1 s2.0 S1044500513000401 Main

Management Accounting Research 25 (2014) 93112
Contents lists available at ScienceDirect
Management Accounting Research

journal homepage: www.elsevier.com/locate/mar
The effects of organic and mechanistic control in exploratory

and exploitative innovations
Mika Ylinen a, , Benita Gullkvist b,1
a
b
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
Corresponding author. Tel.: +358 40 011 3998.

E-mail addresses: mika.ylinen@puv. (M. Ylinen),
benita.gullkvist@hanken. (B. Gullkvist).
1
Tel.: +358 (0)40 3521727.
1044-5005/$ see front matter 2013 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.mar.2013.05.001
changes in technology and represent clear departures from

existing practice (Ettlie et al., 1984) by developing new
products and services for emerging customers or markets and pursuing new knowledge. In contrast, exploitative
(incremental) innovations are other changes in products
and processes, which are generally less signicant or which
do not introduce considerable novelty (OECD, 2004) as
they extend existing products and services for existing
customers and build on existing knowledge (Benner and
Tushman, 2003).
Previous research has asserted that control mechanisms
exert differing inuences on exploratory and exploitative
innovations (e.g., Benner and Tushman, 2003; Davila et al.,
2009b; Hill and Rothaermel, 2003), but empirical studies
examining such relationships have produced mixed results
(Cardinal, 2001; Damanpour, 1991; Dewar and Dutton,
1986; Ettlie et al., 1984; Jansen et al., 2006). For example, results by Cardinal (2001) at the organizational level
show that input, behavior, and output control enhance
exploratory (radical) innovation, and input and output controls enhance exploitative (incremental) innovation, and
94
M. Ylinen, B. Gullkvist / Management Accounting Research 25 (2014) 93112
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
because they provide the information necessary to close

the information gap between the information required for
a particular task and that already available (Davila, 2000;
Tushman and Nadler, 1978). As the degree of uncertainty
will also vary with innovation type, it is likely that the need
for information as well as for coordination and control will
differ in different types of innovations (Davila, 2000; Olson
et al., 1995).
Whereas some researchers dismiss the use of formal controls as totally incompatible with innovation (e.g.,
Ouchi, 1977) there are those who advocate more organic
forms of controlmeaning a uid style of management
dealing with issues as they emergeto foster creativity
and improvisation (e.g., Dougherty, 1992). At the other end
of the scale, are those who prescribe more mechanistic
forms of control like disciplined planning as a way to focus
on and speed project efforts (e.g., Wheelwright and Clark,
1992; Zirger and Maidique, 1990). For example, Chiesa
et al. (2009) found that project managers, in response to
the greater uncertainty inherent in exploratory projects,
appeared to be more determined to adopt boundary
and interactive control mechanisms, a leaning manifested
in more frequent internal meetings. Simons framework
(1995) emphasizes the importance of the interactive use
of formal MCS in ensuring the success of innovation initiatives, but does not clearly distinguish between different,
conceptually distinct, types of potential effects of the use
of MCS, nor is such research known of in regard to project
innovation and performance. Furthermore, the more recent
ndings of Bisbe and Otley (2004) relating to an organizational level do not support the positive effect of an
interactive use of MCS on innovation. They found that
the relationship between innovation and performance was
moderated by the style of use of the MCS.
Proponents of either organic or mechanistic forms of
control have, however, also recognized the need for a
balance between the competing roles of controlling the
achievement of predetermined goals while simultaneously
enabling employees to explore alternative options to solve
problems and develop new suggestions (e.g. Ahrens and
Chapman, 2004; Chenhall and Morris, 1995; Simons, 1995).
Lewis et al. (2002) suggest that managing tensions between
exibility and discipline may provide a key to high project
performance, and argue for a blend of organic and mechanistic forms of control, or as they put it, emergent and
planned approaches. In project environments with wellknown operations, where stability and conformity are
appreciated, tension may, however, be less useful and disturb well-established organizational routines, roles and
internal processes (Henri, 2006). While there is a growing
body of empirical research on the tension between the use
of mechanistic (diagnostic) and organic forms of MCSs (e.g.,
Ahrens and Chapman, 2004; Chenhall and Morris, 1995;
Henri, 2006; Jrgensen and Messner, 2009; Marginson,
2002; Mundy, 2010; Widener, 2007), previous studies have
largely been conducted at the organizational level.
The current research aims to increase understanding of
the use of project controls in innovation environments and
to examine two potential effects of MCSs. To do so, a conceptual model is developed. The proposed research model
is illustrated in Fig. 1 and explained below.
96
Organic control (OC)
Innovation type
(Degree of newness)
Innovativeness
Mechanistic control (MC)
Control variables
Project performance
MC x OC
Direct effect ____

Moderating effect -----Fig. 1. Theoretical model. Direct effect (), moderating effect (-----).
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
to come indirectly through innovativeness rather than

through a direct effect. The direct effect is, however, controlled for in the model.
2.1. The organic form of project control in exploratory
innovations
Dent (1990) proposes that control systems can stimulate and foster curiosity and experimentation. In particular,
the more OC mechanisms build on communication and
decision processes that are participative, exible and open,
and therefore that can enhance opportunities to identify
problems or new ideas (Kamm, 1987; Morse and Lorsch,
1970). Organic forms of control are also needed to ensure
that individuals are motivated to participate in creative
decision making and to provide the free ow of ideas that is
essential for developing entrepreneurial strategies (Burns
and Stalker, 1961) or helping focus attention on strategic issues (Simons, 1995). Scholars (e.g., Montoya-Weiss
and Calantone, 1994; Olson et al., 1995) have also suggested that OC may be best suited for project environments,
because decentralization, autonomy and empowerment
may lead to conict resolution and effective decision making, whereas more formalized, mechanistic approaches
may inhibit the diffusion of ideas among project team
members.
Providing an organic form of control entails informal
communication and the free ow of information, which
also enables project team members to explore alternative
ways to solve problems (e.g., Chenhall, 2003; Wageman,
2001) and develop new suggestions (Wilson et al., 2007),
which in turn form a basis for knowledge generation and
foster innovativeness. We suggest that this is especially
important in exploratory projects, characterized by high
uncertainty and radical change. Moreover, organic forms of
project control might provide the means for project managers to indicate to team members that innovativeness is
highly valued and consistent with project aims. Therefore,
it is proposed that the extent of the use of organic project
control in exploratory innovation projects will be positively

