Professional Documents
Culture Documents
Vaasan ammattikorkeakoulu, University of Applied Sciences, Raastuvankatu 31-33, FI-65100 Vaasa, Finland
Hanken School of Economics, Handelsesplanaden 2, FI-65100 Vaasa, Finland
a r t i c l e
i n f o
Keywords:
Management control systems
Exploitation
Exploration
Innovation project
Performance
Tension
a b s t r a c t
This study investigates the indirect effects of mechanistic and organic types of control on
project performance acting through innovativeness in exploratory and exploitative innovation projects. It also examines the interaction effect of these controls on performance. The
research model is empirically tested with survey data from 119 projects in various project
organizations, using Partial Least Squares (PLS) with controls for the size of the project and
task uncertainty. The results illustrate that organic control, acting through innovativeness
on project performance is an important form of control in exploratory innovations, and also
enhances performance in exploitative innovations. In addition, the results indicate that the
interaction effect of organic and mechanistic control types enhances performance in both
exploratory and exploitative innovation projects, suggesting a complementary effect. The
ndings are discussed in relation to theory and their managerial implications.
2013 Elsevier Ltd. All rights reserved.
1. Introduction
Scholars have long considered innovation a major determinant of organizational long-term performance (e.g.,
Bisbe and Otley, 2004; Kanter, 2001) and an effective
management of innovation projects is a challenge facing
todays organizations (e.g., Jansen et al., 2006; Tushman
and OReilly, 1996). Empirical studies investigating the
innovationperformance relationship have also suggested
that the relationships strength is moderated by the type
of innovation (Calantone et al., 2010). As an innovation
project is the most widespread vehicle for organizing
and managing innovation activities (Chiesa et al., 2009;
Martino, 1995), this study takes exploratory and exploitative innovation projects as its unit of analysis. Exploratory
(radical) innovations cause fundamental, revolutionary
94
Cardinal concluded that incremental and radical innovations should not be managed differently. Conversely,
results by Jansen et al. (2006) at an organizational unit level
indicate that centralization negatively affects exploratory
innovation; formalization positively inuences exploitative innovation; and connectedness (social relations among
unit members) appears to be an important antecedent
of both exploratory and exploitative innovations. Thus,
the issue of whether exploratory and exploitative innovations require different control mechanisms remains largely
unresolved. Examining these innovations separately, but
within the same empirical study, offers a means to analyze
whether project controls differ across innovation projects.
Drawing on the classication in Chenhall (2003), this
study adopts the concepts of the mechanistic control
(MC) and organic control (OC) forms of project control
mechanisms to represent two opposing forms of control.
Mechanistic project controls rely on formal rules, standardized operating procedures and routines, whereas organic
project controls are more exible, responsive, involve
fewer rules and standardized procedures and tend to be
richer in data (Chenhall, 2003). Organic project control
as used here reects two important characteristics: (i)
informal control reecting norms of cooperation, communication and emphasis on getting things done, and (ii)
open channels of communication and free ow of information between project manager and subordinates (Burns
and Stalker, 1961).
Prior studies (Burns and Stalker, 1961) maintain that
a formal management control system (MCS) supports the
periodic execution of the same routines in organizations
where changes are small or non-existent. Empirical evidence also conrms this (e.g., Ouchi, 1979). In this regard,
mechanistic forms of project controls would appear to be of
little relevance to the innovation process associated with
high level of uncertainty. These limitations proposed for
the traditional MCS have, however, been questioned and
proved unfounded in more recent studies, as researchers
nd that these systems may be important in providing
the discipline to help manage uncertainty, and show that
there is also a need for formal MCSs in uncertain settings, such as project environments (see e.g. Abernethy and
Brownell, 1999; Bisbe and Otley, 2004; Cardinal, 2001;
Davila et al., 2009a). Furthermore, Adler and Borys (1996),
distinguishing between coercive and enabling bureaucracies, found that an MCS may be instrumental to innovation,
and Simons (1995) that an interactive systems concept can
play an explicit role in sparking innovation around strategic
uncertainties. Thus, for the most part recent empirical evidence indicates that innovation processes may gain from
the presence of an MCS.
More recent studies have also suggested that opposing
control mechanisms should be implemented simultaneously to foster innovativeness and performance (e.g.,
Chenhall and Morris, 1995; Henri, 2006; Lewis et al., 2002;
Sheremata, 2000). Despite prior studies, scholars claim
that there is little systematic evidence of potential indirect
effects or whether the effects of one form of control are
governed by the level of simultaneous reliance on another
form of control (Abernethy and Brownell, 1997; Malmi and
Brown, 2008).
Moreover, although scholars generally agree that innovation contributes to rm performance and that the
understanding of innovation and control issues requires
a unit of analysis other than the organizational level (e.g.
Davila et al., 2009b), there are few accounting studies that
have investigated the relevance of MCSs in project environments (Chenhall, 2008). In projects resembling temporary
matrix organizations that draw on resources from many
functions and are characterized by a high level of uncertainty (Tatikonda and Rosenthal, 2000a), project managers
may face issues managing the dynamics of their project
teams. That is because innovation and development require
a high degree of exibility in the structural and communication processes (Burns and Stalker, 1961; Van de
Ven, 1986) as well as efciency. Therefore, drawing on
Dougherty (1996), it is suggested that a focus on the relationships between project controls, innovativeness and
performance at the project level permits a more thorough
treatment of the particular project controls acting at this
level and will likely produce greater stability in the proposed relationships.
Therefore, the objective of this study is to examine
the effects of mechanistic and organic forms of control on project performance through innovativeness in
exploratory and exploitative innovation projects. Innovativeness or innovative accomplishments are here dened
very broadly to include any policy, structure, method or
process, product or market opportunity that the project
manager perceives to be new (Kanter, 1983; Zaltman et al.,
1973). In comparison, innovation in addition to novelty
also comprises commercialization and implementation of
accomplishments (e.g., Dewar and Dutton, 1986). Adopting the approach introduced by Gupta and Govindarajan
(1984), project performance was measured by comparing actual project performance and a priori expectations
rather than measuring it on an absolute scale. By assessing
project performance relative to targets and other projects,
the effects of strategic choice on project performance are
indirectly controlled for.
