Journal of Operations Management 25 (2007) 871–884

www.elsevier.com/locate/jom

Innovation-supportive culture: The impact of organizational

values on process innovation
Shalini Khazanchi a,*, Marianne W. Lewis b,1, Kenneth K. Boyer c,2
a
Rochester Institute of Technology, Department of Management, Marketing & International Business,
108 Lomb Memorial Drive, Rochester, NY 14623, United States
b
University of Cincinnati, Management Department, P.O. Box 210165, Cincinnati, OH 45221-0165, United States
c
Michigan State University, Department of Marketing & Supply Chain Management,
N370 North Business Complex, East Lansing, MI 48824-1122, United States
Received 30 June 2004; received in revised form 25 October 2005; accepted 20 August 2006
Available online 13 November 2006

Abstract
For managers, innovation is vital, but paradoxical, requiring flexibility and empowerment, as well as control and efficiency.
Increasingly, studies stress organizational culture as a key to managing innovation. Yet innovation-supportive culture remains an
intricate and amorphous phenomenon. In response, we explore how organizational values – a foundational building block of culture
– impact a particular process innovation, the implementation of advanced manufacturing technology (AMT). To unpack this
scarcely studied construct, we examine three-dimensions of organizational values: value profiles, value congruence and value–
practice interactions.
# 2006 Elsevier B.V. All rights reserved.

Keywords: Empirical research methods; Flexible manufacturing systems; Innovation

1. Introduction market demands. Likewise, process innovation involves

Innovation offers a critical source of sustainable
competitive advantage. Indeed, Christensen (1999)
describes the management of innovation – in its many
forms – as an overriding responsibility of today’s
managers. Product innovation, for instance, entails
developing new goods and services. Managing such
innovation may help firms meet or even drive changing
* Corresponding author. Tel.: +1 585 475 7126;
fax: +1 585 475 5989.
E-mail addresses: skhazanchi@cob.rit.edu (S. Khazanchi),
marianne.lewis@uc.edu (M.W. Lewis), boyerk@msu.edu (K.K.
Boyer).
1
Tel.: +1 513 556 7124; fax: +1 513 556 4891.
2
Tel.: +1 517 353 6381.
creating or improving methods of production, service or
administrative operations. Effective process innovation
may enhance organizational efficiency and responsive-
ness (Damanpour and Gopalakrishnan, 2001).
Despite its promised potential, however, innovation
poses tremendous challenges. For example, innovation
efforts may be highly disruptive, altering relationships
across functional and occupational boundaries or
requiring changes to the organizational structure and
climate (Baer and Frese, 2003; Black et al., 2004; Detert
et al., 2000; McDermott and Stock, 1999). Paradoxes of
innovation may pose even more fundamental and less
understood challenges (Lewis et al., 2002). As Dougherty
(1996) explains, successful innovation requires mana-
ging flexibility-control tensions. Flexibility enables
creativity, empowerment and change vital for the

