Innovation Theory of Planned Behaviour 1
DEVELOPING AN INTEGRATIVE MODEL FOR UNDERSTANDING
KERRIE L. UNSWORTH
UWA Business School
University of Western Australia
Crawley WA 6103, Australia
Queensland University of Technology
Queensland University of Technology
We develop and test a theoretically-based integrative model of organizational innovation
adoption. Confirmatory factor analyses using responses from 134 organizations showed that
the hypothesized second-order model was a better fit to the data than the traditional model of
independent factors. Furthermore, although not all elements were significant, the
hypothesized model fit adoption better than the traditional model.
Innovation Theory of Planned Behaviour 2
The interplay between the micro and macro in innovation adoption (that is, the role
played by individuals and groups affecting organizational-level adoption) has produced
swathes of literature identifying numerous predictor variables. However, much of this
literature has examined the factors separately with no integration across factors (e.g. 1995;
Tabak & Barr, 1996). Those researchers who have attempted to organize these variables into
categories have done so either according to level (e.g., individual, organizational,
contextual:Meyer & Goes, 1988; Wejnert, 2002) or descriptive categorization (e.g., perceived
benefit, external pressure and organizational readiness: Iacovou, Benbasat, & Dexter, 1995),
rather than underlying theory. Indeed, one of the most widely-cited of these pieces of
research explicitly stat their atheoretical examination of innovation adoption (Kimberly &
Evanisk, 1981). Although these piecemeal studies and atheoretical categorizations have
contributed greatly to our knowledge of innovation adoption and the role that people play
within that process, they do not allow us to understand why the factors are important. An
underlying theoretical framework would allow us to have a deeper understanding of adoption,
promote the integration of research findings into a reasoned set of strategies, and most
importantly, allow for the prediction of new areas of research. We approached this need for
an underlying theoretical framework by initially reviewing the extant literature in the area. In
doing so, we found that categories emerged that were analogous to the Theory of Planned
Behavior (TPB: Ajzen, 1985, 1991).
The Innovation Theory of Planned Behavior (ITPB)
Innovation Theory of Planned Behaviour 3
In the ITPB model, we suggest that attitudes towards innovation, subjective norms for
innovation and perceived control over innovation will affect innovation adoption behaviour.
We now briefly review the literature to show how extant variables are representative of each
of these factors before discussing the tests of these propositions. A summary of a more
detailed review of the innovation adoption literature can be obtained from the first author.
Attitudes towards innovation have been manifested in a number of forms within the
organizational innovation literature. Most explicit is the meta-analytic finding by Damanpour
(1991) that senior managers’ attitudes towards the innovation was significantly related to
organizational innovation adoption. Attitudes towards risk- taking within the organization
have also been related to innovation adoption (e.g. Damanpour, 1991; Nystrom,
Ramamurthy, & Wilson, 2002), and, because innovation is inherently risky, may be seen as
attitudes towards innovation. Finally, the perceived benefits of innovating are also proposed
to be manifestations of attitudes towards innovation as they reflect the cognitive component
of the attitude such benefits have been found to be related to innovation adoption of
information technology (Mehrtens, Cragg, & Mills, 2001; Min & Galle, 2003)
Subjective norms are perceived pressures, exerted by significant others, to adopt
innovations. In previous research, these norms have typically come from customers and
suppliers (Iacovou et al., 1995; Mehrtens et al., 2001), competitors, and government
departments (Drazin & Schoonhoven, 1996).
Perhaps the most widely-studied category of factors is that of perceived behavioural
control (PBC). PBC is defined as the perceived ease (or difficulty) of adopting and
implementing an innovation; when the process is perceived as easy, then the organizational
decision-makers feel as though they have control. We suggest that factors such as financial
resources (Bates & Flynn, 1995) and organizational readiness, including knowledge,
technical, and staff readiness (Iacovou et al., 1995; Lehman, Greener, & Simpson, 2002;
Innovation Theory of Planned Behaviour 4
Snyder-Halpern, 2001) all act as indicators of ease of adoption. In other words, when an
organization has the required resources and readiness to adopt, its decision-makers will
perceive that process to be relatively easy and hence feel greater control over adoption.