The current research develops a conceptual model and
tests it through PLS analysis on a sample of 119 projects,
divided into two sub-samples: exploitative and exploratory
settings. Previous studies (e.g., Bisbe and Otley, 2004;
Jansen et al., 2006) suggest an indirect positive effect of an
organic form of control on performance through innovativeness in exploratory and a similar effect brought about
by a mechanistic form of control in exploitative projects.
Moreover, prior research (e.g., Chenhall and Morris, 1995;
Henri, 2006; Lewis et al., 2002) indicates that performance
within different innovation projects can be enhanced by
the effects of combined use of organic and mechanistic
project control.
Although prior research on opposing control forces in
exploratory innovation settings does exist (e.g., Lewis et al.,
2002; Sheremata, 2000), empirical research reporting on
the indirect and interaction effects of opposing forms of
project control in both exploratory and exploitative innovative project settings was not found. Thus, this study
contributes to literature by extending prior research in
MCSs (Chenhall and Morris, 1995; Bisbe and Otley, 2004;
Henri, 2006; Jrgensen and Messner, 2009; Mundy, 2010)
to another level of analysisthe project level. In particular, and contrary to the ndings of Bisbe and Otley
(2004) at the organizational level, this study reports indirect effects of organic project controls on performance
via innovativeness in exploratory settings. This study also
offers empirical evidence of existing tension, resulting from
the joint use of mechanistic and organic forms of control
that inuence performance in exploratory and exploitative
innovation project settings, and thus claries how mechanistic and organic forms of control interact at the project
level. Furthermore, examining project controls separately
in exploratory and exploitative innovation projects makes
it possible to study possible differences and thus enhances
our understanding of the role of MCS in two different innovation settings.
The remainder of this paper is structured as follows:
Section 2 provides a brief overview of extant literature,
resulting in the formulation of the hypotheses. Section 3
presents the design of the empirical survey study conducted to collect data. Section 4 reports the tests of the
hypothesis. Sections 5 and 6 conclude the paper with a discussion of the ndings and their implications as well as
limitations and directions for future research.
2. Literature review and development of the
hypotheses
The increasing number of innovation projects in todays
business has highlighted the need for research on the applicability of MCSs for managing innovations (e.g., Davila,
2000; Davila et al., 2009b). Such systems can facilitate coordinating and controlling the project process and
stimulating dialog and idea generation to enhance innovativeness and project performance (Davila et al., 2009b;
Dougherty, 1996; Lewis et al., 2002). Scholars emphasize that exploitative and exploratory innovations may
need different management (Rogers, 1995; Van de Ven
et al., 1999) as they require different degrees of change,
most likely stemming from a different mix of environmental, organizational, managerial, and structural forces.
For example, Van de Ven et al. (1999) argue that different degrees of novelty need to be managed differently and
maintain that structural variables decreasing the degree
of radical (exploratory) innovation may simultaneously
increase the degree of incremental (exploitative) innovation. Furthermore, both exploratory and exploitative
innovations generate knowledge, but for different purposes, and the extent of new knowledge differ (Un, 2010).
Dewar and Dutton (1986) argue that exploratory (radical) innovations contain a high degree of new knowledge,
whereas exploitative (incremental) innovations build on
a low degree of new knowledge. Moreover, especially
in the early stages, radical change creates a high degree
of uncertainty in the exploratory project as well as the
whole organization, whereas the level of uncertainty is
regarded as much lower in incremental (exploitative) innovations (e.g., Chiesa et al., 2009). Information processing
theory suggests that an increasing level of uncertainty will
increase the need for information and coordination of the
project efforts (Olson et al., 1995). It has been suggested
that MCSs are effective tools for managing uncertainty,
95
96
Innovation type
(Degree of newness)
Innovativeness
Control variables
Project performance
MC x OC
Regarding the direct effects of OC on project performance, prior research provides limited empirical evidence
that may be divided into two views. The rst view maintains that OC, implying an increasing ow of information,
higher transfer of knowledge as well as shared interpretations of project goals and the conduction of project tasks
(Ayers et al., 1997; Turner and Makhija, 2006), will provide
project members with the exibility and opportunity to
adjust their outcome so as to increase the market value
of the product or service provided (Rijsdijk and van den
Ende, 2011). The second view argues for a negative effect
on project performance, as OC is seen to lead to unfocused efforts among project members, resulting in the loss
of innovative design features, which would hamper the
success of a new product or service in particular (Ayers
et al., 1997). Regarding the direct effects of MC on project
performance, prior studies have also found contradictory
results. For example, one line of research (Dvir et al., 2003;
Shenhar et al., 2002) has found that a high level of formality and detailed project planning regarding schedules,
responsibilities, budgets and goals is especially important
for complex projects involving high uncertainty and will
increase project performance, whereas Song and MontoyaWeiss (1998) found that detailed project planning could
have negative effects on the outcome of highly innovative
projects. Further, Tatikonda and Montoya-Weiss (2001)
argue that the positive impact on project operational outcomes is evident irrespective of the technical uncertainty
faced by the product or service.
While the research model contemplates the possibility
that the use of some form of OC and MC might directly
inuence project performance, no formulation of hypotheses for the potential direct effects was developed, as neither
prior research evidence nor the theoretical development
would provide clear arguments for a potential direct effect.
Consistent with reasoning by Bisbe and Otley (2004), the
potential direct effect is also expected to be relatively small
and the major proportion of the potential relationship
between the use of an MCS and performance is proposed
97
empirical evidence of the need for MCs in uncertain situations such as project environments (e.g., Cardinal, 2001;
Davila et al., 2009a).