0272-6963/$ – see front matter # 2006 Elsevier B.V. All rights reserved.
doi:10.1016/j.jom.2006.08.003
872 S. Khazanchi et al. / Journal of Operations Management 25 (2007) 871–884
exploration that fuels innovation. Control, on the other
hand, provides discipline, focusing innovation initiatives,
for instance, on achieving long-term goals, leveraging
core competencies, and meeting budgets. As such,
innovation paradoxes may foster mixed messages and
role ambiguity that adversely affect performance.
Studies increasingly tout organizational culture as a
key to managing innovation. Jassawalla and Sashittal
(2002) define an innovation-supportive culture as a
firm’s ‘‘social and cognitive environment, the shared
view of reality, and the collective belief and value
systems reflected in a consistent pattern of behaviors
among participants’’ (2002: 43). They propose that
culture may provide an overarching frame of
reference, helping align employee behavior with
organizational objectives of innovation and meet
paradoxical demands for control and flexibility. Yet
researchers stress that innovation-supportive culture
remains an intricate and amorphous phenomenon
(e.g., Frohman, 1998; Higgins and McAllaster, 2002).
In their review, Detert et al. (2000) argue that the
complexity of accumulating theories of culture is
matched by a dearth of corresponding empirical
research. They conclude by calling for studies that
investigate specific components of innovation-suppor-
tive culture and their impact.
In response, this study explores organizational
values, a foundational building block of innovation-
supportive culture that has received scant attention in
empirical research. To unpack this phenomenon, we
examine three-dimensions of organizational values.
First, we examine the role of value profiles, clusters of
interwoven values representing paradoxical orientations
of flexibility or control (Quinn and Rohrbaugh, 1983).
Second, value congruence refers to the extent to which
there is an agreement or consensus about organizational
values amongst organizational members (Chatman,
1989; Kalliath et al., 1999). Value congruence draws
attention to the importance of common goals and
consistent expectations for firm performance (Deal and
Kennedy, 1982; Peters and Waterman, 1982). Lastly,
value–practice interactions address consistency
between organizational values and practices, drawing
attention to the potential for mixed messages (e.g.,
Argyris and Schön, 1974) in an innovation-supportive
culture. In sum, this study seeks to contribute an
expanded understanding of how values may affect
innovation.
To enable focus, we explore a particular context of
process innovation: the implementation of advanced
manufacturing technology (AMT). AMT represents
‘‘a range of programmable machinery that execute,
monitor and connect the production process, including
computer-aided manufacturing, flexible manufactur-
ing systems, and computer numerically controlled
machines’’ (Lewis and Boyer, 2002: 111). Given its
computerization, AMT often marks a dramatic change
from mechanized technologies and manual opera-
tions. Process innovation occurs throughout imple-
mentation and beyond—from initial AMT adoption,
through system design and programming, to use and
continuing improvement within the manufacturing
plant.
This process innovation offers potential insights for
two primary reasons. First, like many cases of
innovation, AMT holds great, but often-untapped
potential. Since the 1970s, AMT has been praised for
its ability to enhance plant performance by improving
production quality, efficiency and responsiveness
(Hayes et al., 1988). Yet reports find that only 25–
50% of implementations live up to expectations
(Cleland et al., 1995), with researchers often blaming
a lack of supportive culture (e.g., Bates et al., 1995;
Lewis, 1998). Second, AMT research stresses the
flexibility-control tensions reflected in wider innovation
literature (Adler, 1993). Successful implementation
requires flexibility for operators to explore, debug and
customize the technology in use. Yet control remains
critical. Clear management objectives, for instance,
help guide implementation and continuous improve-
ment to ensure that the AMT adds value (Kern and
Schumann, 1992).
Manufacturing plants serve as the unit of analysis.
The plant level reduces the impact of technological
heterogeneity and matches the locus of innovation,
as implementation – from initial adoption through
debugging and ongoing use – occurs within a plant or
final home of the AMT (Dean et al., 1992).
Furthermore, when studying culture, staying close to
the phenomena is vital (Hofstede et al., 1990). The
distance between values espoused at corporate head-
quarters and those perceived on a factory floor
exacerbates the potential for value incongruence and
confounding variables that reduce meaningfulness of
findings.
2. Literature review and hypotheses
Although conceptualizations vary, organizational
culture is defined broadly as a collection of values,
beliefs and norms shared by its members and reflected
in organizational practices and goals (Hofstede et al.,
1990). Applying this definition, Jassawalla and Sashittal
(2002) further describe innovation-supportive cultures
 S. Khazanchi et al. / Journal of Operations Management 25 (2007) 871–884
as fostering expectations and guidelines for member
creativity, experimentation and risk taking.
Values are a primary building-block for culture
(Quinn and Rohrbaugh, 1983). Studies suggest that an
innovation-supportive culture derives from values,
which inform an underlying belief structure and
reinforce daily practice (e.g., Frohman, 1998; Higgins
and McAllaster, 2002). According to Detert et al.
(2000), values serve as the backbone of cultures that
foster process innovation, thereby enabling or hindering
performance improvements. Zammuto and O’Connor
(1992) theorize that values play a subtle, yet powerful
role. During AMT implementation, values may affect
critical decisions and emerging norms (e.g., how should
AMT be integrated within the production process; to
what degree should operators experiment with the new
AMT).
Research suggests that three-dimensions of values
may influence innovation: value profiles, value con-
gruence and value–practice interactions. This following
section details each dimension, developing related
hypotheses focused on the AMT implementation
context.
2.1. Value profiles
Quinn and Rohrbaugh (1983) stress the impact of
different value profiles. They define profiles as cohesive
sets of organizational values that orient its members,
guide their expectations, decisions and actions. While a
flexibility value profile stresses creativity, change and
empowerment, a control value profile encourages
efficiency, productivity and stability.
Given their paradoxical demands, however, success-
ful firms rarely hold a single profile, and may even
thrive with a paradoxical mix of values (Quinn and
Kimberly, 1984). Innovation research often highlights
the importance of value profiles that support both
flexibility and control (Dougherty, 1996). Lewis et al.
(2002) explain that product development requires an
emphasis on employees’ applying creative problem
solving, and on managers’ setting clear objectives and
deadlines. Frequently, cases of successful AMT
implementation depict cultures built on seemingly
conflicting, yet complementary values (Lewis, 1998).
Likewise, in their survey study, McDermott and Stock
(1999) find that plants espousing high values for
flexibility and control are most satisfied with AMT
implementation.
Zammuto and O’Connor (1992) theorize that
flexibility values enable innovation-supportive culture
by fostering experimentation. Flexibility values, they
873
claim, underlie a decentralized structure and supportive
climate that enables operators to debug and tailor AMT
to their work. By espousing empowerment, managers
help operators develop new skills, fostering trust and
reducing potential resistance (Lewis and Boyer, 2002).
According to Cleland et al. (1995), flexibility values
enable higher performance by encouraging employee
commitment and problem solving necessary to adjust
production processes to AMT.
Yet studies also suggest that control values support
AMT implementation. Tyre and Orlinkowski (1993)
propose that control values encourage stable routines,
which help operators surface and solve problems during
process innovation. Similarly, Lewis (1998) claims
that discipline is vital to leveraging AMT’s potential.
By stressing explicit goals for responsiveness and
productivity, for instance, control values may enable
computerized coordination. Customer demands (iden-
tified by sales representatives) and systems efficiencies
(analyzed by engineers) may be programmed within
AMT machinery (managed by operators).
In sum, we propose that paradoxical value profiles
support innovation: the higher the flexibility and control
values, the greater the improvement in plant perfor-
mance following AMT implementation. Although
related survey studies are scarce, McDermott and
Stock (1999) examine the impact of the value profiles
perceived by plant managers. Managers are those most
likely to promote the espoused (i.e., formal or overt)
values of the organization (Argyris and Schön, 1974;
Kotter and Heskett, 1992). Hence, a managerial focus is
a useful starting point.
Hypothesis 1a. Flexibility values (value profile per-
ceived by managers) will be positively related to plant
performance.
Hypothesis 1b. Control values (value profile perceived
by managers) will be positively related to plant perfor-
mance.
2.2. Value congruence
Value profiles provide a basis for organizational
culture, but perceptions of an organization’s values may
vary across hierarchical levels, functional departments
and geographic locations (Quinn and Rohrbaugh,
1983). As mentioned earlier, value congruence refers
to the extent to which organizational members perceive
similar organizational values (O’Reilly and Chatman,
1986). Congruence enables a more cohesive culture,