related to innovativeness and eventually to project performance. Accordingly, the following hypothesis is proposed:
H1a. There is a positive effect of OC on innovativeness in
exploratory innovations.
In contemporary management research, empirical studies have repeatedly conrmed a positive relationship
between, for example, product innovativeness and project
performance in terms of new product success and failure (e.g., Calantone et al., 2010; Kleinschmidt and Cooper,
1991; Song and Montoya-Weiss, 1998). In addition to
its direct effect on performance, Van de Ven and Polley
(1992) emphasize that learning during the innovation process generates absorptive capacity (dened by Cohen and
Levinthal (1990) as the capability to identify, assimilate,
and apply knowledge) through which competitive advantages (Zahra and George, 2002) are generated. Although
exploratory innovations in the early stages of the innovation process, might have little or no economic impact
(Popadiuk and Choo, 2006), the effect of innovativeness
on performance has generally been regarded as positive,
and mainly long-term (e.g., Lewis, 2000; Tushman and
OReilly, 1996). Researchers (e.g., Droge et al., 2008) argue
that for example innovating rms with unique knowledge, capabilities and superior products/services should
record high performance. Likewise, it is here proposed that
exploratory projects, drawing on new ideas and developing new products, processes and so forth, will contribute
to strong project performance. The following hypothesis is
proposed:
H1b. There is a positive effect of innovativeness on project
performance in exploratory innovations.
In summary, if the organic form of project control can be
linked to innovativeness and innovativeness can be linked
to project performance, then the use of OC can be expected
to have implications for project performance through an
induced increase in innovativeness. As a result, this study
anticipates OCs acting through innovativeness having an
indirect effect on project performance in exploratory innovation.
2.2. The mechanistic form of project control in
exploitative innovations
Mechanistic control, in its various forms, provides formal rules and regulations for project team members, which
may enhance project efciency (Lewis et al., 2002; Shenhar
and Dvir, 1996). Milestones based on schedules or budgets help ensure teams are more aware of scarce resources
(Wheelwright and Clark, 1992) and guide decision making regarding resource allocation (Lewis et al., 2002). In
addition, a mechanistic form of control, particularly a more
directive one, may also dene project boundaries and thus
reduce the probability of wasteful explorations and costly
errors (McDonough and Barczak, 1991). Whereas previous
research has traditionally associated MCSs with mechanistic organizations (Burns and Stalker, 1961), and suggested
that they are of little or no use in research and development
settings (Davila et al., 2009a), more recent studies provide
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
and regulations as rigid and repressive (Dougherty, 2006;

Abernethy and Brownell, 1997). Consequently, knowledge generation will suffer (Amabile et al., 2005), which
would have a negative effect on team innovation and
project performance. Similarly, providing OC entails informal communication and free ow of information, which
would enable project team members to explore alternative
options to solve problems (e.g., Chenhall, 2003; Wageman,
2001) and develop new suggestions (Wilson et al., 2007).
Nevertheless, if only OCs are provided and at the same time
little or no mechanistic control is applied, the team members may utilize the available freedom too expansively
and the increased search for alternative opportunities may
reduce the level of coordination within the project team
(Sheremata, 2002). That in turn can lead to unfocused and
ineffectual efforts which would not promote project performance.
Thus, when considering the above separate effects of
organic and mechanistic project controls, it appears likely
that a combination of opposing forms of project control
would yield a balanced set of forces enhancing and impeding project performance. Although both OC and MC may
yield effects when applied individually, and the combined
use of OCs and MCs is likely to create either negative or
positive tension (conict, paradox) due to the different
controls being contradictory (Lewis, 2000), we consider
the simultaneous implementation of opposing OC and
MC necessary to enhance project performance, because
the combination should create a positive synergy that
enhances project performance. For example, an opposing form of control is called for to hinder the negative
effect of relying solely on OC. Mechanistic forms of control, through coordinated action plans, will have an indirect
positive effect by preventing the exploration of alternative
undesired options and ideas, something that could otherwise escalate into negative spirals (Argyris and Schn,
1983; Sundaramurthy and Lewis, 2003), and ultimately
make the project team unmanageable. These direct and
indirect effects of OC and MC create the necessary imbalance of forces that is believed to favor innovativeness, and
also make project performance more focused and effectual. Therefore, mechanistic and organic forms of control
are believed to mutually reduce the undesired side effects
of the respective opposing control mechanism and enable
the positive effects of the respective opposite strategy to
be fully realized to create synergy. For this reason, a multiplicative (rather than an additive) effect of these two types
of project control on project performance is posited in this
study.
Following previous research, which have has mostly
found complementary effects of control combinations (e.g.,
Anderson and Dekker, 2005; Widener, 2007), this is considered a likely outcome in this study, too. For example, in
exploitative innovation projects, where MC is essential for
controlling and coordinating innovation projects and keeping them on track, OC could fulll a complementary role by
assisting MC through improving exibility and enabling the
adaptability required when dealing with unforeseen events
(Turner and Makhija, 2006). Thus, the complementary role
of OC would strengthen the effect of MC and help achieve
the desired project performance outcomes. Likewise, in
exploratory projects, MC could provide additional benets

on top of OC to enhance project performance. Despite this
reasoning, the technical tests of the relationships around
the two control dimensions in this study follow an interaction approach, which will permit either relation. Thus,
what is tested is the more general proposition that the two
project controls used jointly produce a positive synergy
effect on project performance.
In summary, in this study OC and MC are believed to
mutually reduce the undesired side effects of the respective control mechanism and thus enable the positive effects
of the respective opposite strategy to be fully realized
to create synergy. Positive effects of tension might be
fully realized in project environments where simultaneous
requirements for both control mechanisms exist. This is
proposed to hold true for both exploratory and exploitative
innovations. Therefore, the third hypothesis of this study is
as follows:
H3. There is a positive interaction effect of OC and MC on
project performance in exploratory and exploitative innovations.
The research model was controlled for perceived task
uncertainty and project characteristics such as size, here
measured as the number of employees working on the
project and the duration of the project. Previous research
has consistently shown that innovation projects are characterized by considerable uncertainty and ambiguity,
and that exploratory innovations show a higher level
of uncertainty throughout the development phase than
exploitative projects (e.g., Chiesa et al., 2009). Prior studies have also used the level of uncertainty to distinguish
between exploratory and exploitative innovation projects.
Furthermore, previous research has found team size to be
related to cohesiveness and internal communication for
groups (e.g., Ancona and Caldwell, 1992). Scholars have also
found that larger groups show greater potential for team
conict (Williams and OReilly, 1998), which may result in
reduced team performance (De Dreu and Weingart, 2003).
In addition, project duration is used here as a proxy for
project size, because long-term projects will necessitate
extensive use of tangible and intangible resources.
3. Research design
The research model was tested with a stratied sample of project organizations randomly extracted from the
database of the Project Management Association of Finland.
True random sampling, requiring the a priori establishment
of the population from which a sample could be drawn,
was not possible as detailed and complete public data on
projects or project managers were not available. Thus, there
is a danger that the sample will not represent the broader
population. Data were collected by a survey questionnaire
administered to 240 project managers from 145 different
project organizations.
Project managers of innovative projects were initially
targeted with a personally addressed email providing
response instructions and guaranteeing anonymity. The
initial request was supported by one reminder email,
following which, 119 usable responses were received, a
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
No. of items used
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