Despite this apparent mapping of the ITPB model onto previous literature we need to
test the model’s validity. Thus, we will use structural equation modelling to conduct
confirmatory factor analyses (CFA) where an overall attitudes latent factor and an overall
PBC latent factor load onto variables that are currently used within the literature (general
attitude toward innovation, risk-taking culture, and experience of innovation benefits; and
financial resources, knowledge and technical readiness, human resources readiness, and
perceived implementation efficacy; respectively). However, to ensure that this model is the
most appropriate, two theoretically-plausible alternative models are also specified. The first
of these is a nested traditional first-order CFA where all variables are correlated but
independent, and the second is a second-order, one-factor model where one factor only loads
onto all variables (that is, the currently identified variables all represent an underlying
Hypothesis 1: The second-order, two- factor CFA model will provide a significantly
better fit to the data than a nested traditional first-order model or a second-order, one-factor
We also hypothesize that overall attitudes, overall PBC and overall subjective norms
will be more strongly related to product, process and technological innovation adoption than
the individual variables will be on their own. In other words, we hypothesize that:
Hypothesis 2: The second-order model with latent attitudes, PBC and subjective
norms will fit innovation adoption data better than the traditional first-order model.
Innovation Theory of Planned Behaviour 5
Sample and Procedure
The survey was sent to the Managing Director of 864 organizations from business
register databases. Thirty- three questionnaires were unable to be delivered. A reminder was
sent to those who had not returned their survey a fortnight later. In total, 134 organizations
responded (16.1% response rate). The majority of respondents were from small firms (62%
has less than 50 employees) in the manufacturing industry (55%).
Attitudinal Variables. Risk-taking culture was measures by a four-item factor taken
from Litwin and Stringer (1968). An example item is “The philosophy of our management is
that in the long run we get ahead playing it slow, safe and sure.” The organization’s general
attitude towards innovation was defined as an organizations’ overall evaluation of
innovation. The scale was developed based on the methods used by researchers of the Theory
of Planned Behaviour (Ajzen, 1991). Respondents were asked how their organization viewed
innovation along five adjective pairs (e.g., a bad idea- a good idea) along a scale from -3 to
+3. Experiences of innovation benefits indicate an organization’ realization of the intended
benefits of previously- adopted innovations across financial, customer and employee benefits
(Totterdell, Leach, Birdi, Clegg, & Wall, 2002).
Perceived Behavioral Control Variables. Financial resource availability was
measured by four items from Klein, Conn and Sorra (2001), an example item is “Money is
readily available to pay for special projects in the organization”. Innovation efficacy refers to
perceptions of confidence that the organization is capable of adopting and implementing
Innovation Theory of Planned Behaviour 6
innovations. We developed three items based on Theory of Planned Behaviour research to
measure the likelihood that organizations perceived they were capable of introducing
innovation (Ajzen, 1991) – e.g., “I am confident that innovation would be successful in this
organization”. The human resources readiness scale included two items on availability of
skilled labour resources and managerial talent (Nystrom et al., 2002). Technical and
knowledge readiness measured the degree to which organizations possessed existing
knowledge and technologies to support any new innovation. Nine questions were adapted
from Iacavou et al.’s (1995) organizational readiness framework, which included perceptions
about adequacy of innovative knowledge, technical knowledge, and availability of hardware
Subjective Norms for Innovation. This measure examined the perceived external
pressure from suppliers, customers, competitors, technology diffusion agencies, universities,
and government departments to engage or not to engage in innovation adoption. Subjective
norms for innovation are compromised of two elements- the degree to which the external
stakeholder supports innovation adoption, and the degree to which the organization values the
opinion of that stakeholder. As such, there were two sets of questions that were then
multiplied to obtain measures of perceived pressure from each type of stakeholder (Ajzen,
1991). Using the criteria outlined by MacKenzie, Podsakoff and Jarvis (2005) we suggest that
the indicators of perceived pressure from various stakeholders are defining and causal
characteristics of overall subjective norms but that they do not necessarily share a common
theme or antecedents and are therefore an index with formative rather than reflective
indicators. Therefore, the indicators were summed to give an overall index of subjective
Innovation Adoption. Following Totterdell et al. (2002), we asked respondents to
identify the innovations that their organizations had adopted over the last three years. The
Innovation Theory of Planned Behaviour 7
innovations were product-oriented (new products, changes to existing products, changes in
business services), technologically-oriented (new plant or machinery, new manufacturing or
product-based technology), and process-oriented (new processes or work design systems,
new administrative systems, HRM innovations and organizational restructuring innovations).