setting consistent and common expectations for
behavior. Shared values enhance goal alignment and,
874 S. Khazanchi et al. / Journal of Operations Management 25 (2007) 871–884
in turn, may positively impact performance (Denison,
1989; Gordon and DiTomaso, 1992). Moreover, it has
been shown that organizations may benefit from
employees who are dedicated to common goals (e.g.,
Deal and Kennedy, 1982; Peters and Waterman, 1982).
In the context of AMT, congruence between
managers and operators is particularly vital. Managers
are the decision makers, determining AMT goals,
guiding its initial design, and monitoring its effective-
ness (Lewis, 1998). Operators, on the other hand, are
AMT users. Operators may adjust the machinery to
improve its speed and accuracy (Kelley, 1990). In
addition, AMT’s informational capabilities enable
greater problem solving on the factory floor, rendering
process innovation an ongoing, operator activity
(Zammuto and O’Connor, 1992).
Given managers and operators’ interdependent roles,
value congruence may be critical to support a common
vision of AMT and its potential to improve plant
performance. Building from the previous hypotheses,
we propose that values should not only be shared, but
also should reflect the paradoxical nature of innovation.
In sum, the positive effects of congruence may be
greater when managers and operators share higher
levels of flexibility and control values.
Hypothesis 2. Plant performance (a) will be positively
related to the congruence between flexibility values as
perceived by managers and operators; and (b) will be
positively related to the level of flexibility values as
perceived by managers and operators.
Hypothesis 3. Plant performance (a) will be positively
related to the congruence between control values per-
ceived by managers and operators; and (b) will be
positively related to the level of control values as
perceived by managers and operators.
2.3. Value–practice interactions
In an innovation-supportive culture, organizational
practices complement values (Frohman, 1998; Higgins
and McAllaster, 2002). Practices moderate the impact
of values by reinforcing or contradicting their message.
As Detert et al. (2000) explain, practices interact with
values, ideally forging a consistent understanding of
‘‘what matters most’’ in daily life. In contrast, when
practices contradict values, mixed messages arise.
According to Argyris and Schön (1974), mixed
messages denote conflicts between espoused values
(i.e., as perceived by managers) and values in practice
(i.e., as applied by employees). They offer the common
example of managers claiming to value empowerment,
while employees feel closely monitored and act
accordingly. The result is a sense of growing distrust
between the groups.
AMT studies suggest countless work, human
resource, coordination and strategic practices that
may interact with flexibility or control value profiles
(e.g., Cleland et al., 1995; Dean et al., 1992; Lewis,
1998; Zammuto and O’Connor, 1992; Zhao and Co,
1997). As a comprehensive review and test is beyond
this paper’s scope, we pair each value profile with a
single practice. Our goal here is to aid future research
on innovation-supportive culture by offering theore-
tical rationale for value–practice interactions and
examples. The specific practices also were chosen to
illustrate that flexibility and control are not simple
opposites—e.g., control is not the absence of flexibility
or visa versa. Rather they denote a more intricate
paradox.
Applying insights from Poole and Van de Ven
(1989), we chose practices that illustrate how the
flexibility-control tension may operate on multiple
levels. Flexibility may emerge from the bottom-up, as
creativity and change are driven often by those on the
front line—or in the case of AMT, by those on the
factory floor. In contrast, control typically descends
from the top, as managerial direction provides formal
goals, guidelines and constraints. In this light, we pair
flexibility values with operator discretion and control
values with formalized AMT objectives.
Operator discretion reinforces flexibility values and
their emphasis on empowerment. AMT literature takes
a very focused view of discretion, examining operators’
ability to decide how they use AMT—e.g., by setting
the pace of their work, choosing the order in which they
complete tasks, and varying their work methods (e.g.,
Jackson et al., 1993; Kelley, 1990). The interaction of
flexibility values and operator discretion may facilitate
the ongoing AMT adjustments and factory floor
problem solving needed to enhance performance.
Similarly, AMT objectives may complement control
values. Successful implementation requires that man-
agers – AMT decision makers and planners – clearly
communicate the purpose of this process innovation
and its strategic role (Zhao and Co, 1997). In addition,
Small and Yasin (1997) claim that well-defined, long-
term objectives signal management’s commitment to
AMT and their expectations for improved plant
performance.
Hypothesis 4a. Operator discretion (as perceived by
operators) will moderate the relationship between flex-
ibility values (as perceived by managers) and plant
 S. Khazanchi et al. / Journal of Operations Management 25 (2007) 871–884
performance. Specifically, the positive impact of higher
flexibility values on plant performance will be greater
when operator discretion is high.
Hypothesis 4b. AMT objectives (as perceived by
operators) will moderate the relationship between con-
trol values (as perceived by managers) and plant per-
formance. Specifically, the positive impact of higher
control values on plant performance will be greater
when AMT objectives are more explicit.
3. Methods
Studies of AMT implementation, as well as of other
process innovations, pose challenging research trade-
offs (Dean et al., 1992; Lewis, 1998). Rare, large sample
studies enable comparisons, but introduce numerous
confounding variables (e.g., variations in type of new
technology, degree of innovation, timing of implemen-
tation, etc.). In contrast, case studies are prevalent,
particularly when examining innovation-supportive
culture. Yet cases enable depth at the expense of
generalizability. Responding to these tradeoffs, we
sought a relatively large, yet focused sample of
manufacturing plants that had recently implemented
the same AMT.
To identify potential respondents, we worked with
one of the world’s largest AMT manufacturers. Their
North American distributor sells computerized die/
mold machinery (e.g., machining centers, graphite
milling machines) to nearly 100 plants per year, and
provided us access to their entire customer network.
Engineers from the distributor served as a pilot for the
survey, helping ensure that items (all adapted from
previous studies and shown to be valid and reliable)
were appropriate and clearly worded for this particular
sample.
Focusing on computerized die/mold machinery
served two purposes. First, this is an excellent
example of AMT. Successful implementation can
dramatically enhance production quality and produc-
tivity, while enabling greater customization. The
second reason for examining this AMT is the diversity
of its users. Die/mold manufacturers vary widely in
terms of their size and sales, their primary industries,
and their investments in AMT. The initial sampling
pool included all 271 plants that had purchased
computerized die/mold machinery from the distribu-
tor within the past 3 years. The distributor provided
considerable a priori data, such as customer name and
address, primary contact, and AMT purchased (six
closely related machines).
875
The initial pool was well targeted toward our
objectives, but limited in size. To increase our response
rate, we paid close attention to data collection
techniques. First, we sent the survey packet to the
primary contact at each customer plant, typically the
plant manager. The packet included a cover letter from
the distributor, introducing the survey and requesting
responses; a survey and return envelope for the manager
overseeing the implementation (usually completed by
the primary contact); and a separate survey and return
envelope for an AMT operator. The cover letter asked
the manager to provide the operator survey to the
operator working most closely with the new AMT [the
distributor’s engineers stressed that a new AMT rarely
had more than one to three primary operators, offering
limited choice for managers in choosing an operator].
One week later, all contacts received a follow-up letter.
A second follow-up letter and survey packet were sent 1
month later to plants that had not yet responded. The
final sample consisted of 110 plants (each with
completed, usable surveys from both a manager and
an operator, for a response rate of 40.6%). Table 1 offers
a profile of the resulting sample, illustrating its
diversity.
3.1. Value profiles
Following McDermott and Stock (1999), we
measured value profiles using Likert-scale items
consistent with the competing values framework (Quinn
and Rohrbaugh, 1983). Variables were calculated
separately for managers and operators. Flexibility
values were measured by averaging responses to five
items, which gauge the degree to which the manager or
operator perceived organizational values of empower-
ment, growth, change, creativity and flexibility (e.g., To
what extent does your plant value empowerment? 1: low
emphasis to 7: high emphasis). Control values were
measured by averaging responses to four similar items
that assess to what extent the manager or operator
perceived organizational values of stability, efficiency,
predictability and control (e.g., to what extent does your
plant value stability? 1: low emphasis to 7: high
emphasis).
3.2. Value congruence
Traditional measures of congruence, such as score
differences, have been criticized for conceptual
ambiguity and discarded information (Edwards,
1993). For this study, polynomial regression analyses
provided a more sophisticated means of examining
876 S. Khazanchi et al. / Journal of Operations Management 25 (2007) 871–884

