Change,
University of Wisconsin Press, Madison.
Kaiser, H.F. (1974) ‘An index of factorial simplicity’, Psychometrika, Vol. 39, No. 1, pp.31–36.
Kaplan, R.S. (1985) ‘Accounting lag: the obsolescence of cost accounting systems’, in Clark, K.
and Lorenze, C. (Eds.): Technology and Productivity: The Uneasy Alliance, Harvard Business
School Press, Boston, MA.
Kaplan, R.S. and Cooper, R. (1998) Cost & Effect: Using Integrated Cost Systems to Drive
Profitability and Performance, Harvard Business School Press, Boston, MA.
Karahanna, E., Straub, D.W. and Chervany, N.L. (1999) ‘Information technology adoption across
time: a cross-sectional comparison of pre-adoption and post-adoption beliefs’, MIS Quarterly,
Vol. 23, No. 2, pp.183–213.
The effect of innovation characteristics on activity-based costing adoption 311
Krumwiede, K.R. (1998) ‘The implementation stages of activity-based costing and the impact of
contextual and organizational factors’, Journal of Management Accounting Research, Vol. 10,
pp.239–277.
Lai, V.S. (1997) ‘Critical factors in ISDN implementation: an exploratory study’, Information &
Management, Vol. 33, No. 2, pp.87–97.
Lamminmaki, D. and Drury, C. (2001) ‘A comparison of New Zealand and British product costing
practices’, International Journal of Accounting, Vol. 36, No. 3, pp.329–347.
Lukka, K. and Granlund, M. (1996) ‘Cost accounting in Finland: current practice and trends of
development’, European Accounting Review, Vol. 5, No. 1, pp.1–28.
Malmi, T. (1999) ‘Activity-based costing diffusion across organizations: an exploratory
empirical analysis of Finnish firms’, Accounting, Organizations and Society, Vol. 24, No. 8,
pp.649–672.
Moore, G.C. and Benbasat, I. (1991) ‘Development of an instrument to measure the perceptions of
adopting an information technology innovation’, Information Systems Research, Vol. 2, No. 3,
pp.192–222.
Pollanen, R.M. and Abdel-Maksoud, A.B. (2010) ‘The deployment of contemporary management
accounting practices in Canadian firms: a contingency approach’, International Journal of
Managerial and Financial Accounting, Vol. 2, No. 2, pp.134–152.
Price, J.L. and Mueller, C.W. (1986) Handbook of Organizational Measurement, Pitman,
Marshfield, MA.
Rogers, E.M. (1983) Diffusion of Innovations, 3rd ed., Free Press, New York, NY.
Rogers, E.M. (2003) Diffusion of Innovations, 5th ed., Free Press, New York, NY.
Siegel, S. and Castellan, N.J. (1988) Non-parametric Statistics for the Behavioral Sciences,
2nd ed., McGraw Hill, New York, NY.
Swanson, E.B. (1994) ‘Information systems innovation among organizations’, Management
Science, Vol. 40, No. 9, pp.1069–1088.
Wynekoop, J.L., Senn, J.A. and Conger, S.A. (1992) ‘The implementation of CASE tools: an
innovation diffusion approach’, in Kendall, K.E., Lyytinen, K. and DeGross, J.L. (Eds.): The
Impact of Computer Supported Technology in Information Systems Development, Elsevier,
Amsterdam.
Yakhou, M. and Dorweiler, V.P. (1995) ‘Advanced cost management systems: an empirical
comparison of England, France and United States’, Advances in International Accounting,
Vol. 8, pp.99–127.
Notes
1 Al-Omiri and Drury’s (2007) definition of adoption is obtained from examining this paper’s
research questionnaire in conjunction with the research paper.
2 The effect observed in Grover (1993) and Lai (1997) is the negative influence of complexity
on adoption.
3 From the data supplied by Drury et al. (1993), it is possible to calculate a chi-square statistic
(X2 = 1.425, df = 1, p > 0.05).
4 From the data supplied by Cagwin and Bouwman (2002), it is possible to calculate a chi-
square statistic (X2 = 7.228, df = 1, p < 0.01).
5 A copy of the questionnaire is available from the second author.
6 The appropriateness of the factor analysis was confirmed by the Kaiser-Meyer Olkin measure
of sampling adequacy (Kaiser, 1974), Bartlett’s Test of Specificity (Bartlett, 1950, 1951) and
reviewing the off-diagonal elements of the anti image covariance matrix (Kaiser, 1963;
Dziuban and Shirkey, 1974).
312 D. Askarany et al.
7 The factor loadings for the measure of relative advantage were 0.642, 0.723, 0.552, 0.651 and
0.597; for the measure of comparability they were 0.632, 0.857 and 0.813; for the measure of
result demonstrability they were 0.831, 0.858, 0.643 and 0.653; and for the measure of
trialability they were 0.839 and 0.871.
8 These are the first and third items listed under the ease of use measure in the Appendix.
9 Cramér’s coefficient C measures the degree of association between one or more constructs
that are coded on a nominal scale (see Siegel and Castellan, 1988). In this paper, these
constructs are the adoption of ABC, industry and the two country constructs.
10 For the number of employees construct, the multiple comparisons between treatments
procedure indicates that organisations that are considering ABC have significantly more
employees than those that have not considered it (p < 0.05). For the Country 1 construct, for
those organisations that have not implemented ABC on a permanent basis, a significantly
higher proportion of organisations in Australasia have not considered ABC relative to those
that have considered it some extent (X2 = 5.775, df = 3, p = 0.016).
11 The percentages are calculated as Number adopted ABC/Number adopted and rejected ABC.
Appendix
Moore and Benbasat’s (1991) 18-item short-form measure of the characteristics of
innovation.
Relative advantage
Its ability to get the job/service done quicker
Its ability to improve the quality of the job/service
Its ability to do the job/service easier
Its ability to increase the overall effectiveness of the job/service
Its ability to offer greater control over work processes (job/service).
Compatibility
Being compatible with all aspects of existing processes (job/service)
Fitting well with the way I/organisation like to work
Fitting into my/organisation work style.
Ease of use
Being easy to learn how to operate
Offering clear and understandable interaction with the technique
Being easy to use/implement
Being easy to get the technique to do what I/organisation want it to do.
The effect of innovation characteristics on activity-based costing adoption 313
Result demonstrability
Having no difficulty telling others about the results of using the technique
Being able to communicate to others the consequence of using the technique
Being able to see the results of using the technique clearly
Being able to explain why using the technique may or may not be beneficial.
Trialability
Being able to try the technique before deciding to implement it (or not)
Being allowed to use the technique on a trial basis long enough to see what it could do.