Innovativeness
Project performance
Combination (OC MC)
Project duration
No. of people
Task uncertainty (a)
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
(a) Reverse coded.
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
extremely high. While the degree to which subjective

self-ratings correspond to objective performance measures is debatable, evidence from prior research has shown
that managers subjective self-ratings of performance and
objective measures are highly correlated (see e.g., Shields
et al., 2000). Similar approaches at the project level have
been used by Sheremata (2002).
The combined effect of MC and OC is operationalized as
a product term between organic and mechanistic forms
of control. Following previous research (e.g., Poppo and
Zenger, 2002; Siggelkow, 2002), the interaction effects
method for testing complementarity between control combinations and project performance was used. Prior research
suggests that there will be evidence of complementarity
between control combinations and project performance if
the coefcient of the interaction term is positive. In the context of this study, a signicant positive interaction effect
between OC and a given type of MC would suggest that
they are complementary and a negative one would suggest
that they are substitutional.
The control variable project size was operationalized
as two project characteristicsthe number of full-timeequivalent workers assigned to the project and the project
duration measured in months. The second control variable, task uncertainty was operationalized as project
task uncertainty resulting from the characteristics of the
project teams work process or set of value activities.
Task uncertainty (TU) was measured using the three items
representing task difculty (task analyzability) of the measurement instrument developed by Withey et al. (1983).
The items were anchored with to a small extent (1) and
to a great extent (7), but the scale was revised in the
analysis to reect that the highest numbers accorded with
the higher perceived TU.
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
ability. All items suggest good indicator reliability with few

indicator loadings below 0.7.
Convergent validity can be assessed by examining composite reliability (CR) and average variance extracted (AVE)
measures (Fornell and Larcker, 1981; Hulland, 1999). As
shown in Table 2, the CR values indicate good consistency
of all constructs bar one, when using the threshold of 0.7
(Nunnally, 1978). The proportion of total variance in a construct that is extracted by the component set of indicator
variables (AVE) should be greater than 0.50 (Chin, 1998;
Hair et al., 1998). This is true for all but one variable, which
is 0.38. Therefore, most measures indicate good convergent
validity.
Discriminant validity is evaluated by comparing the AVE
of each construct and the variance shared between such
constructs and other constructs in the model. Furthermore,
the cross-loadings of indicators on other constructs should
be minimal. Table 3 shows the correlations between different constructs in the lower left, off-diagonal elements of the
matrix, and the square root of the AVE value calculated for
each of the constructs along the diagonal (marked in bold).
For adequate discriminant validity, as occurs in this analysis, the diagonal elements should be signicantly greater
than the off-diagonal elements in the corresponding rows
and columns (Fornell and Larcker, 1981; Hulland, 1999).
In addition, a matrix of cross-loadings was constructed to
test discriminant validity on the item level (Table 4). All
items demonstrated higher loadings on their associated
constructs (marked in bold) compared with their crossloadings. While cross-loadings derived from this procedure
will inevitably be higher than those derived from typical exploratory factor analysis (Gefen and Straub, 2005),
the cross-loading differences were much higher than the
suggested threshold of 0.1 (Gefen and Straub, 2005). The
above assessment suggests that all scales behaved reliably
in both groups, demonstrated satisfactory convergent and
discriminant validity and exhibited adequate psychometric
properties.
In addition, the correlation analysis in Table 3 indicates a strong relationship between the two variables OC
and MC. The control variable Task Uncertainty correlates
strongly with the dependent variable PERF. A high level
of TU is considered a typical feature of the development
project environment, and has also been found to relate to
project performance in previous studies (e.g., Tatikonda
and Rosenthal, 2000b).
4.2. Common method bias
4.1. Measurement model

Table 2 illustrates the measurement model parameters
for both subgroups. Regarding item reliability, Fornell and
Larcker (1981) suggest that loadings of indicators on latent
constructs greater than 0.7 are sufcient to establish reli-
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.
As with all self-reported data, there is a potential

for common method bias (CMB) resulting from multiple
sources such as consistency motif and social desirability
(Podsakoff et al., 2003; Podsakoff and Organ, 1986). Several actions were taken to address the potential threat
of CMB. First, a Harmon one-factor test (Podsakoff and
Organ, 1986) was conducted on the three variables in the
theoretical model including MC, OC, INN and PERF. The
results from this test showed that four factors are present
and the most covariance explained by one factor was
approximately 28%, indicating that CMB is not a likely contaminant of the results. Second, following Podsakoff et al.
102
Table 2
Estimation of the measurement model parameters.
Original sample
Sample mean
Standard deviation
Exploratory projects (n = 68)

OC1
OC2
OC3
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

MC1
MC2
MC3
MC4
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
Exploitative projects (n = 51)

OC1
OC2
OC3
0.538
0.719
0.565
0.473
0.668
0.544
0.231
0.248
0.272
2.324
2.901
2.078

MC1
MC2
MC3
MC4
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).
method variance was about 111:1. In addition, all method

factor loadings bar one are non-signicant. Given the small
magnitude and insignicance of method variance, evidence
collectively suggests that common method bias is not a
signicant issue in this study.
4.3. Structural model
The second step in the PLS analysis is the estimation of
the specied structural equations. As PLS does not make
any assumption about data distribution, the statistical
signicance of examined paths is evaluated by nonparametric techniques. Bootstrapping with replacement
103
Table 3
Discriminant validity coefcients.
Exploratory projects (n = 68)
OC
MC
Inn.
Perf.
Duration
No.of people
TU

Innovativeness (INN)
Project performance (PERF)
Project duration
No. of people
Task uncertainty (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
Exploitative projects (n = 51)

Innovativeness (INN)
Project performance (PERF)
Project duration
No. of people
Task uncertainty (TU)
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
Each cell reports the path coefcient (t-value).