Table 1 indicating the extent to which they felt empowered to

Sample profile determine the timing, method and pace of their work on
Characteristics Grouping Number Percentage the AMT machinery. Items assessed operators’ agree-
Primary Primary metal 11 10.0 ment with descriptions of their AMT work (e.g., The
industry(ies)— industry (SIC machine operators can vary how they do their work. 1:
check all that apply code: 33) strongly disagree to 7: strongly agree). AMT objectives
Industrial machinery 13 11.8 were calculated by averaging responses to three items,
& equipment (35)
indicating the extent to which the objectives of
Transportation 22 20.0
equipment (37) purchasing the new technology were well-defined,
Misc. manufacturing 21 19.1 long-term and supported by top management. Adapted
industries (39) from Zhao and Co (1997), these Likert-scale items
Fabricated metal 16 14.5 gauged to what extent operators agreed with descrip-
products (34)
tions (e.g., the objectives of the new machine were well-
Electronic & other 18 16.4
electric equipment defined. 1: strongly disagree to 7: strongly agree).
(36)
Instruments and 10 9.10 3.4. Plant performance
related products
(38)
We measured plant performance by averaging
Other primary 40 36.0
industries managers’ responses to five, Likert-scale items. These
items gauged the extent to which performance has
Sales (in million) $0–5 43 39.1
$5–10 26 23.6
changed since implementation: ‘‘Please indicate how
$10–20 17 15.5 much your plant’s performance has changed over the
$20+ 24 21.8 past 2 years in the area of (product quality, scrap
Plant size 1–200 84 76.3 minimization, on-time delivery, equipment utilization,
(total number of full- or manufacturing lead-time)’’ (1: significant decline to
time employees) 4: no change to 7: significant improvement). Such
201–500 12 10.9 subjective measures potentially introduce manager bias,
501+ 14 12.7
but are common practice in manufacturing and AMT
AMT investment (total $0–150,000 22 20.0 research for two reasons (see Boyer et al., 1997). First,
investment in the studies find that respondents are less willing to provide
implementation)
$150,000–300,000 32 29.1
objective performance data. Second, more objective
$300,000–500,000 33 30.0 measures may limit the accuracy and comparability of
$500,000+ 23 20.9 responses (Dess and Robinson, 1984; Porter, 1979). For
Timing (months 0–9 34 30.9 example, other AMT studies use these same indicators,
since last AMT stating that a host of contextual factors impede the
implementation) usefulness of objective data (e.g., different metrics for
9–14 22 20.0 quality, efficiency, productivity) (Dean and Snell, 1991;
14–25 31 28.2
Zhao and Co, 1997). Ketokivi and Schroeder’s (2004)
25–36 23 20.9
extensive multitrait–multimethod analysis concludes
that such perceptional performance measures satisfy
congruence through the use of three-dimensional reliability and validity requirements.
surface graphs (Edwards and Parry, 1993). Our use of
this approach is detailed in Section 4. 3.5. Control variables

3.3. Practices A major challenge of culture research is the

To examine potentially complementary practices, we
asked operators – those using the new technology on a
daily basis – a series of survey questions. Operator
discretion was measured using Likert-scale items
adapted from Jackson et al. (1993). We calculated this
variable by averaging operators’ responses to six items,
potentially confounding effects of numerous, interwoven
factors, requiring thoughtful inclusion of control vari-
ables. As our focus is on the plant level, we included three
variables that could represent important, contextual
differences and contingencies that might influence plant
performance following AMT implementation: plant size,
AMT investment, and AMT training. Plant size denotes
 S. Khazanchi et al. / Journal of Operations Management 25 (2007) 871–884 877

the total number of full-time employees. AMT investment
was measured by asking survey respondents how much
their plant had invested in implementing the new
machinery, including training, related software pur-
chases and development, test runs, etc., but not including
the actual cost of the machinery. Finally, AMT training
was measured using a single item, indicating the degree
of training received on the new AMT prior to the
implementation (plant employees received extensive
training on the new machine prior to implementation. 1:
strongly disagree to 7: strongly agree).
4. Results
Correlations and descriptive statistics variables are
reported in Table 2. Cronbach’s alphas, shown along
diagonal, indicate that the measures are reliable (alphas
from .68 to .81).
Using regression analyses, we first tested the impact
of flexibility and control value profiles on plant
performance (see Table 3). Plant size, AMT investment,
and AMT training acted as controls in the base model.
When entered separately, flexibility and control values
were significantly and positively correlated to perfor-
mance. Yet when both variables were entered into the
full model only flexibility values remained significant,
offering partial support for Hypotheses 1a and 1b.
Post hoc analysis suggests that flexibility values
completely mediated the control values–plant perfor-
mance relationship. The mediation path was as follows:
control values to flexibility values, flexibility values to
plant performance. Following Baron and Kenny (1986),
we estimated three equations: (1) regress flexibility
values on control values ( p < .01; see Table 2), (2)
regress plant performance on control values ( p < .01;
see Table 3), and (3) regress plant performance on
flexibility and control values (only flexibility values
significant, p < .01; see Table 3). Hence, all conditions
for complete mediation were satisfied.
Next, we tested value congruence. Edwards and
Parry (1993) and Kalliath et al. (1999) detail serious
flaws of the traditional congruence methods: score
difference and square of score difference. Both methods
discard vital information by reducing two data points

Table 2
Descriptive statistics and correlations
Variable Mean S.D. 1 2 3 4 5 6 7 8 9 10
1. Plant size 1.34 .68
2. AMT investment 2.58 1.03 .12
3. AMT training 3.68 1.86 .03 .14 n.a
4. Flexibility values  mgrs 5.34 .92 .02 .02 .17y .68
5. Flexibility values  oprs 5.02 1.09 .07 .16 .14 .25 * .78
6. Control values  mgrs 5.52 .92 .06 .20* .13 .50 ** .22* .77
7. Control values  oprs 5.30 1.02 .20y .01 .13 .20 .74** .31 ** .78
8. AMT objectives 6.01 1.00 .04 .13 .06 .06 .31** .10 33** .72
9. Operators discretion 5.21 1.22 .08 .19y .02 .17 .04 .05 .04 .12 .81
10. Plant performance 5.13 0.73 .02 .06 .21 * 37 ** .12 .26 ** .09 .20y .09 .76
Cronbach’s alpha provided along the diagonal, shown in bold (n.a. signifies that alpha is not appropriate for scale comprised of less than three items).
**
p < .01.
*
p < .05.
y
p < .10.