Contemporary research on the effects of MC on innovation has indicated a positive effect of mechanistic
organizational structures (Calantone et al., 2010; Jansen
et al., 2006). Jansen et al. maintain that rules and procedures might not be as detrimental to innovation efforts
as was previously assumed. In exploitative innovations,
MC can facilitate the generation of proposals to improve
existing routines; make existing knowledge and skills
more explicit, and diminish possible variance through
improvements in existing processes and outcomes (Benner
and Tushman, 2003). Furthermore, MC can be used to
codify best practice, making it easier to adopt more efciently (Zander and Kogut, 1995). Chiesa et al. (2009)
reported that managers in exploitative projects, less
affected by uncertainty, relied more on formalized control systems throughout the project compared to managers
in exploratory projects. In summary, high levels of MC are
likely to produce inertial forces and a focus on exploitation
(e.g., Cardinal, 2001) as well as aid planning and restrain
excessive innovation (Chenhall, 2003). Therefore, drawing
on the above ndings, a positive effect of MC on innovation
in exploitative innovation is proposed. The above analysis
generates the following hypothesis:
H2a. There is a positive effect of MC on innovativeness in
exploitative innovations.
Regarding the relationship between innovativeness
and performance in exploitative environments, scholars
have suggested that exploitative innovations engender
less uncertainty and build on rm synergies and existing
knowledge, through which performance is generated (e.g.,
Calantone et al., 2010). In addition, although the returns
from exploitative innovations may be considered shortterm, they are typically positive and more predictable
than those from exploratory innovations (Menguc and
Auh, 2008). Accordingly, this study also hypothesizes that
exploitative innovations will have a positive effect on
project performance, as formally stated in the following
hypothesis:
H2b. There is a positive effect of innovativeness on project
performance in exploitative innovations.
To sum up, if the use of MC can be associated with innovativeness and innovativeness can be related to project
performance, then the use of MC can be expected to affect
project performance through inducing increased innovativeness. Therefore, MC is proposed to have an indirect
effect on project performance through innovativeness in
exploitative innovations.
2.3. Combination of project controls
Scholars maintain that control mechanisms may behave
as either substitution or complementary controls (Fisher,
1995). This means that control mechanisms can be effective and achieve the same desired outcome individually, in
other words, one control may be substituted for another (a
substitution effect) or two or more controls may reinforce
98
each other in pursuit of the desired outcome (complementary controls), making both controls necessary to achieve
the desired results (Widener, 2004). Milgrom and Roberts
(1995) illustrated analytically that control mechanisms can
be complementary, as an increase in the emphasis on
one control component increases the benets gained from
increasing the emphasis on another control. Investigating
controls in combination is therefore important because a
control may add to, detract from, or multiply the effects of
another. A combination can also bring about equinality, in
which different combinations of control practices produce
equivalent results (Anderson and Dekker, 2005).
While researchers agree on the general assumption
that MCSs are inter-dependent (e.g., Anderson and Dekker,
2005; Milgrom and Roberts, 1995; Widener, 2007), empirical studies have found evidence of both complementary
and substitutional activity. In the LOC framework (Simons,
1995) control of business strategy is achieved by integrating the four levers of beliefs systems, boundary systems,
diagnostic control systems, and interactive control systems. Simons suggests that these four levers create tension
as follows: the beliefs and interactive control system create
positive energy, and the boundary and diagnostic control systems create negative energy. He also shows how
opposing forces such as diagnostic and interactive control systems may complement each other over time in the
implementation of strategic change, and balance inherent
organizational tension. Widener (2007) provides empirical evidence on the relations among the various control
systems in the LOC framework and nds complementary
effects. She reports that when enterprises emphasize their
belief system, they also emphasize each of the three other
control systems. Henri (2006) nds empirical evidence for
managers using performance measures in both a diagnostic
and interactive way and of that use resulting in a desirable
state of dynamic tension enhancing organizational capabilities and performance.
Drawing on the rationale of positive and negative
energy (Henri, 2006; Simons, 1995), the organic form of
control, stimulating the generation of new ideas, represents a positive force when OC is used to expand
opportunity seeking and learning within and throughout
the project. It represents a negative force when used to
mold team members actions to predetermined goals and
curb what is deemed excessive innovation activity. Likewise, the mechanistic form of project control, including
monitoring, coordinating and focusing on pre-set goals and
correcting performance deviations from those goals, may
represent both positive and negative forces. The focus on
deviations may provide a negative or positive feedback
signal to the project manager, and the possible corrective
actions undertaken will be either positive or negative, i.e.
the reverse of the feedback signal to correct the innovation process. Thus, OC and MC have different purposes,
but could be used simultaneously in the project environment to manage inherent tension within the project. This
is exemplied as follows.
While MCs, providing formal rules and regulations for
project team members, are useful in assisting planning and
curbing excessive innovation (Chenhall, 2003), team members might increasingly come to view those same rules
99
nal response rate of approximately 50%. The respondents were asked to classify their project type as either
exploratory or exploitative. Of the projects, 68 represented
exploratory innovation projects and 51 exploitative innovation projects. The relatively high number of exploratory
innovation projects in this study may be explained by
the fact that the majority of the responding project managers worked in high-tech industries. A comparison of
means on all the measured variables was undertaken to
test for potential non-response bias by comparing the
mean responses of the questionnaires received prior to
the reminder email to those received after the reminder
email. A two-sample t-test revealed no signicant differences, indicating that non-response bias is unlikely to affect
the results.
The project organizations in this study represent a variety of industries including information technology (28
projects), telecommunications (20 projects), consulting (11
projects), automation (9 projects), electronics (9 projects),
metal industry (7 projects), construction (7 projects),
engineering (6 projects), banking and nancial services
(6 projects) and miscellaneous (11 projects). Descriptive statistics indicate the following data for explorative
innovation projects and exploitative innovation projects
respectively: the average previous project experience of
the project manager was 8.61 years and 8.82 years; the
average project budget was EUR 1.11 million for the explorative projects and EUR 1.47 million for the exploitative;
the average project duration was 15.21 months and 11.25
months; and nally, the average number of employees
working on an explorative project was 14 as against 10
employees for an exploitative one.
Variables were measured using multiple indicators and
operationalized through multi-item constructs on Likerttype scales. Established and reliable scales for measuring
were used, with a modication to t the present research
context. To check on the relevance of these measures,
the web-questionnaire was pre-tested on ve academics
and four project managers with expertise in the areas of
management control and project management. Final measures were then developed and rened. The measurement
instrument is provided in Appendix A and the descriptive
statistics in Table 1.