Stage I reports test results without interaction.
Stage II reports test results with interaction.
**
p < 0.010 (one-tailed tests).
***
p < 0.001 (one-tailed tests).
using 500 samples (generated from the original dataset)

was undertaken, as suggested by Chin (1998). Table 5
presents the structural model and outlines the results of
the hypothesis testing. The structural model was independently tested for the exploratory and exploitative project
subgroups. All control variables were included in the
research model from stage 1, which represents the model
without the interaction variable. The control variable
measuring project size by number of people was not found
to be signicant at any stage of the analysis, whereas the
duration of project variable was signicant in exploitative
projects in both stages. Furthermore, the control variable
Task Uncertainty was found to have a signicant effect on
PERF in both subgroups and at both stages.
Although no hypotheses were proposed for the direct
effects of OC and MC on project performance, the results
will be briey reported. First, the test of the initial research
model without the mediator variable INN indicates direct
effects of OC on performance in exploratory ( = 0.240,
p < 0.05) and exploitative innovation ( = 0.326, p < 0.01),
but non-signicant effects of MC ( = 0.104, p > 0.05 and
= 0.010, p > 0.05, respectively). Second, after including
the mediator INN, the evidence indicates that there is a
signicant positive direct effect of OC on PERF in exploitative innovation projects ( = 0.267, p < 0.01), whereas there
is a signicant direct positive relationship between MC and
PERF in exploratory innovation projects ( = 0.160, p < 0.05)
(Stage I in Table 5). This would imply that OC enhances
project performance in exploitative settings by increasing

the shared understanding and common values of project
goals and, as a consequence, are likely to increase project
efciency and performance. In addition MC, including the
focus on detailed performance evaluation and target deviations, ensures discipline and denes project boundaries,
and is therefore likely to make the team aware of scarce
resources, to reduce costly errors and thereby safeguard
project performance in exploratory innovation projects
(Chenhall and Morris, 1995).
The results in Table 5 (Stage I) also indicate a signicant path between OC and INN ( = 0.411, p < 0.01)
in exploratory innovations. Thus, OC seems to be useful
in driving project innovativeness in exploratory projects,
where innovation is highly valued and generation of
new knowledge is more important. Moreover, the evidence indicates a signicantly positive effect of INN
on PERF in exploratory projects ( = 0.402, p < 0.001,
Table 5).
Contrary to expectations, no evidence of a signicant
positive relationship between MC and INN was found in
the exploitative innovation subgroup ( = 0.023, p > 0.05,
Stage I in Table 5). The empirical evidence, however, indicates a signicant positive effect of INN on PERF ( = 0.270;
p < 0.01, Table 5). This positive signicant relation between
INN and PERF conrms the strong relationship between
innovativeness and performance in both types of innovation, something also noted in previous studies.
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
It was hypothesized that OC and MC respectively would

affect project performance through innovativeness. Stage
II in Table 5 shows that OC in exploratory innovation
has a positive effect on INN ( = 0.411; p < 0.001), which
in turn has a positive effect on PERF measured through
INN ( = 0.402; p < 0.001). Further, the results indicate in
exploitative innovation a non-signicant positive effect of
MC on INN ( = 0.023; p > 0.05) and a positive effect of INN
on PERF ( = 0.233; p < 0.01, Stage II in Table 5). Together,
the results in exploratory innovation fulll the conditions
(Preacher et al., 2007) to consider that INN mediates the
relationship between OC and PERF, but not in exploitative innovation. More specically, the results in exploratory
innovation support a pattern of partial mediation (Mathieu
and Taylor, 2006), as the effect of OC appears to still exist
after including INN. Thus, these results suggest that the
association between OC and PERF is partially accounted for
by INN, but that OC most likely also has a direct effect on
project performance in exploratory innovation.
Further, following previous practice (Sarkar et al., 2001)
the indirect effect was calculated as a multiplication of
the statistically signicant indirect effects on and from
innovativeness. The indirect effect for exploratory innovations was derived by multiplying the signicant path
coefcients, OCINN 0.411 and INNPERF 0.402, which
generated an indirect and total effect of 0.165. The indirect
effect for exploitative innovation could not be analyzed as
the path between MCINN was not signicant. Thus, the
results reveal that only the indirect effect of OC on project
performance through innovativeness can be analyzed in
this study.
The statistical signicance of the indirect effect of OC
on project performance through innovativeness was tested
using a bootstrapping procedure with 1000 bootstraps.
Compared to common techniques such as the three-step
procedure of Baron and Kenny (1986) or the Sobel test, the
bootstrapping procedure has been used more recently (e.g.,
2012; Hall, 2011) and is suggested
Bisbe and Malagueno,
to be more appropriate for small samples not following a
multivariate normal distribution, like the sample in this

study (e.g., Preacher and Hayes, 2004, 2008; MacKinnon
et al., 2004). Following the procedure of Preacher and Hayes
(2008, pp. 883884), for each of the 1000 bootstraps the
estimated coefcients for each direct path were multiplied in order to calculate an estimated coefcient for the
indirect effect. The 1000 indirect effect coefcients were
then rank-ordered and analyzed. The bootstrapped 95%
condence intervals did not record a zero in exploratory
projects, indicating a signicant indirect effect whereas
in exploitative projects the bootstrapped 95% condence
intervals included zero, indicating a non-signicant indirect effect. Thus, the results of the undertaken procedure
provide support for Hypothesis 1, but not Hypothesis 2.
In addition to the indirect effects, the results of the
test for the combined effect of MC and OC indicate a signicant positive interaction effect ( = 0.218; p < 0.01) for
the exploratory project subgroup and a signicant positive
interaction effect ( = 0.341; p < 0.001) for the exploitative
projects, when project size and uncertainty are controlled
for (Table 5). Thus, Hypothesis 3 was supported. To shed
some light on the nature of the interactions, low and
high subgroups (A, B, C and D) were created for OC
and MC, and the differences in the PERF and INN mean
scores for each of the four subgroups in exploratory and
exploitative projects respectively were examined (Table 6).
As expected, the lowest mean scores of performance PERF
and innovativeness INN occur where both OC and MC are
low (cells A1, A2, A3, but with the exception of A4, instead
C4). The highest PERF occurs in exploratory projects where
OC is high and MC is low (cell B1) and in exploitative
settings where both OC and MC are high (cell D3). The
highest INN occurs in both exploratory and exploitative
projects where OC is high and MC is low (cell B2 and B4).
An ANOVA test of the differences in mean scores between
the subgroups revealed statistically signicant differences
in exploratory innovations for INN (3.913, p = 0.013) and
in exploitative innovations for PERF (3.586, p = 0.02). The
Scheff test for unplanned multiple comparisons, however,
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
***p <0.001, **p <0.010