Table 3
Regression analyses: flexibility and control value profiles (Hypotheses 1a and 1b)
Variable Base model Flexibility model Control model Full model
2
R .07 .19 .15 .20
Adj. R 2 .04 .16 .12 .17
F 2.42y 6.17** 4.57** 5.34**
Plant size .12 .17y .15 .17
AMT investment .10 .10 .09 .10
AMT training .19 * .13 .22 ** .15y
Flexibility values  mgrs .37 ** .29 **
Control values  mgrs .29 ** .14
Dependent variable is plant performance yp < .10; *p < .05; **
p < .01.
878 S. Khazanchi et al. / Journal of Operations Management 25 (2007) 871–884

into one, and do not predict outcomes at various levels Table 4

of congruence. In response, researchers increasingly Polynomial analysis: flexibility value congruence (Hypotheses 2a and
2b)
employ polynomial regression analysis (see Edwards
and Parry, 1993). Variable Model 1 Model 2
Polynomial regression analysis involves polynomial Flexibility values  mgrs Betas (S.E.) Betas (S.E.) 
regression equations and response surface methodology. .30 (.09)** .32 (.98)
Analysis first requires regressing the dependent variable Flexibility values  oprs .02 (.07) .21 (.57)
Flexibility values  .01 (.09)
on the individual main terms to determine main effects
mgrs squared
and overall model significance. Higher order terms – the Flexibility values  .04 (.06)
quadratic and interaction terms – are introduced only if oprs squared
main effects and the model are significant. The significant Flexibility values  .12y (.06)
beta coefficients of the full model then are used to mgrs  flexibility
values  oprs
compute scores for the linear and curvilinear slopes of
R .38 .42
two lines. The line of perfect fit signifies complete R2 .14 .18
congruence (e.g., managers’ perceptions of flexibility F 6.78** 3.14*
values = operators’ perceptions of flexibility values), DR2 .04
whereas, the line of misfit represents incongruence. DF 1.14
These scores are then used to plot the three-dimensional Slopes for line of perfect fit; linear: a1 = b1 + b2 = .53; Curvilinear:
graph needed to interpret findings of polynomial a2 = b3 + b4 + b5 = .072; Slopes for line of misfit; linear:
regression. This process enabled us to test the overall x1 = b1  b2 = .11; Curvilinear: x2 = b3  b4 + b5 = .15; yp < .10;
*
p < .05; **p < .01.
impact of value congruence on plant performance, and
how that impact varies at different levels of values (e.g., at
high versus low flexibility values). Following guidelines specified by Edwards and
Applying this method to test Hypothesis 2, we Parry (1993), we regressed plant performance on the
regressed plant performance on flexibility values main effects of flexibility values of managers and
perceived by managers (FM), flexibility values per- operators (model 1). We then included the cross-product
ceived by operators (FO), the interaction between FM of flexibility values of managers and of operators and
and FO, the square of FM, and the square of FO. Z the square of flexibility values of managers and of
represents the dependent variable (plant performance), operators (model 2). Results shown in Table 4 indicate
while e represents the error term. The polynomial that value congruence explained significant variance
regression equation is: (R2 = .18, p < .05).
Our findings can be interpreted using the three-
Z ¼ b0 þ b1 ðFMÞ þ b2 ðFOÞ þ b3 PðFOÞ2 þ b4 ðFMÞ dimensional graph shown in Fig. 1. The dotted line
 ðFOÞ þ b5 ðFMÞ2 þ e represents the line of misfit. At the left-hand extreme,

Fig. 1. Flexibility value congruence predicting plant performance.

 S. Khazanchi et al. / Journal of Operations Management 25 (2007) 871–884 879

coordinates of this line for flexibility values of Table 5

managers and operators are (7, 1). As one moves along Polynomial analysis: control value congruence (b)
the line of misfit toward the line of perfect fit (solid Variable Model 1 Model 2
line), congruence between managers and operators’ Control values  mgrs Betas (S.E.) N.S.
flexibility values rises. In other words, congruence .22 (.09)**
increases along the line of misfit, until it reaches a point Control values  oprs .01 (.08)
of intersection with the line of perfect fit. Beyond the Control values 
mgrs squared
line of perfect fit, congruence decreases.
Control values 
As predicted, plant performance is highest along the oprs squared
line of perfect fit. Plant performance rises as congruence Control values 
between flexibility values of managers and operators mgrs  control
increases, indicated by the upward slope of the line of values  oprs
R .27
misfit (dotted line) as it moves toward the line of perfect
R2 .07
fit (left-hand side of solid line). Furthermore, plant F 3.16**
performance declines as congruence between managers DR 2
and operators’ values decreases. This trend is indicated DF
by the downward slope of the line of misfit as it moves Slopes for line of perfect fit; linear: a1 = b1 + b2 = .23; Slopes for line
away from the line of perfect fit. These findings support of misfit; linear: a1 = b1 + b2 = .21. Note: no higher order effects were
Hypothesis 2a. significant; only main effects can be estimated. +p < .10; *p < .05;
**
Hypothesis 2b, however, is not supported. There is p < .01.
no significant variation in performance along the line of
perfect fit, indicating that congruence at varying levels centered the measures of flexibility values and of
does not impact performance. In other words, high operator discretion on the mean. Next, we used the
flexibility values of both managers and operators do not unstandarized beta coefficients and constants from the
necessarily translate into better performance than saturated regression equation to plot the relation
congruent, but low flexibility values. between operator discretion and performance at
Similarly, the polynomial equation for control values different levels (i.e., one standard deviation above the
is represented as follows: mean indicated that the managers perceived high
flexibility values; one standard deviation below the
Z ¼ b0 þ b1 ðCMÞ þ b2 ðCOÞ þ b3 PðCOÞ2 þ b4 ðCMÞ mean indicated low flexibility values). The interaction
 ðCOÞ þ b5 ðCMÞ2 þ e illustrated in Fig. 2, however, is contrary to Hypothesis
3a. When managers perceived high flexibility values,
To test Hypothesis 3, we first regressed plant perfor- plant performance decreased with greater levels of
mance on the main effects of managers and operators’ operator discretion.
control values (model 1). Although the main effect for
operators’ control values was found to be significant, no
Table 6
higher order effects were found to be significant (model Regression analysis: value–practice interactions (Hypotheses 4a and
2). Therefore, results shown in Table 5 do not support 4b)
Hypotheses 3a and 3b. Variable Flexibility Control Full
Table 6 summarizes the regression analysis of value– 2
practice interactions. We examined whether the R .18 .10 .23
Adj. R 2 .15 .07 .17
relationship between values (as perceived by managers) F 6.18 3.12 3.97
and plant performance is moderated by complementary p .001 .03 .002
practices (as perceived by operators). Specifically, we Flexibility values  mgrs 1.23** 1.18**
tested for interaction effects between flexibility values Operator discretion 1.13* 1.66*
and operator discretion, and between control values and Flexibility values  1.52* 1.57*
mgrs  operator
AMT objectives. We found support for Hypothesis 4a, discretion
but not for Hypothesis 4b. Results indicate a statistically Control values  mgrs .60 .36
significant interaction effect between flexibility values AMT objectives .50 .26
and operator discretion. Control values  .51 .25
To interpret the interaction effect, we applied the mgrs  AMT objectives
procedures detailed by Aiken and West (1991). We p < .10; *p < .05; **
p < .01.
880 S. Khazanchi et al. / Journal of Operations Management 25 (2007) 871–884