Because we were interested in innovation adoption per se, rather than breadth of adoption
types or amount of adoption, an organization was given a score of 1 for the innovation type if
they had adopted at least one innovation from within that category.
Industry & Radicalness Controls. Because close to half of the participating
organizations belonged to the manufacturing industry, a dummy variable for manufacturing
was created. In addition respondents were asked to rate on a 5-point Likert scale, the extent to
which the adopted innovations were radically different from what the organization had or did
before. An index was then created to represent the overall radicalness of the innovations.
These two controls were included in all analyses.
Modeling Latent Factors of Attitudes and Perceived Behavioral Control
Initial congeneric models showed that four items did not load onto their hypothesized
scale or cross-loaded to other items, thus we removed these items from our analysis (items
available on request). The hypothesized second-order, two-factor model had a good fit to the
data (χ2 = 321.73, df= 265, p<.05; RMSEA=.04 (CI=.02-.05); CFI= .95). The composite
reliabilities for the two second-order factors were both high (.91 and .95 for the overall
attitudes and overall PBC, respectively), indicating high convergent validity (Fornell &
Larcker, 1981). The average variances extracted were .45 for attitudes and .53 for PBC. The
Innovation Theory of Planned Behaviour 8
average variance extracted for attitudes was, therefore, slightly below that recommended by
Fornell and Larcker (1981) to provide strong evidence for discriminant validity. Thus, we
compared this model with a second-order, one-factor model. This model also had a
reasonable fit data (χ2 = 410.85, df = 316, p<.05; RMSEA = .05 (CI=.03- .06); CFI=.93), but
the difference in the chi-squares obtained was significant, indicating that the inclusion of a
second latent factor significantly increased the goodness of fit (∆χ2 =89.12, df = 51, p<.01)
and providing support for the discriminant validity of the two factors.
Finally, the traditional, nested first-order model was tested. Because HR readiness
consisted of only two items, it was unable to be identified. Therefore, we used the factor
regression weights obtained from the hypothesized model (the best-fitting model) to fix the
item regression weights. The fit to data of this nested first-order model was poor (χ2 = 453.1,
df= 271, p<.05; RMSEA=.07 (CI=.06-.09); CFI=.83), particularly in comparison to the
hypothesized second-order, two factor model (∆χ2 = 131.79, df=6, p<.05).
Although we do not differentiate between types of innovation adoption in the ITPB
model, we checked the robustness of the hypothesized model fit across different types. As
expected, the fit of the hypothesized model was similar for organizations that adopted product
innovations, process innovations, and technological innovations. The pattern of standardized
regression weights for the different innovation adoption types was also similar (findings
available on request). Most importantly, no regression weights changed significance across
the different sub-samples. Therefore, we suggest that the mapping of variables onto the ITPB
framework holds across small organizations adopting different types of innovation.
When examining the standardized regression weights in the hypothesized model, we
found that, as hypothesized, risk-taking (β = .51, p<.05), general attitude towards innovation
(β = .53, p<.05), and previous experiences with innovation (β = .88, p<.05) were all
significantly related to the second-order latent attitudes factor. Also as hypothesized,
Innovation Theory of Planned Behaviour 9
financial resources (β = .50, p<.05), or organizational efficacy β = .59, p<.05), knowledge
and technical readiness (β = .82, p<.05), and HR readiness (β = .73, p<.05) were all
significantly related to the second-order latent PBC factor. Interestingly, the latent attitudes
factor and latent PBC factor were significantly correlated (r=.64, p<.05).