Results for exploratory projects are marked in bold and exploitative projects in italic.
Fig. 2. Results of PLS analysis. *** p < 0.001, ** p < 0.010. Results for exploratory projects are marked in bold and exploitative projects in italic.
indicates marginally signicant differences in exploratory

projects only between cells A2B2 and A2D2 for INN
(p = 0.055 and p = 0.058 respectively) and in exploitative
projects between cells A3 and D3 for PERF (p = 0.072). In
summary, this analysis indicates that project performance
in both projects is lowest where both OC and MC are lowest, and highest where OC is high and MC is either low (in
exploratory projects) or high (in exploitative projects). In a
similar way, innovativeness is lowest where OC is low and
MC is either low (exploratory projects) or high (exploitative
projects), and INN is highest with high OC and low MC in
both projects. These results suggest that OC plays an important role in both exploratory and exploitative projects,
whereas MC has a minor role.
Finally, as the objective of PLS is to maximize variance
explained rather than t, prediction-orientated measures,
such as R-square (R2 ), were used to evaluate the PLS models (Chin, 1998). The R2 for the main effects and interaction
models are presented in Table 5. To see how much predictive value the interaction terms add to the model, the R2
from the main model without interactions (Stage 1) was
compared with the R2 from the interaction effects model
(Stage 2). The additive explanatory power of the interaction model was determined by calculating Cohens f2 effect
size measure (Chin et al., 2003; Cohen, 1988). The results
show that the interaction constructs have a total effect
size f of 0.23 in exploitative innovation projects and 0.07
in exploratory innovation projects, which indicates that
the inclusion of the interaction terms does improve the
explanatory power of the model.
The predictive validity of the parameter estimates can
be assessed by use of a cross-validated redundancy index
or a StoneGeisser Q2 -test (Geisser, 1974; Stone, 1974).
As PLS models lack an index for goodness of t statistics,
Tenenhaus et al. (2005) and Vandenbosch (1996) argue that
besides the reliability and validity of constructs, the signicance of variance explained and positive Q2 s for all but one
of the constructs provide sufcient evidence of model t.
The results in Table 5 suggest that the model has predictive
relevance.
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
approaches such as detailed project planning including

budgets and timetables do not generate forces and a
focus on increasing performance in innovative project
settings. On the other hand, the results also indicate that
MC may not be as detrimental to exploratory innovation
efforts as some previous studies have found (e.g., Song and
Montoya-Weiss, 1998).
In exploratory innovations, with high levels of uncertainty, radical change and strong emphasis on new
knowledge generation, it appears that OC works best
indirectly through innovativeness. One possible explanation is that OC stimulates idea generation, exibility, and
interaction among team members and thus enhances innovativeness, which is of vital importance in exploratory
innovations, rather than project completion and overall
project performance. In exploitative settings, with regard
to MC, there is no evidence of any indirect effects of MC on
project performance through innovativeness.
Third, the results suggest that performance is positively
affected by tension emerging from the combined use of OC
and MC in both exploratory and exploitative innovation
projects. These results extend the ndings of Henri (2006),
who in an ex post analysis revealed that dynamic tension
might have a direct positive and signicant impact on performance, in particular for rms facing high environmental
uncertainty and with values encompassing exibility. The
results are also in line with Lewis et al. (2002), who
argued that meeting tenuous demands for innovation and
efciency requires a blend of emergent (organic) and
planned (mechanistic) approaches, and that managing
tensions between exibility and efciency may prove key
to strong project performance. Combining opposing forms
of control such as OC and MC enables the desired direct
and indirect effects of OC and MC to simultaneously yield
high positive levels of both. Thus, MC used jointly with OC
appears to create a tension that enhances project performance in both exploratory and exploitative innovations.
This combined effect is particularly useful in exploitative
settings, where MC interacts with OC to enhance project
performance. Therefore, the tension created in exploratory
and exploitative innovation projects has complementary
effects. This nding is interesting and also consistent with
more recent theorizing that MC is not without value in
uncertain environments.
In summary, OC appears to be the project managers
chosen tool for managing project tensions. The results
support an indirect effect of OC on project performance
through innovativeness in exploratory projects (H1), but do
not support indirect effects of MC on project performance
through innovativeness in exploitative settings (H2). The
interaction effect of OC and MC was supported (H3) in
both innovation environments. Thus, it is the conclusion
of the current research examining project controls in two
different innovation types that exploratory and exploitative innovation projects may indeed benet from different
types of management controls.
6. Conclusion
The aim of this study was to investigate the indirect effects of mechanistic and organic forms of control
107
on project performance through innovativeness as well

as the interaction effect of those controls on project
performance in exploratory and exploitative innovation
projects. Overall, the results support the importance of
OC, showing it acting through innovativeness on project
performance in exploratory innovations and enhancing project performance in exploitative innovations. The
results also suggest that the interaction effect of OC and MC
enhances performance in both exploratory and exploitative innovation projects, and support a complementary
effect.
This study contributes to literature in three ways.
Firstly, this study draws on prior research in management accounting and control (Chenhall and Morris,
1995; Henri, 2006; Jrgensen and Messner, 2009; Mundy,
2010) to extend those to another level of analysis, the
project level rather than the organizational level. While
accounting research has studied control mechanisms in
other organizational contexts, the project level has largely
been forgotten (Chenhall, 2008), although scholars argue
it to be more appropriate for studying control systems in
innovation settings (Davila et al., 2009b). Secondly, this
study offers empirical evidence of the impact of indirect effects of project controls on project performance.
By distinguishing between exploratory and exploitative
innovative project settings, this study nds indirect effects
of organic project controls on performance via innovativeness in exploratory settings. These ndings are seen
as extending the ndings of Bisbe and Otley (2004) at
the organizational level. This study also nds empirical evidence of tension on project performance resulting
from a combined use of mechanistic and organic forms
of control in both exploratory and exploitative innovation project settings. As the results indicate, the tension
caused by a combination of project control mechanisms
appears to be complementary, but not dependent on the
degree of innovation newness. Although prior research on
opposing control forces in exploratory innovation settings
exists, empirical research comparing the effects of combined use of two opposing forms of project control in both
exploratory and exploitative innovative project settings is
non-existent.
This study also has important managerial implications.
First, it emphasizes the important role of organic communication processes in enhancing project performance
through innovativeness in exploratory innovation settings. It also shows that project managers can identify how
the different forms of control they already use or intend
to use can most benet innovation by combining those
controls with their respective complementary opposite
coordination mechanisms. The variety of possible combinations of different forms of MCS or style of use of MCS
provides the project manager with a pool of options. A
project manager who is informed about the desirable and
undesirable effects and who pays them due attention,
can enhance performance and innovativeness. However,
combining opposing control strategies may not be easy
in practice within the same social system, and evidence
from prior studies also indicates that practitioners often
do not combine opposing action strategies (e.g. Sheremata,
2002).
108
Some limitations of this study need to be noted. As