5. Discussion

This study seeks to deepen understandings of

Fig. 2. Interaction between flexibility values and operator discretion.
innovation-supportive culture by exploring the impact
of organizational values. Specifically, we examined the
role of value profiles, value congruence and value–
practice interactions. We now discuss three, related
findings (for summary, see Table 7).
First, innovation studies often note the need for
value profiles that stress flexibility and control (e.g.,
Dougherty, 1996; Quinn and Kimberly, 1984). At
first glance, our results appear more one-sided. Only
a flexibility value profile seems critical for AMT
implementation—a finding predicted by some
theorists (e.g., Cleland et al., 1995; Zammuto
and O’Connor, 1992). Our exploratory, post hoc
analysis, however, suggests that control values may
play a more subtle role as flexibility values

Table 7
Summary findings
Value profile
H1a Flexibility values (value profile perceived by managers) Partial supporta
will be positively related to plant performance
H1b Control values (value profile perceived by managers) Partial supporta
will be positively related to plant performance
Value congruence
H2a Plant performance will be positively related to the Supported
congruence between flexibility values as perceived
by manager and by operators
H2b Plant performance will be positively related to the Not-supported
level of flexibility values perceived by managers
and operators
H3a Plant performance will be positively related to the Not-supported
congruence between control values perceived by
managers and by operators
H3b Plant performance will be positively related to the Not-supported
level of control values as perceived by managers
and operators
Value–practice interactions
H4a Operator discretion (as perceived by operators) Supported
will moderate the relationship between flexibility
values (as perceived by managers) and plant performance.
Specifically, the positive impact of higher flexibility
values on plant performance will be greater when
operator discretion is high
H4b AMT objectives (as perceived by operators) Not-supported
will moderate the relationship between control
values (as perceived by managers) and plant performance.
Specifically, the positive impact of higher control values
on plant performance will be greater when AMT objectives
are more explicit
a
Post hoc analysis revealed complete mediation effect (see Table 3).
 S. Khazanchi et al. / Journal of Operations Management 25 (2007) 871–884
completely mediated the control values–performance
relationship.
One explanation is that control values enable
flexibility values and their benefits. Flexibility values
encourage employee empowerment and creative free-
dom. Yet, as Mills and Ungson (2003) explain,
managers may avoid flexibility for fear of losing
control. A foundation of control values and supporting
practices may alleviate such concerns, as stable routines
and explicit goals, for example, facilitate trust in
operators to innovate within appropriate boundaries.
The enabling potential of control values is consistent
with improvisation approaches to innovation. For
instance, Lewis et al. (2002) stress that control aids
improvisation during product development. They find
that a planned style of project management (e.g., formal
reviews, milestones) provides a vital framework for
brainstorming and experimentation. Moreover, control
can aid managers in monitoring and evaluating the
impact of innovation (Dougherty, 1996). For example,
Spear and Bowen (1999) state that the Toyota
Production Process – a system that enables continuous
process innovation – requires a culture of discipline
(e.g., a mindset and routines that stress the monitoring
of quality to identify problems and spur creative
problem solving).
Second, although we find that flexibility values most
directly affect innovation, results regarding value
congruence elaborate this understanding. Specifically,
the more managers and operators shared perceptions of
flexibility values (high or low), the more plant
performance improved after AMT implementation. In
comparison, congruence around control values was not
found to significantly affect plant performance. One
explanation is that the ambiguity of flexibility-related
norms (e.g., empowerment and creativity) increases the
importance of shared values, whereas, norms influenced
by control values (e.g., efficient routines) may be
promoted via explicit policies and procedures. For
example, imagine a case where managers perceive high
flexibility values, while operators perceive low flex-
ibility values. Managers may presume that operators are
creatively debugging the AMT as needed, making
continuous adjustments that improve performance.
Meanwhile, operators assume that creative problem
solving is a low priority. In response, operators view
themselves – then act – as ‘‘button-pushers’’ and
‘‘screen watchers’’ for the new automation. Managers
become frustrated by this lack of initiative, while
operators feel increasingly alienated from their work.
The resulting vicious spiral impedes performance.
While extreme, studies depict such scenarios in plants
881
implementing AMT (e.g., Cleland et al., 1995; Kelley,
1990).
Collectively, our findings suggest that the role of
control values, however, is complex. For instance,
findings suggest that when perceived by managers,
control values work through flexibility values to impact
plant performance. In contrast, when operators’
perceptions of control values and practices are taken
into account, as revealed through the insignificant roles
of control value congruence and control value–practices
interaction (see Table 7), control does not seem to affect
performance. A possible explanation is that control
perceptions reflect distinct managerial and operator
sub-cultures, whereas, perceptions of flexibility may be
more reflective of overarching organizational culture
(e.g., Cameron and Quinn, 1999; Von Meier, 1999).
Finally, the most counterintuitive finding involves
the interaction between flexibility values and operator
discretion. Results indicate that performance is highest
when flexibility values are high and operator discretion
is low. This combination suggests a strong mixed
message: ‘‘the organization values empowerment, but
operators cannot determine the timing, methods and
pace of their own work.’’ Historically, researchers have
stressed that a key issue when implementing AMT is
operator discretion—determining whether operators
manage the AMT (making day-to-day decisions
regarding its use) or whether engineers program the
machinery and operators simply monitor operations.
The former task design signals a culture of flexibility,
while the latter indicates a deskilling approach (Cleland
et al., 1995; Kelley, 1990).
Our study, however, supports an alternative view of
operator discretion. As Dean et al. (1992) explain in
AMT contexts the most critical discretion entails
autonomous problem solving away from the machin-
ery. Minimizing operators’ machine responsibilities
allows focus on higher impact issues—those that
foster continuous process innovation. For instance,
operator teams may put their knowledge of the factory
floor to greater use by informing the design of the
production system (Adler, 1993; Kern and Schumann,
1992). Indeed, Garud and Kotha (1994) propose that