Degree of innovation newness (Innovation type) was
measured using a dummy variable where different degrees
of newness were condensed into the following two denitions: exploratory (radical) innovations aim to produce
fundamental changes in the products, services, processes
or activities of the organization and represent clear departures from existing practice; exploitative (incremental)
innovations are projects that aim to produce only minor
changes and thus require less departure from existing practices (Dewar and Dutton, 1986; Ettlie et al., 1984).
Organic form of control (OC) was operationalized as a
three-item construct modied from the instruments developed by Van der Stede (2001) and Chenhall and Morris
(1995). The adopted items are considered to capture the
communicative and interactive project control mechanisms and expected to measure the level of use of informal
and face-to-face meetings as well as the free ow of information as one form of OC. The items were measured on a
100
Table 1
Descriptive statistics on research variables.
N = 119
Theoretical range
Min
Max
Median
Mean
SD
Organic control
Mechanistic control
Innovativeness
Project performance
3
4
3
3
17
17
17
17
1.00
1.00
1.00
1.00
7.00
7.00
7.00
7.00
6.00
5.00
5.00
5.00
5.55
4.45
4.44
5.06
1.53
1.73
1.23
1.14
Exploratory (n = 68)
Minimum
Maximum
Median
Mean
SD
1.00
1.00
1.00
1.00
1.00
3.00
2.00
1.00
7.00
7.00
7.00
7.00
49.00
50.00
80.00
7.00
6.00
5.00
5.00
5.00
25.00
12.00
8.00
4.00
5.69
4.40
4.67
5.16
25.65
15.21
13.65
3.78
1.54
1.68
1.22
1.19
12.67
11.00
15.89
1.40
Exploitative (n = 51)
Organic control
Mechanistic control
Innovativeness
Project performance
Combination (OC MC)
Project duration
No. of people
Task uncertainty (a)
1.00
1.00
1.00
2.00
1.00
2.00
3.00
1.00
7.00
7.00
7.00
7.00
49.00
42.00
50.00
7.00
6.00
5.00
4.00
5.00
24.00
9.00
7.00
3.00
5.37
4.50
4.13
4.94
24.70
11.25
10.33
3.53
1.50
1.82
1.17
1.07
12.84
8.22
8.72
1.31
seven-point Likert scale, where 1 reected strongly disagree and 7 strongly agree.
The mechanistic form of control (MC) was operationalized
as a four-item variable, measuring the extent to which the
project manager has a detailed interest in specic project
performance line-items in an evaluation and does not tolerate deviations from interim project performance targets.
The instrument was developed based on previous research
by Van der Stede (2001). The items were scored on a sevenpoint Likert scale with 1 reecting strongly disagree and
7, strongly agree.
Innovativeness (INN) was operationalized as a threeitem variable and measured using a modication of
the job-performance instrument of Shields et al. (2000).
Respondents were provided with the following denition of innovativeness: Innovative accomplishments are
dened here very broadly to include any policy, structure,
method or process, product or market opportunity that
you as the manager of the project perceived to be new,
and asked to indicate the level of the project team innovativeness (the number of innovations) on a seven-point
Likert scale ranging from 1 indicating extremely low to 7
reecting extremely high.
Performance of development project (PERF) was measured using subjective self-assessment of a projects
performance relative to predetermined targets and other
similar projects. Performance was operationalized as a
multi-item construct capturing the most important dimensions relative to project performance targets (standards),
other project teams and overall assessment. These items
are a modication of the job-performance construct of
Shields et al. (2000). The level of performance of the
project team was scored on a seven-point Likert scale
ranging from 1 indicating extremely low to 7 reecting
4. Results
The model was estimated using a structural equation
modeling approach. The path analytic modeling technique of PLS was chosen because of the predictive nature
of this study (Jreskog and Wold, 1982) and because it
is a method that can provide unbiased estimates with
small sample sizes (Falk and Miller, 1992). The Smart
PLS software (Ringle et al., 2005) was used in this study.
In accordance with common procedure, the model was
analyzed and interpreted in two stages (Barclay et al.,
1995; Hulland, 1999): the rst involved the assessment
of the reliability and validity of the measurement model
and the second, the assessment of the structural model.
This was done to ensure that the constructs measures
were reliable and valid before assessing the nature of
the relations between the constructs (Barclay et al., 1995;
Hair et al., 1998; Hulland, 1999). For the measurement
model, each construct was modeled to be reective. As the
reective measurement model requires, the indicators of
each construct are highly inter-correlated (test of crossloadingsTable 4) and the indicators share a common
theme, which makes them interchangeable (Chin, 1998).
Causality is thus assumed to ow from each construct to
its indicators. All items in the analysis are univariate normal, but some data were missing at the item level. As the
amount of missing data was very small and the data missing completely at random, the missing data were replaced
using the mean imputation approach (Hair et al., 1998).2
The PLS analysis of the current research contains multiplicative interaction terms, which are developed following
the procedure outlined in Chin et al. (2003). Because interaction terms increase the potential for multicollinearity,
all items reecting the predictor and moderator constructs
were standardized (m = 0; s2 = 1) (Aiken and West, 1991;
Chin et al., 2003; Cronbach, 1987; Drazin and Van de Ven,
1985; Venkatraman, 1989). Doing so minimizes the degree
of multicollinearity among the variables and improves the
interpretability of the results. We tested for interaction
effects using the product indicators procedure suggested
by Chin et al. (2003). The statistical power of PLS in analyzing interaction effects with a product indicator approach
has been conrmed using Monte Carlo simulation (Chin
et al., 2003; Goodhue et al., 2007), and a recent study by
Goodhue et al. (2007) suggests that signicant interaction
results revealed using the PLS product indicator approach
are reliable.
101
2
An examination of means, standard deviations, and correlations
before and after the replacement revealed only minor differences. Further, the proposed research model maintains high robustness even after
observations with missing variables were dropped.
102
Table 2
Estimation of the measurement model parameters.