Evaluating the Overall Model of Innovation Adoption
To test the advantages provided by the ITPB, we compared the structural models
containing latent factors for overall attitudes and overall PBC with those where the first-order
factors for attitudinal and PBC indicators were related directly to innovation adoption. The
hypothesized model was a better fit for product innovation adoption (∆χ2 = 96.7, df=2,
p<001). These findings suggest that the inclusion of the two additional latent factors provide
a better fit to the data than the traditional first-order models and provide some support to our
Nevertheless, it is important to note that not all hypotheses were supported. In
particular, while the effects of overall PBC were significant and in the hypothesized
direction, the effects of overall attitudes and subjective norms on innovation adoption were
generally non-significant. Furthermore, the regression weight between overall attitudes and
technological innovation adoption was negative. In investigating this surprising finding
further, we conducted a partial correlation between overall attitudes and technological
adoption controlling for the industry and radicalness variables and found a significant,
positive relationship between the two (r=.17, p<.05). It is likely therefore that the negative
relationship for attitudes in technological adoption is due to a suppression effect from overall
PBC due to its high correlation with overall attitudes. Thus, while there is overall support for
the effectiveness of the hypothesized model in fitting the data for innovation adoption, certain
elements of the model were not supported.
Innovation Theory of Planned Behaviour 10
In this paper, we developed a framework to integrate and understand organizational
innovation adoption research from the micro-macro perspective and conducted preliminary
analyses to support such a model. In contrast to previous frameworks, we used a well-studied
theory to provide a coherent reasoning for categorising variables within the extant innovation
adoption literature. Not only did we find that the proposed overall constructs accounted for
many of the variables previously identified in the literature, but that structural models
including these latent factors fit the innovation adoption data significantly better than the
traditional first-order models. Our novel approach of applying the TPB to innovation
adoption has three implications. First it allows us to understand why certain factors are
important in predicting innovation adoption, at least among small organizations. Second, it
identifies a number of areas for future research that might prove fruitful in our understanding
of successful innovation adoption. Finally, it represents a practical contribution by focusing
change efforts on those proximal factors that are most likely to create an effect in small firms.
A checklist of predictive factors is a useful starting point for an area of study, however, a
theoretical underpinning is necessary to move forward in the literature. We believe that the
ITPB model provides such an underpinning.
Surprisingly, although the overall ITPB model provided better fit to the adoption data,
attitudes and subjective norms were generally not related to innovation adoption. One
possible explanation for this finding comes from the way in which we measured attitudes and
subjective norms. Because we were using existing theory and measures, we examined
indicators of an organization’s general attitude towards innovation and indicators of its
general subjective norms. Most of the previous research examining components of these
Innovation Theory of Planned Behaviour 11
constructs has been focused on specific innovations (e.g., Nystrom, et al., 2002). It could be
that both attitudes and subjective norms need to be specifically related to the particular
innovation under consideration. Indeed, the original Theory of Planned Behavior works best
when the three factors (attitude, subjective norms and PBC) are specific to the behaviour
under question (Azjen, 2001). Now that our initial research has provided support for the ITPB
model, future research can develop specific measures to further understand these anomalies.
Of course, the study was not without limitations. Because some of the variables tested
had emerged from qualitative research, we needed to create our own measures, and additional
validity studies would be useful for these measures. The cross- sectional, retrospective nature
of the data collection precludes statements about the causality of the relationships within the
model. Furthermore, the need to test the ITPB using existing operationalizations meant that
we examined general factors rather than factors rather than factors specifically related to the
innovation being adopted; this may have resulted in a decrease in predictive power for
subjective norms, attitudes and PBC.