flexible manufacturing demands such adaptive cap-
abilities. A related explanation, however, is that there
may be a time lag in the need for operator discretion.
The potential of AMT requires the capacity to learn.
Over time, operators, engineers and managers may
become increasingly capable of revising the flow and
layout of production activities (e.g., inventory,
loading/unloading of machinery, etc.) for ongoing
improvement.
882 S. Khazanchi et al. / Journal of Operations Management 25 (2007) 871–884
6. Implications
This study explores innovation-supportive culture by
examining the basic building block of organizational
values. For researchers, findings offer theoretical
insights and may spark subsequent studies. Consistent
with past work, this research supports and elaborates a
paradoxical view of innovation-supportive culture.
Specifically, results suggest possible, underlying
sources of paradox. For example, we found that
flexibility values may mediate the role of control
values. Flexibility values foster a culture of experi-
mentation and empowerment, whereas, control values
may set boundaries that facilitate managerial trust and
evaluation. Further, while flexibility values enable
operators to engage in creative problem solving or
debug routine machine-related problems (e.g., Zam-
muto and O’Connor, 1992), operators may see control
as inhibiting innovation. Thus, innovation-supportive
culture may appear paradoxical because of flexibility
and control co-existing in underlying values and
practices, but also may stem from conflicting views
held by occupational and hierarchical sub-cultures
within the organization.
Future work could elaborate the paradoxical inter-
play between flexibility and control values, investigat-
ing whether this relationship varies during the
innovation process or in the presence of other contextual
factors (e.g., organizational structure, environmental
uncertainty). Subsequent studies might suggest how
managers can foster shared values, while identifying
practices that reinforce those values to fuel innovation.
For instance, in AMT settings, flexibility values may
need to be paired with more specialized conceptualiza-
tions of operator discretion or related practices of
empowerment. Lastly, research could benefit by going
beyond managers and operators to study other groups
involved in process innovation, such as suppliers and
customers (Ettlie and Reza, 1992).
Similarly, our results offer insights for managers. For
example, this study highlights the paradoxical need for
control and flexibility value profiles. Such findings
encourage managers to avoid viewing such values as
conflicting, seeking instead to empower employees and
to establish supporting policies and systems. In essence,
managers should nurture an organization that offers
explicit controls for evaluating and most decision-
making, but also offers the flexibility for operators to
depart from routine work procedures. Likewise,
effective manages may provide independence, while
interacting frequently. Such managers may encourage
AMT operators to be creative and autonomous problem
solvers, while interacting with operators frequently to
set limits and help codify the learning that occurs via
experimentation.
To illustrate, we offer an expanded example from one
firm. Following AMT implementation, the company
used a 2-month training and ramp-up period. Once
‘‘steady state’’ production was reached, the company
continued to hold monthly meetings with all of the
equipment operators and the department supervisor.
These meetings were used to discuss problems with the
equipment, brainstorm for solutions and improved
approaches, and document the learning that occurred.
The goal was to encourage flexibility, but also to
develop controls that could be shared with future users
of the equipment and to standardize the knowledge.
7. Limitations
Limitations of our study also suggest research needs
and opportunities. For instance, our subjective measures
of plant performance present a serious limitation. As
with any subjective measures of performance, caution
must be exerted in interpreting the results because of the
possibility of managerial bias. Findings should be
replicated in future studies that use objective measures,
such as quality improvement, and/or multiple data
sources, such as suppliers and customers’ assessments.
Another limitation is that we examined innovation
within the context of AMT. Focusing on a particular
process innovation reduces the effect of confounding
variables, but raises concerns over generalizability.
Subsequent work may attempt to replicate these
findings in different settings of process or product
innovation. While controlling for variables such as plant
size, AMT investment, and AMT training build
confidence in our findings, future work could benefit
from including other indicators of a firm’s support for
manufacturing R&D. For example, Ettlie and Reza
(1992) note that integrating mechanisms, such as
centers for advanced manufacturing engineering, may
serve as enablers of process innovation.
Furthermore, the cross-sectional nature of this study
calls for future longitudinal studies that might account
for potential lagged effects associated with process
innovation. For instance, it might prove insightful to
investigate the changing roles of operators and
managers over time. Finally, our research design relied
on self-report surveys, raising the threat of common
method variance. Gathering responses from both
managers and operators builds confidence, but cannot
eliminate concerns. Future studies could benefit from
using different respondents to examine complementary
 S. Khazanchi et al. / Journal of Operations Management 25 (2007) 871–884
practices and plant performance, and from supplement-
ing subjective measures with more objective data. In
addition, the proliferation of innovation research poses
tremendous potential. As qualitative and quantitative
studies accumulate, triangulating results may deepen
understandings (Lewis, 1998). Weaving together stu-
dies of innovation and culture may identify supportive
configurations of values and practices. As Jassawalla
and Sashittal explain, ‘‘the challenge of developing
significantly higher levels of collaboration, creativity,
and innovation appears to relate to the way in which the
fuzzy, amorphous nature of culture is integrated with the
hard, cold analysis of technology, customers, markets,
and competitors’’ (2002: 53).
References
Adler, P.S., 1993. Time-and-motion regained. Harvard Business
Review 97–108.
Aiken, L.S., West, S.G., 1991. Multiple Regression: Testing and
Interpreting Interactions. Sage, Newbury Park, CA.
Argyris, C., Schön, 1974. Theory in Practice: Increasing Professional
Effectiveness. Jossey-Bass, San Francisco.
Baer, M., Frese, M., 2003. Innovation is not enough: climates for
initiative and psychological safety, process innovations, and firm
performance. Journal of Organizational Behavior 24 (1), 45–69.
Baron, R.M., Kenny, D.A., 1986. The moderating-mediator variable
distinction in social psychological research: conceptual, strategic
and statistical considerations. Journal of Personality and Social
Psychology 51 (6), 1173–1182.
Bates, K.A., Amundson, S.D., Schroeder, R.G., Morris, W.T., 1995.
The crucial interrelationship between manufacturing strategy and