survey research, the study is vulnerable to the typical
weaknesses relating to validity and reliability of items and
tests, although best practice was followed in the development and pre-testing of the instrument (Van der Stede
et al., 2005). One of the obvious limitations of this study is
the small sample size and the fact that its sample was not
strictly randomly selected; meaning any inferences from
the results must be drawn cautiously. The relatively high
number of exploratory projects compared to exploitative
projects in this study differs from a normal innovation setting, and may suggest a bias. The combined use of organic
(OC) and mechanistic (MC) forms of project control was
also examined using a product term, and a complete and
reliable test for complementarity between the two forms of
control and project performance would include incorporating appropriate bi-directional links into the model, which
was not possible in this study. In addition, the limited
measurement of mechanistic and organic forms of control
and the use of a categorical (and not continuous) measure
of innovation type would suggest the results should be
interpreted with caution. Furthermore, it would be wise to
be cautious when interpreting the statistical associations
as causal relationships, owing to the cross-sectional nature
of the study. Although project managers are considered
the best judges of their own actions (Brownell, 1995), a
further limitation is the use of self-assessed use ratings.
Previous research has, however, found a signicant positive correlation between superiors subjective rating and
objective measures of subordinate performance (Bommer
et al., 1995; Furnham and Stringeld, 1994). It is also
important to note that the effects obtained in the different
runs of the model show high robustness and the results of
the preliminary and nal analyses are consistent.
It is suggested that more accounting studies on control and coordination mechanisms in project environments
would be worthwhile. Studies examining MCS from different innovation perspectivessuch as other types or stages
of innovationmay nd different outcomes or allow replications that make these results generalizable. The tension
arising from the simultaneous use of various forms of
control or various styles of use of control mechanisms
may change over time, and enhance project performance
differently in the early stages of project development
than they do during late development and commercialization. Thus, studies using longitudinal data might reveal
whether project managers adjust their actions to contextual changes, and if so, how (Lewis et al., 2002; Pennings,
1992). In addition, qualitative methodologies may provide
new insights into the ways in which project managers
in different kinds of organizations and project settings
reinforce and manage tension in practice. Moreover, the
personality traits of the project manager are likely to be
of considerable importance when implementing and using
opposing forms of MCS simultaneously and well worth
investigation in future research.
Acknowledgments
The authors would like to thank two anonymous
reviewers, Professor Robert H. Chenhall and the
participants of the MONFORMA conference in Melbourne in November, 2010 for their very insightful and
helpful comments.
Appendix A. Measurement instrument

A.1. Organic form of control (OC)
Level of agreement with the statements (1 = strongly
disagree, 7 = strongly agree).
OC1 . I call my project team subordinates in to discuss
project performance deviations in face-to-face meetings.
OC2 . My own project team subordinates and I often discuss
and resolve project performance issues together informally.
OC3 . I place considerable emphasis on open channels of
communication and the free ow of information between
myself and my subordinates.
A.2. Mechanistic form of control (MC)
Level of agreement with the statements (1 = strongly
disagree, 7 = strongly agree).
Detail of performance evaluation and importance of
project performance target deviations.
MC1 . I judge my project team performance with
performance measures that explain in detail project performance variances on a line-by-line basis.
MC2 . I am not only interested in how well my project
team achieves the overall project performance targets,
but I also evaluate the extent to which my project team
is on target in each of the project performance lineitems.
MC3 . I attach a great deal of importance to interim
project performance target deviations from budgeted performance and project milestones.
MC4 . I require my project team subordinates to report
the actions taken to correct causes of deviation from the
interim project performance targets.
A.3. Innovativeness (INN)

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. Anchored by 1 = extremely
low, and 7 = extremely high.
INN1 . The level of my project team innovation performance (number of innovations) measured relative to
project performance targets (standards).
INN2 . The level of my project team innovation performance (number of innovations) relative to other project
teams measured innovation performance working on the
same type of project.
109
Table B1
Common method bias analysis.
Item
Substantive factor loading (R1)
R12
Common method factor loading (R2)
R22
Mechanistic control (item 1)

Organic control (item 1)
Project Performance (item 1)
Innovativeness (item 1)
Task Uncertainty (item 1)
Project Duration (item 1)
No. People in Project (item 1)
Average
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.
INN3 . The level of my project team measured innovation

performance.
A.4. Project performance (PERF)
Level of performance of your
(1 = extremely low, 7 = extremely high).
project
team
Perf1 . The level of my project team measured performance

relative to my performance targets (standards).
Perf2 . The level of my project team measured performance
relative to other project teams measured performance
working on the same kind of project.
Perf3 . The level of my project team measured performance.
A.5. Task uncertainty (TU)
Level of agreement with the statements (1 = to a small
extent, 7 = to a great extent).
The scale was revised in the analysis, that is, a larger
number now indicates a greater uncertainty.
TU1 . To what extent was there a clearly understood way
of doing the major types of work normally encountered in
your project?
TU2 . To what extent was there a clearly dened body of
knowledge of subject matter, which could guide the work
done in your project?
TU3 . To do the work of your project, to what extent could
personnel actually rely on established procedures and
practices?
Appendix B.
Table B1.
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.