Original sample
Sample mean
Standard deviation
t-statistic
0.801
0.76
0.561
0.781
0.747
0.54
0.102
0.105
0.156
7.814
7.253
3.607
0.775
0.839
0.672
0.705
0.710
0.775
0.616
0.656
0.210
0.154
0.182
0.166
3.696
5.447
3.700
4.260
Innovativeness
Inn1
Inn2
Inn3
0.872
0.754
0.930
0.873
0.742
0.932
0.032
0.074
0.013
27.017
10.242
71.082
Project performance
Perf1
Perf2
Perf3
0.874
0.836
0.874
0.879
0.833
0.873
0.026
0.043
0.070
33.753
19.360
12.533
Control variable
TU1
TU2
TU3
0.710
0.793
0.755
0.693
0.773
0.733
0.136
0.117
0.133
5.225
6.745
5.672
0.538
0.719
0.565
0.473
0.668
0.544
0.231
0.248
0.272
2.324
2.901
2.078
0.523
0.800
0.754
0.765
0.536
0.760
0.737
0.718
0.163
0.127
0.122
0.125
3.204
6.280
6.175
6.124
Innovativeness
Inn1
Inn2
Inn3
0.615
0.898
0.966
0.666
0.827
0.885
0.223
0.146
0.167
2.761
6.140
5.770
Project performance
Perf1
Perf2
Perf3
0.926
0.821
0.934
0.924
0.825
0.932
0.014
0.042
0.012
65.866
19.516
76.626
Control variables
TU1
TU2
TU3
0.789
0.909
0.714
0.777
0.903
0.712
0.085
0.044
0.098
9.299
20.590
7.304
AVE
CR
0.511
0.754
0.563
0.836
0.731
0.890
0.742
0.896
0.567
0.797
0.375
0.639
0.517
0.807
0.706
0.875
0.801
0.923
0.653
0.848
Loadings of indicators on latent constructs (original sample), 0.7 or above indicates good indicator reliability.
AVE: average variance extracted, 0.5 or above indicates good convergent reliability.
CR: composite reliability, 0.7 or above indicates good convergent reliability.
(2003) and Williams et al. (2003), a common method factor was included in the PLS model. The indicators of all
constructs were associated reectively with the method
factor. Next, each indicator variance explained by the principle construct and by the method factor was computed.3
As shown in Appendix B (Table B1), the results demonstrate that the average substantively explained variance of
the indicators was 0.669, while the average method based
variance was 0.006. The ratio of substantive variance to
3
This study employed the analytical procedure previously used by
Liang et al. (2007) and Hsieh et al. (2008).
103
Table 3
Discriminant validity coefcients.
Exploratory projects (n = 68)
OC
MC
Inn.
Perf.
Duration
No.of people
TU
0.715
0.493**
0.357**
0.220
0.131
0.096
0.054
0.751
0.093
0.243
0.217
0.172
0.256
0.855
0.429**
0.061
0.032
0.041
0.862
0.020
0.053
0.331**
N/A
0.304*
0.196
N/A
0.203
0.753
0.612
0.437**
0.157
0.353*
0.003
0.221
0.141
0.719
0.087
0.276
0.023
0.336**
0.391**
0.841
0.268
0.129
0.036
0.014
0.895
0.215
0.158
0.422**
N/A
0.157
0.049
N/A
0.361**
0.808
The square root of the AVE value for each of the constructs along the diagonal (in bold).
Correlations between different constructs in the lower leftoff-diagonal elements of the matrix.
*
p < 0.050 (two-tailed test).
**
p < 0.010 (two-tailed test).
Table 4
Cross-loadings.
Exploratory projects
OC
MC
Inn
Perf.
Duration
No. people
TU
0.801
0.760
0.561
0.318
0.404
0.224
0.442
0.225
0.214
0.424
0.224
0.180
0.162
0.131
0.097
0.073
0.039
0.110
0.530
0.204
0.287
0.775
0.839
0.672
0.705
0.061
0.031
0.124
0.269
0.161
0.193
0.217
0.172
0.162
0.182
0.234
0.348
0.208
0.173
0.064
0.152
0.014
0.098
0.872
0.754
0.930
0.313
0.474
0.325
0.061
0.032
0.097
0.097
0.128
0.156
0.253
0.009
0.210
0.219
0.029
0.152
0.354
0.295
0.430
0.874
0.836
0.874
0.020
0.053
0.199
0.290
0.246
0.085
0.114
0.094
0.181
0.201
0.215
0.101
0.006
0.053
0.086
0.017
0.042
0.077
1.000
0.304
0.051
0.196
0.253
0.024
0.166
0.123
0.037
0.189
0.146
0.216
0.017
0.036
0.030
0.013
0.058
0.069
0.304
1.000
0.108
0.123
0.228
0.088
0.026
0.055
0.334
0.213
0.029
0.076
0.013
0.036
0.050
0.378
0.128
0.345
0.196
0.203
0.710
0.793
0.755
Exploitative projects
0.538
OC1
0.719
OC2
0.565
OC3
0.138
MC1
0.413
MC2
0.166
MC3
0.456
MC4
0.088
Inn1
0.102
Inn2
Inn3
0.175
0.316
Perf1
0.284
Perf2
0.346
Perf3
0.003
Duration
People
0.221
0.154
TU1
0.088
TU2
0.110
TU3
0.351
0.200
0.316
0.523
0.800
0.754
0.765
0.110
0.106
0.075
0.264
0.221
0.254
0.024
0.336
0.322
0.313
0.318
0.154
0.085
0.087
0.054
0.118
0.008
0.109
0.615
0.898
0.966
0.313
0.143
0.239
0.129
0.036
0.301
0.390
0.324
0.060
0.265
0.247
0.121
0.239
0.214
0.194
0.021
0.177
0.305
0.926
0.821
0.934
0.215
0.158
0.097
0.068
0.047
0.130
0.071
0.022
0.216
0.065
0.026
0.059
0.096
0.174
0.089
0.238
0.061
0.244
1.000
0.157
0.051
0.104
0.136
0.085
0.045
0.260
0.350
0.276
0.259
0.163
0.071
0.028
0.041
0.078
0.246
0.134
0.157
1.000
0.288
0.314
0.272
0.065
0.009
0.182
0.151
0.364
0.366
0.194
0.200
0.046
0.014
0.395
0.352
0.386
0.049
0.361
0.789
0.909
0.714
OC1
OC2
OC3
MC1
MC2
MC3
MC4
Inn1
Inn2
Inn3
Perf1
Perf2
Perf3
Duration
People
TU1
TU2
TU3
Discriminant validity on item level indicated as higher loadings on associated constructs compared to cross-loadings.