Although our research supports much of the previous micro-macro literature in
innovation adoption, our work extends the field by developing and testing an integrative,
theoretically- derived framework. Our research, examining the measurement and predictive
properties of the ITPB model, has found that existing variables do map onto the ITPB and
that the model fits innovation adoption data to some extent. By finding at least some support
for the ITPB model, we believe that we have provided a clearer framework for both
academics and practitioners to understand and improve organizational innovation adoption.
Innovation Theory of Planned Behaviour 12
Ajzen, I. (1985). From intentions to actions: A theory of planned behavior. In J. K. J.
Beckmann (Ed.), Action control: From cognition to behavior (pp. 11-39): Heidelberg:
Ajzen, I. (1991). The theory of planned behavior. Organizational Behavior and Human
Decision Processes, 50, 179-211.
Bates, K. A., & Flynn, J. E. (1995). Innovation history and competitive advantage: A
resource-based view analysis of a manufacturing technology innovations. Academy of
Management Journal, Best paper proceedings 1995.
Damanpour, F. (1991). Organizational innovation: A meta-analysis of effects of determinants
and moderators. Academy of management Journal, 34(3), 555-590.
Drazin, R., & Schoonhoven, C. B. (1996). Community, population, and organization effects
on innovation: A multilevel perspective. Academy of Management Journal, 39(5),
Iacovou, C. L., Benbasat, I., & Dexter, A. S. (1995). Electronic data interchange and small
organizations: Adoption and impact of technology. Management Information Systems
Kimberly, J. R., & Evanisk, M. J. (1981). Organizational innovation: The influence of
individual, organizational, and contextual factors on hospital adoption of
technological and administrative innovations. Academy of Management Journal,
Klein, K. J., Conn, A. B., & Sorra, J. S. (2001). Implementing computerized technology: An
organization analysis. Journal of Applied Psychology, 86(5), 811-824.
Lehman, W. E. K., Greener, J. M., & Simpson, D. D. (2002). Assessing organizational
readiness for change. Journal of Substance Abuse Treatment.
Litwin, G. H., & Stringer Jr, R. A. (1968). Motivation and Organizational Climate: Boston:
Harvard University Press.
MacKenzie, S. B., Podsakoff, P. M., & Jarvis, C. (2005). The Problem of Measurement
Model Misspecification in Behavioral and Organizational Research and Some
Recommended Solutions. Journal of Applied Psychology, 90(4), 710-730.
Mehrtens, J., Cragg, P. B., & Mills, A. M. (2001). A model of Internet adoption by SMEs.
Information & Management, 39(3), 165-176.
Innovation Theory of Planned Behaviour 13 Download full-text
Meyer, A. D., & Goes, J. B. (1988). Organizational assimilation of innovations: A multilevel
contextual analysis. Academy of Management Journal, 31(4), 897-923.
Min, H., & Galle, W. P. (2003). E-Purchasing: Profiles of Adopters and Non-Adopters.
Industrial Marketing Management, 32(2), 227-233.
Nystrom, P. C., Ramamurthy, K., & Wilson, A. L. (2002). Organizational context, climate
and innovativeness: Adoption of imaging technology. Journal of Engineering and
Technology Management, 19(3,4), 221.
Rogers, E. M. (1995). Diffusion of innovations (4 ed.). New York: The Free Press.
Snyder-Halpern, R. (2001). Indicators of organizational readiness for clinical information
technology/systems innovation: A delphi study. International Journal of Medical
Informatics, 63, 179-204.
Tabak, F., & Barr, S. H. (1996). Innovation attributes and category membership: Explaining
intention to adopt technological innovations in strategic decision making contexts.
The Journal of High Technology Management Research, 9(1), 17-34.
Totterdell, P., Leach, D., Birdi, K., Clegg, C., & Wall, T. (2002). An Investigation of the
Contents and Consequences of Major Organizational Innovations. International
Journal of Innovation Management, 6(4), 343.
Wejnert, B. (2002). Integrating models of diffusion of innovations: A conceptual framework.
Annual Review of Sociology, 28, 297-326.