Marginson, D., 2002. Management control systems and their effects on
strategy formation at the middle-management levels: evidence from a
U.K. organization. Strategic Management Journal 23 (11), 10191031.
Martino, J., 1995. Research and Development Project Selection. Wiley and
Sons, New York, NY.
Mathieu, J.E., Taylor, S.R., 2006. Clarifying conditions and decision points
for mediational type inferences in organizational behavior. Journal of
Organizational Behavior 27, 10311056.
McDonough III, E.F., Barczak, G., 1991. Speeding up new product development: the effects of leadership style and source of technology. Journal
of Product Innovation Management 8 (3), 203211.
Menguc, B., Auh, S., 2008. Innovation and knowledge creation:
the asymmetric moderating role of market orientation on the
ambidexterityrm performance relationship for prospectors and
defenders. Industrial Marketing Management 37 (4), 455470.
Milgrom, P., Roberts, J., 1995. Complementarities and t: strategy,
structure, and organizational change in manufacturing. Journal of
Accounting and Economics 19 (23), 179208.
Montoya-Weiss, M.M., Calantone, R., 1994. Determinants of new product
performance: a review and meta-analysis. Journal of Product Innovation Management 11 (5), 397417.
Morse, J.J., Lorsch, J.W., 1970. Beyond theory. Harvard Business Review 48
(MayJune), 6168.
Mundy, J., 2010. Creating dynamic tensions through a balanced use of
management control systems. Accounting, Organizations and Society
35 (5), 499523.
Nunnally, J.C., 1978. Psychometric Theory. McGraw-Hill, New York.
OECD, 2004. Oslo Manual. The Measurement of Scientic and Technological Activities. Proposed Guidelines for Collecting and Interpreting
Technological Innovation Data. European Commission, Eurostat,
Available at: http://www.oecd.org/dataoecd/35/61/2367580.pdf
Olson, E.M., Walker, O.C., Ruekert, R.W., 1995. Organizing for effective new
product development: the moderating role of product innovativeness.
Journal of Marketing 59 (1), 4862.
Ouchi, W.G., 1977. The relationship between organizational structure
and organizational control. Administrative Science Quarterly 22 (1),
95113.
Ouchi, W.G., 1979. A conceptual framework for the design of organizational control mechanisms. Management Science 25, 833848.
Pennings, J.M., 1992. Structural contingency theory: a reappraisal. In:
Staw, B.M., Cummings, L.L. (Eds.), Research in Organizational Behavior,
14. JAI Press, Greenwich CT, pp. 267309.
Podsakoff, P.M., MacKenzie, S.B., Lee, J.-Y., Podsakoff, N.P., 2003. Common
method biases in behavioral research: a critical review of the literature
and recommended remedies. Journal of Applied Psychology 88 (5),
879903.
Podsakoff, P.M., Organ, D.W., 1986. Self-reports in organizational research:
problems and prospects. Journal of Management 12 (4), 531
544.
Popadiuk, S., Choo, C.W., 2006. Innovation and knowledge creation: how
are these concepts related? International Journal of Information Management 26 (4), 302312.
Poppo, L., Zenger, T., 2002. Do formal contracts and relational governance
function as substitutes or complements? Strategic Management Journal 23, 707725.
Preacher, K.J., Hayes, A.F., 2004. SPSS and SAS procedures for estimating indirect effects in simple mediation models. Behavior Research
Methods 36 (4), 717731.
Preacher, K.J., Rucker, D.D., Hayes, A.F., 2007. Assessing moderated mediation hypotheses: theory, method, and prescriptions. Multivariate
Behavioral Research 42, 185227.
Preacher, K.J., Hayes, A.F., 2008. Asymptotic and resampling strategies for
assessing and comparing indirect effects in multiple mediator models.
Behavior Research Methods 40 (3), 879891.
Rijsdijk, S.A., van den Ende, J., 2011. Control combinations in new product
development projects. Journal of Product Innovation Management 28,
868880.
Ringle, C.M., Wende, S., Will, A., 2005. SmartPLS, 2.0 (beta), Available at:
http://www.smartpls.de. Hamburg, Germany.
Rogers, E., 1995. Diffusion of Innovations. Free Press, New York.
Sarkar, M.B., Echambadi, R., Cavusgil, S.T., Aulakh, P.S., 2001. The inuence of complementarity, compatibility, and relationship capital on
alliance performance. Academy of Marketing Science Journal 29 (4),
358373.
Shenhar, A.J., Dvir, D., 1996. Toward a typological theory of project management. Research Policy 25 (4), 607632.
Shenhar, A.J., Tishler, A., Dvir, D., Lipovetsky, S., Lechler, T., 2002. Rening the search for project success factors: a multivariate, typological
approach. R&D Management 32, 111126.
111
Sheremata, W.A., 2000. Centrifugal and centripetal forces in radical new