104
Table 5
PLS structural model results: path coefcients, t statistics, R2 and Q2 .
Exogenous
variables
Organic control
(OC) Perf
Organic control
(OC) INN
Mechanistic
control
(MC) Perf
Mechanistic
control
(MC) INN
Innovativeness
(INN) Perf
OC MC Perf
Control variables:
Project size:
duration
Project size: no. of
people
Task uncertainty
R2 (Perf)
R2 (INN)
Q2 (Perf)
Q2 (INN)
Exploratory
Exploitative
Stage I
Stage II
Stage I
Stage II
0.004
(0.050)
0.411
(4.862)***
0.160
(1.702)
0.030
(0.377)
0.411
(4.636)***
0.146
(1.533)
0.267
(2.445)**
0.147
(1.316)
0.031
(0.521)
0.254
(2.729)**
0.148
(1.345)
0.015
(0.256)
0.110
(1.084)
0.110
(1.140)
0.023
(0.222)
0.022
(0.213)
0.402
(4.239)***
0.413
(4.442)***
0.218
(2.716)**
0.270
(2.554)**
0.233
(2.277)**
0.341
(4.501)***
0.060
(0.909)
0.056
(0.934)
0.274
(3.147)***
0.303
0.136
0.001
0.083
0.027
(0.454)
0.053
(0.881)
0.279
(3.162)***
0.348
0.136
0.009
0.083
0.269
(4.262)***
0.009
(0.192)
0.417
(5.162)***
0.391
0.025
0.094
0.016
0.249
(4.238)***
0.003
(0.060)
0.404
(4685)***
0.505
0.025
0.092
0.016
105
Table 6
Two-way interactions of OC and MC mean scores of PERF and INN.
Performance
Exploratory projects
Exploitative projects
Mechanistic
Organic
Low
High
Innovation
Organic
Low
High
Mechanistic
Low
High
Low
High
A1
= 4.783
X
S.D. = 1.033
n = 20
B1
= 5.716
X
S.D. = 0.921
n=8
C1
= 5.429
X
S.D. = 1.134
n=7
D1
= 5.406
X
S.D. = 0.948
n = 23
A3
= 4.457
X
S.D. = 0.795
n = 14
B3
= 5.143
X
S.D. = 0.766
n=7
C3
= 5.222
X
S.D. = 1.328
n=6
D3
= 5.266
X
S.D. = 0.807
n = 14
A2
= 4.050
X
S.D. = 1.078
n = 20
B2
= 5.517
X
S.D. = 0.696
n=8
C2
= 4.429
X
S.D. = 0.789
n=7
D2
= 4.942
X
S.D. = 0.925
n = 23
A4
= 4.156
X
S.D. = 0.935
n = 14
B4
= 4.286
X
S.D. = 1.177
n=7
C4
= 3.389
X
S.D. = 1.405
n=6
D4
= 4.043
X
S.D. = 0.725
n = 14
106
Combined effect
MC x OC
0.411***/0.147
-0.004/0.267**
0.218**/0.341***
0.402***/0.270**
Innovativeness
-0.110/0.023
Project performance
0.160*/-0.031
5. Discussion
Overall, the results of this study emphasize the importance of OC being the main form of control in both
exploratory and exploitative innovations. The effect of OC
is however somewhat different, as in exploratory innovation projects the focus of OC appears to be on enhancing
project performance through innovativeness, whereas in
exploitative innovations, OC drives project performance.
The ndings also indicate that MC is not without value,
not when used individually, but rather if used in combination with OC to enhance project performance. MC appears
not to drive innovativeness, but instead to be associated
with performance mainly in exploratory innovation settings. Further, MC appears to interact with organic forms
of control to enhance project performance in both innovation contexts. The ndings are shown in Fig. 2 and will be
discussed in more detail below.
First, although not hypothesized, the direct effects were
tested and analyzed. In exploitative settings the effect of OC
is manifest on project performance, whereas in exploratory
innovations OC works best indirectly through innovativeness. This may be explained by the differences between
the two innovation types, related to the different need for
information in exploratory and exploitative innovations
deriving from a different level of novelty and uncertainty as
well as a different need for new knowledge. The importance
of OC in project environments in this study is consistent with the ndings of previous studies by Jansen et al.
(2006) conducted on the organizational level. There, pursuing exploratory innovation was more effective in dynamic
environments, and centralization was found to have a negative effect on a units exploratory efforts.
Furthermore, while this study registers a slightly
positive direct effect of MC on performance in exploratory
projects, overall MC appears not to make much of a direct
and individual contribution to innovativeness or project
performance. The insignicance of MC in innovation
projects has also been suggested in early research (e.g.,
Ouchi, 1977). It therefore appears as if mere mechanistic
107
108
participants of the MONFORMA conference in Melbourne in November, 2010 for their very insightful and
helpful comments.
109
Table B1
Common method bias analysis.
Item
R12
R22
0.714**
0.639**
0.895**
0.727**
0.700**
0.823**
0.521**
0.817**
0.866**
0.929**
0.910**
0.829**
0.858**
0.789**
0.841**
0.699**
1.000
1.000
0.801
0.510
0.408
0.801
0.529
0.490
0.677
0.271
0.667
0.750
0.863
0.828
0.687
0.736
0.623
0.707
0.489
1.000
1.000
0.669
0.034
0.180*
0.162
0.017
0.068
0.083
0.027
0.074
0.034
0.040
0.107
0.017
0.111
0.001
0.043
0.049
0.000
0.000
0.002
0.001
0.032
0.026
0.000
0.004
0.007
0.001
0.005
0.001
0.002
0.011
0.000
0.012
0.000
0.002
0.002
0.000
0.000
0.006
*
**
p < 0.05.
p < 0.01.
project
team
References
Abernethy, M.A., Brownell, P., 1997. Management control systems in
research and development organizations: the role of accounting,
behavior and personnel controls. Accounting, Organization and Society 22 (34), 233248.
Abernethy, M.A., Brownell, P., 1999. The role of budgets in organizations
facing strategic change: an exploratory study. Accounting, Organizations and Society 24 (3), 189204.
Adler, P.S., Borys, B., 1996. Two types of bureaucracy: enabling and coercive. Administrative Science Quarterly 41 (1), 6189.
Ahrens, T., Chapman, C.S., 2004. Accounting for exibility and efciency:
a eld study of management control systems in a restaurant chain.
Contemporary Accounting Research 21 (2), 271301.
Aiken, L.S., West, S.G., 1991. Multiple Regression: Testing and Interpreting
Interactions. Sage, Newbury Park, CA.
Amabile, T.M., Barsade, S.G., Mueller, J.S., Staw, B.M., 2005. Affect and
creativity at work. Administrative Science Quarterly 50 (3), 367403.
Ancona, D.G., Caldwell, D.F., 1992. Demography and design: predictors of new product team performance. Organization Science 3 (3),
321341.
Anderson, S.W., Dekker, H.C., 2005. Management control for market transactions: the relation between transaction characteristics, incomplete
contract design and subsequent performance. Management Science
51 (12), 17341753.
Argyris, C., Schn, D., 1983. Organizational Learning. Addison-Wesley, MA.
Ayers, D., Dahlstrom, R., Skinner, S.J., 1997. An exploratory investigation
of organizational antecedents to new product success. Journal of Marketing Research 34 (1), 107116.
Barclay, D., Higgins, C., Thompson, R., 1995. The Partial Least Squares (PLS)
approach to causal modeling: personal computer adoption and use as
an illustration. Technology Studies 2 (2), 285324.
Baron, R.M., Kenny, D.A., 1986. The moderator-mediator variable distinction in social psychological research: conceptual, strategic and
statistical considerations. Journal of Personality and Social Psychology
51 (6), 11731182.
Benner, M.J., Tushman, M.L., 2003. Exploitation, exploration, and process
management: the productivity dilemma revisited. Academy of Management Review 28 (2), 238256.
R., 2012. Using strategic performance measurement
Bisbe, J., Malagueno,
systems for strategy formulation: does it work in dynamic environments? Management Accounting Research 23 (4), 296311.
Bisbe, J., Otley, D., 2004. The effects of the interactive use of management
control systems on product innovation. Accounting, Organizations
and Society 29 (8), 709737.
Bommer, W.H., Johnson, J.L., Rich, G.A., Podsakoff, P.M., MacKenzie, S.B.,
1995. On the interchange ability of objective and subjective measures
of employee performance. Personnel Psychology 48, 587605.
110
Brownell, P., 1995. Research Methods in Management Accounting. Coopers and Lybrand.
Burns, T., Stalker, G.M., 1961. The Management of Innovation. Tavistock,
London.
Calantone, R.J., Harmancioglu, N., Droge, C., 2010. Inconclusive innovation
returns: a meta-analysis of research on innovation in new product development. Journal of Product Innovation Management 27 (7),
10651081.
Cardinal, L., 2001. Technological innovation in the pharmaceutical industry: the use of organizational control in managing research and
development. Organization Science 12 (1), 1936.
Chenhall, R.H., 2003. Management control systems design within its organizational context: ndings from contingency-based research and
directions for the future. Accounting, Organizations and Society 28
(23), 127168.
Chenhall, R.H., 2008. Accounting for the horizontal organization: a review
essay. Accounting, Organizations and Society 33 (45), 517550.
Chenhall, R.H., Morris, D., 1995. Organic decision and communication processes and management accounting systems in entrepreneurial and
conservative business organizations. Omega: International Journal of
Management Science 23 (5), 485497.
Chiesa, V., Frattini, F., Lamberti, L., Noci, G., 2009. Exploring management
control in radical innovation projects. European Journal of Innovation
Management 12 (4), 416443.
Chin, W.W., 1998. The partial least square approach for structural equation
modeling. In: Marcoulides, G.A. (Ed.), Modern Methods for Business
Research. Lawrence Erlbaum Associates.
Chin, W.W., Marcolin, B.I., Newsted, P.R., 2003. A partial least squares
latent variable modeling approach for measuring interaction effects:
results from a Monte Carlo simulation study and an electronicmail
emotion/adoption study. Information Systems Research 14 (2),
189217.
Cohen, J., 1988. Statistical Power Analysis for the Behavioral Sciences.
Lawrence Erlbaum Associates, Hillsdale, NJ.
Cohen, W.M., Levinthal, D.A., 1990. Absorptive capacity: a new perspective
on learning and innovation. Administrative Science Quarterly 35 (1),
128152.
Cronbach, L.J., 1987. Statistical tests for moderator variables: aws in
analyses recently proposed. Psychological Bulletin 102 (3), 414
417.
Damanpour, F., 1991. Organizational innovation: a meta-analysis of
effects of determinants and moderators. Academy of Management
Journal 34 (3), 555590.
Davila, T., 2000. An empirical study on the drivers of management control
systems design in new product development. Accounting, Organizations and Society 25 (45), 383409.
Davila, A., Foster, G., Li, M., 2009a. Reasons for management control systems adoption: insights from product development systems choice
by early-stage entrepreneurial companies. Accounting, Organizations
and Society 34, 322347.
Davila, A., Foster, G., Oyon, D., 2009b. Accounting and control,
entrepreneurship and innovation: venturing into new research opportunities. European Accounting Review 18 (2), 281311.
De Dreu, C.K.W., Weingart, L.R., 2003. Task versus relationship conict and
team effectiveness: a meta-analysis. Journal of Applied Psychology 88
(4), 741749.
Dent, J.F., 1990. Strategy, organization and control: some possibilities for
accounting research. Accounting, Organizations and Society 15 (12),
325.
Dewar, R.D., Dutton, J.E., 1986. The adoption of radical and incremental innovations: an empirical analysis. Management Science 32 (11),
14221433.
Dougherty, D., 1992. Interpretive barriers to successful product innovation
in large rms. Organization Science 3 (2), 179202.