product development under time pressure. Academy of Management
Review 25 (2), 389408.
Sheremata, W.A., 2002. Finding and solving problems in software new
product development. Journal of Product Innovation Management 19
(2), 144158.
Shields, M.D., Kato, Y., Deng, J., 2000. The design of control systems: test
of direct and indirect effect models. Accounting, Organizations and
Society 25 (2), 185202.
Siggelkow, N., 2002. Misperceiving interactions among complements and
substitutes: organizational consequences. Management Science 48
(7), 900916.
Simons, R., 1995. Levers of Control: How Managers Use Innovative Control
Systems to Drive Strategic Renewal. Harvard Business School Press,
Boston, MA.
Song, X.M., Montoya-Weiss, M., 1998. Critical development activities for
really new versus incremental products. Journal of Product Innovation
Management 15, 124135.
Stone, M., 1974. Cross-validatory choice and assessment of statistical predictions. Journal of the Royal Statistical Society, Series B 36, 111133.
Sundaramurthy, C., Lewis, M., 2003. Control and collaboration: paradoxes
of governance. Academy of Management Review 28 (3), 397415.
Tatikonda, M.V., Montoya-Weiss, M.M., 2001. Integrating operations
and marketing perspectives of product innovation. The inuence of
organizational process factors and capabilities on development performance. Management Science 47, 151172.
Tatikonda, M.V., Rosenthal, S.R., 2000a. Successful execution of product
development projects: balancing rmness and exibility in the innovation process. Journal of Operations Management 18 (4), 401425.
Tatikonda, M.V., Rosenthal, S.R., 2000b. Technology novelty, project complexity, and product development project execution success: a deeper
look at task uncertainty in product innovation. IEEE Transactions on
Engineering Management 47 (1), 7487.
Tenenhaus, M., Vinzi, V.E., Chatelin, Y., Lauro, C., 2005. PLS path modelling.
Computational Statistics and Data Analysis 48 (1), 159205.
Turner, K.L., Makhija, M.V., 2006. The role of organizational controls
in managing knowledge. Academy of Management Review 31 (1),
197217.
Tushman, M., Nadler, D., 1978. Information processing as an integrating
concept in organizational design. Academy of Management Review 3
(3), 613624.
Tushman, M.L., OReilly, C.A., 1996. Ambidextrous organizations: managing evolutionary and revolutionary change. California Management
Review 38 (4), 830.
Un, C.A., 2010. An empirical multi-level analysis for achieving balance
between incremental and radical innovations. Journal of Engineering
and Technology Management 27 (12), 119.
Wageman, R., 2001. How leaders foster self-managing team effectiveness:
design choices versus hands-on coaching. Organization Science 12 (5),
559577.
Van de Ven, A.H., 1986. Central problems in the management of innovation. Management Science 32 (5), 590607.
Van de Ven, A., Andrew, H., Polley, D.E., Garud, R., Venkataraman, S., 1999.
The Innovation Journey. Oxford University Press, New York.
Van de Ven, A.H., Polley, D., 1992. Learning while innovating. Organization
Science 3 (1), 92116.
Van der Stede, W.A., 2001. Measuring tight budgetary control. Management Accounting Research 12 (1), 119137.
Van der Stede, W.A., Young, S.M., Chen, C.X., 2005. Assessing the quality of evidence in empirical management accounting research: the
case of survey studies. Accounting, Organizations and Society 30 (78),
655684.
Vandenbosch, M.B., 1996. Conrmatory compositional approaches to the
development of product spaces. European Journal of Marketing 30 (3),
2346.
Venkatraman, N., 1989. The concept of t in strategy research: toward verbal and statistical correspondence. Academy of Management Review
14 (3), 423444.
Wheelwright, S.C., Clark, K.B., 1992. Revolutionizing Product Development: Quantum Leaps in Speed, Efciency, and Quality. Free Press,
New York.
Widener, S.K., 2004. An empirical investigation of the relation between
the use of strategic human capital and the design of the management control system. Accounting, Organizations and Society 29 (34),
377399.
Widener, S.K., 2007. An empirical analysis of the levers of control framework. Accounting, Organizations and Society 32 (78), 757788.
Williams, K.Y., OReilly, C.A., 1998. Demography and diversity in organizations: a review of 40 years of research. In: Staw, B.M., Sutton, R.I.
112
(Eds.), Research in Organizational Behavior, 20. JAI Press, Greenwich,

CT, pp. 77140.
Williams, L.J., Edwards, J.R., Vandenberg, R.J., 2003. Recent advances
in causal modeling methods for organizational and management
research. Journal of Management 29 (6), 903936.
Wilson, J.M., Goodman, P.S., Cronin, M.A., 2007. Group learning. Academy
of Management Review 32 (4), 10411059.
Withey, M., Daft, R.L., Cooper, R.H., 1983. Measures of Perrows work-unit
technology: an empirical assessment and a new scale. Academy of
Management Journal 26 (1), 4563.
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.

1 s2.0 S1044500513000401 Main

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

1 s2.0 S1044500513000401 Main

Uploaded by

Copyright:

Available Formats

Management Accounting Research 25 (2014) 93112

Contents lists available at ScienceDirect

Management Accounting Research

The effects of organic and mechanistic control in exploratory

Corresponding author. Tel.: +358 40 011 3998.

changes in technology and represent clear departures from

M. Ylinen, B. Gullkvist / Management Accounting Research 25 (2014) 93112

M. Ylinen, B. Gullkvist / Management Accounting Research 25 (2014) 93112

because they provide the information necessary to close

M. Ylinen, B. Gullkvist / Management Accounting Research 25 (2014) 93112

Organic control (OC)

Mechanistic control (MC)

Direct effect ____

to come indirectly through innovativeness rather than

M. Ylinen, B. Gullkvist / Management Accounting Research 25 (2014) 93112

control in exploratory innovation projects will be positively

M. Ylinen, B. Gullkvist / Management Accounting Research 25 (2014) 93112

and regulations as rigid and repressive (Dougherty, 2006;

M. Ylinen, B. Gullkvist / Management Accounting Research 25 (2014) 93112

exploratory projects, MC could provide additional benets

M. Ylinen, B. Gullkvist / Management Accounting Research 25 (2014) 93112

No. of items used

Organic control (OC)

(a) Reverse coded.

extremely high. While the degree to which subjective

M. Ylinen, B. Gullkvist / Management Accounting Research 25 (2014) 93112

ability. All items suggest good indicator reliability with few

4.1. Measurement model

As with all self-reported data, there is a potential

M. Ylinen, B. Gullkvist / Management Accounting Research 25 (2014) 93112

Exploratory projects (n = 68)

Mechanistic control (MC)

Exploitative projects (n = 51)

Mechanistic control (MC)

method variance was about 111:1. In addition, all method

M. Ylinen, B. Gullkvist / Management Accounting Research 25 (2014) 93112

Organic control (OC)

Exploitative projects (n = 51)

M. Ylinen, B. Gullkvist / Management Accounting Research 25 (2014) 93112

Each cell reports the path coefcient (t-value).

using 500 samples (generated from the original dataset)

project performance in exploitative settings by increasing

M. Ylinen, B. Gullkvist / Management Accounting Research 25 (2014) 93112

It was hypothesized that OC and MC respectively would

multivariate normal distribution, like the sample in this

M. Ylinen, B. Gullkvist / Management Accounting Research 25 (2014) 93112

Organic control (OC)

Mechanistic control (MC)

***p <0.001, **p <0.010

indicates marginally signicant differences in exploratory

M. Ylinen, B. Gullkvist / Management Accounting Research 25 (2014) 93112

approaches such as detailed project planning including

on project performance through innovativeness as well

M. Ylinen, B. Gullkvist / Management Accounting Research 25 (2014) 93112

Some limitations of this study need to be noted. As

Appendix A. Measurement instrument

A.3. Innovativeness (INN)

M. Ylinen, B. Gullkvist / Management Accounting Research 25 (2014) 93112

Substantive factor loading (R1)

Common method factor loading (R2)

Mechanistic control (item 1)

INN3 . The level of my project team measured innovation

Perf1 . The level of my project team measured performance

*p <0.001, p <0.010