Dougherty, D., 1996. Organizing for innovation. In: Clegg, S.R., Hardy, C.,
Nord, W.R. (Eds.), Handbook of Organization Studies. Sage, Thousand
Oaks, CA, pp. 424439.
Dougherty, D., 2006. Organizing for innovation in the 21st century. In:
Clegg, S.R., Hardy, C., Lawrence, T.B., Nord, W.R. (Eds.), The Sage Handbook of Organization Studies. Sage Publications, London, pp. 598
617.
Drazin, R., Van de Ven, A.H., 1985. Alternative forms of t in contingency
theory. Administrative Science Quarterly 30 (4), 514539.
Droge, C., Calantone, R., Harmancioglu, N., 2008. New product success: is
it really controllable by managers in highly turbulent environments?
Journal of Product Innovation Management 25 (3), 272286.
Dvir, D., Raz, T., Shenhar, A.J., 2003. An empirical analysis of the relationship between project planning and project success. International
Journal of Project Management 21, 8995.
Ettlie, J.E., Bridges, W.P., OKeefe, R.D., 1984. Organization strategy and
structural differences for radical versus incremental innovations.
Management Science 30 (6), 682695.
Falk, R.F., Miller, N.B., 1992. A Primer for Soft Modeling. The University of
Akron Press, Akron, OH.
Fisher, J.G., 1995. Contingency-based research on management control
systems: categorization by level of complexity. Journal of Accounting
Literature 14, 2453.
Fornell, C., Larcker, D.F., 1981. Evaluating structural equation models with
unobservable variables and measurement error. Journal of Marketing
Research 18 (1), 3950.
Furnham, A., Stringeld, P., 1994. Congruence of self and subordinate
ratings of managerial practices as a correlate of superior evaluation.
Journal of Occupational and Organizational Psychology 67, 5767.
Gefen, D., Straub, D., 2005. A practical guide to factorial validity using
PLS-Graph: tutorial and annotated example. Communications of the
Association for Information Systems 16 (5), 91109.
Geisser, S., 1974. A predictive approach to the random effect model.
Biometrika 61 (1), 101107.
Goodhue, D., Lewis, W., Thompson, R., 2007. Statistical power in analyzing interaction effects: questioning the advantage of PLS with
product indicators. Information Systems Research 18 (2), 211
227.
Gupta, A.K., Govindarajan, V., 1984. Business unit strategy, managerial
characteristics, and business unit effectiveness at strategy implementation. Academy of Management Journal 27, 2541.
Hall, M., 2011. Do comprehensive performance measurement systems
help or hinder managers mental model development? Management
Accounting Research 22 (2), 6883.
Hair, J.F., Anderson, R.E., Tatham, R.L., Black, W.C., 1998. Multivariate Data
Analysis. Prentice-Hall.
Henri, J.F., 2006. Management control systems and strategy: a resourcebased perspective. Accounting, Organizations and Society 31 (6),
529558.
Hill, C.W.L., Rothaermel, F.T., 2003. The performance of incumbent rms in
the face of radical technological innovation. Academy of Management
Review 28 (2), 257274.
Hsieh, P-O.J.J., Rai, A., Kell, M., 2008. Understanding digital inequality: comparing continued use behavioural models of the socioeconomically advantaged and disadvantaged. MIS Quarterly 32 (1),
97126.
Hulland, J., 1999. Use of partial least Squares (PLS) in strategic management research: a review of four recent studies. Strategic Management
Journal 20 (2), 195204.
Jansen, J.J.P., van den Bosch, F.A.J., Volberda, H.W., 2006. Exploratory
innovation, exploitative innovation, and performance: effects of organizational antecedents and environmental moderators. Management
Science 52 (11), 16611674.
Jrgensen, B., Messner, M., 2009. Management control in new product
development: the dynamics of managing exibility and efciency.
Journal of Management Accounting Research 21 (1), 99124.
Jreskog, K.G., Wold, H., 1982. The ML and PLS techniques for modeling with latent variables: historical and comparative aspects. In:
Wold, H., Jreskog, K. (Eds.), Systems Under Indirect Observation:
Causality, Structure, Prediction (Vol. I). North-Holland, Amsterdam,
pp. 263270.
Kamm, J.B., 1987. An Integrative Approach to Managing Innovation. Lexington Books, Lexington.
Kanter, R., 1983. The Change Masters. Simon & Schuster, New York.
Kanter, R., 2001. Evolve! Succeeding in the Digital Culture of Tomorrow.
Harvard Business School Press, Boston.
Kleinschmidt, E.J., Cooper, R., 1991. The impact of product innovativeness on performance. Journal of Product Innovation Management 8,
240251.
Lewis, M.W., 2000. Exploring paradox: toward a more comprehensive
guide. Academy of Management Review 25 (4), 760776.
Lewis, M.W., Welsh, M.A., Dehler, G.E., Green, S.G., 2002. Product development tensions: exploring contrasting styles of project
management. Academy of Management Journal 45 (3), 546
564.
Liang, H., Saraf, N., Hu, Q., Xue, Y., 2007. Assimilation of enterprise systems: the effect of institutional pressures and the mediating role of
top management. MIS Quarterly 31 (1), 5987.
MacKinnon, D.P., Lockwood, C.M., Williams, J., 2004. Condence limits for
the indirect effect: distribution of the product and resampling methods. Multivariate Behavioral Research 39 (1), 99128.
Malmi, T., Brown, D.A., 2008. Management control systems as a package:
opportunities, challenges and research directions. Management
Accounting Research 19 (4), 287300.
111
112
Zahra, S.A., George, G., 2002. Absorptive capacity: a review, reconcepualisation, and extension. Academy of Management Review 27 (2),
185203.
Zaltman, G., Duncan, R., Holbek, J., 1973. Innovations and Organizations.
Wiley, New York.
Zander, U., Kogut, B., 1995. Knowledge and the speed of the transfer and
imitation of organizational capabilities: an empirical test. Organization Science 6 (1), 7692.
Zirger, B.J., Maidique, M.A., 1990. A model of new product development:
an empirical test. Management Science 36 (7), 867883.