organizational culture. Management Science 41 (10), 1565–1580.
Black, L.J., Carlile, P.R., Repenning, N.P., 2004. A dynamic theory of
expertise and occupational boundaries in new technology imple-
mentation: building on Barley’s study of CT scanning. Adminis-
trative Science Quarterly 49 (4), 572–607.
Boyer, K.K., Leong, G., Ward, P.T., Krajewski, L., 1997. Unlocking
the potential of advanced manufacturing technologies. Journal of
Operations Management 15 (4), 331–347.
Cameron, K.S., Quinn, R.E., 1999. Diagnosing and Changing Orga-
nizational Culture: Based on Competing Values Framework.
Addison-Wesley, Reading, MA.
Chatman, J.A., 1989. Improving interactional organizational research:
a model of person-organization fit. Academy of Management
Review 14 (3), 333–349.
Cleland, D.I., Bidanda, B., Chung, C.A., 1995. Human issues in
technology implementation. Industrial Management 37 (4), 22–
26.
Christensen, C.M., 1999. Innovation and the General Manager.
McGraw-Hill.
Damanpour, R., Gopalakrishnan, S., 2001. The dynamics of the
adoption of product and process innovations in organizations.
Journal of Management Studies 38 (1), 45–66.
Deal, T.E., Kennedy, A.A., 1982. Corporate Cultures. Perseus Books,
Reading, MA.
Dean, J.W., Snell, S.A., 1991. Integrated manufacturing and job
design: moderating effects of organizational inertia. Academy
of Management Journal 34 (4), 459–480.
883
Dean, J.W., Yoon, S.J., Susman, G.I., 1992. Advanced manufacturing
technology and organization structure: empowerment or
subordination? Organization Science 3 (2), 203–229.
Denison, D., 1989. Corporate Culture and Organizational Effective-
ness. John Wiley, New York.
Dess, G.R., Robinson Jr., R.B., 1984. Measuring organizational
performance in the absence of objective measures: the case of
the privately held firm and conglomerate business unit. Strategic
Management Journal 5 (3), 265–274.
Detert, J.R., Schroeder, R.G., Mauriel, J.L., 2000. A framework for
linking culture and improvement initiatives in organizations.
Academy of Management Review 25 (4), 850–863.
Dougherty, D., 1996. Organizing for innovation. In: Clegg, S.R.,
Handy, C., Nord, W.R. (Eds.), Handbook of Organization Studies.
Sage, Thousand Oaks, CA, pp. 424–439.
Edwards, J.R., 1993. Problems with the use of profile similarity
indices in the study of congruence in organizational research.
Personnel Psychology 46 (3), 641–666.
Edwards, J.R., Parry, M.E., 1993. On the use of polynomial regression
equations as an alternative to difference scores in organizational
research. Academy of Management Journal 36 (6), 1577–1613.
Ettlie, J.E., Reza, E.M., 1992. Organizational integration and process
innovation. Academy of Management Journal 35 (4), 795–828.
Frohman, A.L., 1998. Building a culture for innovation. Research
Technology Management 41 (2), 9–12.
Garud, R., Kotha, S., 1994. Using the brain as a metaphor to model
flexible production systems. Academy of Management Review 19
(4), 671–698.
Gordon, G.G., DiTomaso, N., 1992. Predicting corporate performance
from organizational culture. Journal of Management Studies 29
(6), 783–799.
Hayes, R.H., Wheelwright, S.C., Clark, K.B., 1988. Dynamic Man-
ufacturing: Creating the Learning Organization. Free Press, New
York.
Higgins, J.M., McAllaster, C., 2002. Want innovation? Then use
cultural artifacts that support it. Organizational Dynamics 31
(1), 74–85.
Hofstede, G., Neuijen, B., Ohayv, D.D., Sangers, B., 1990. Measur-
ing organizational cultures: a qualitative and quantitative study
across twenty cases. Administrative Science Quarterly 35 (2),
286–316.
Jackson, P.R., Wall, T.D., Martin, R., David, K., 1993. New measures
of job control, cognitive demands, and production responsibility.
Journal of Applied Psychology 78 (5), 753–762.
Jassawalla, A.R., Sashittal, H.C., 2002. Cultures that support product
innovation processes. Academy of Management Executive 16 (3),
42–53.
Kalliath, T.J., Bluedorn, A.C., Strube, M., 1999. A test of value
congruence effects. Journal of Organizational Behavior 20 (7),
1175–1199.
Kelley, M.R., 1990. New process technology, job design, and work
organization: a contingency model. American Sociological
Review 55 (2), 191–208.
Kern, H., Schumann, M., 1992. New concepts of production and the
emergence of the systems controller. In: Adler, P.S. (Ed.), Tech-
nology and the Future of Work. Oxford University Press, New
York, pp. 111–148.
Ketokivi, M., Schroeder, R.G., 2004. Perceptual measures of perfor-
mance: fact or fiction? Journal of Operations Management 22/23,
247–265.
Kotter, J.P., Heskett, J.L., 1992. Corporate Culture and Performance.
Free Press, New York.
884 S. Khazanchi et al. / Journal of Operations Management 25 (2007) 871–884
Lewis, M.W., 1998. Iterative triangulation: a theory development
process using existing case studies. Journal of Operations Manage-
ment 16 (4), 455–469.
Lewis, M.W., Boyer, K.K., 2002. Factors impacting AMT implemen-
tation: an integrative and controlled study. Journal of Engineering
Technology Management 19 (2), 111–130.
Lewis, M.W., Welsh, A.W., Dehler, G.E., Green, S.G., 2002. Product
development tensions: exploring contrasting project management
styles. Academy of Management Journal 45 (3), 546–564.
McDermott, C.M., Stock, G.N., 1999. Organizational culture and
advanced manufacturing technology implementation. Journal of
Operations Management 17 (5), 521–533.
Mills, P.K., Ungson, G.R., 2003. Reassessing the limits of structural
empowerment: organizational constitution and trust as controls.
Academy of Management Review 28 (1), 143–154.
O’Reilly, C.A., Chatman, J., 1986. Organizational commitment and
psychological attachment: the effects of compliance, identifica-
tion, and internalization on prosocial behavior. Journal of Applied
Psychology 71 (3), 492–499.
Peters, T.J., Waterman, R.H., 1982. In Search of Excellence. Harper
Collins Trade, New York, NY.
Poole, M.S., Van de Ven, A.H., 1989. Using a paradox to build
management and organization theories. Academy of Management
Review 14 (4), 562–578.
Porter, M.E., 1979. The structure within industries and companies’
performance. Review of Economics and Statistics 61, 214–227.
Quinn, R.E., Kimberly, J.R., 1984. Paradox, planning, and persever-
ance: guidelines for managerial practice. In: Kimberly, J.R.,
Quinn, R.E. (Eds.), Managing Organizational Transitions. Down
Jones-Irwin, Homewood, IL, pp. 295–313.
Quinn, R.E., Rohrbaugh, J., 1983. A spatial model of effectiveness
criteria: towards a competing values approach to organizational
analysis. Management Science 29, 363–377.
Small, M., Yasin, M., 1997. Developing a framework for the effective
planning and implementation of advanced manufacturing. Inter-
national Journal of Operations and Production Management 17 (5/
6), 468–489.
Spear, S., Bowen, K.H., 1999. Decoding the DNA of the Toyota
production system. Harvard Business Review 77 (5), 96–107.
Tyre, M.J., Orlinkowski, W.J., 1993. Exploring opportunities for
technological improvements in organizations. Sloan Management
Review 35 (1), 13–27.
Von Meier, A., 1999. Occupational cultures as a challenge to tech-
nological innovation. IEEE Transactions of Engineering Manage-
ment 46 (1), 101–114.
Zammuto, R.F., O’Connor, E.J., 1992. Gaining advanced manufac-
turing technologies benefits: the role of organizational design
and culture. Academy of Management Review 17 (4), 701–
728.
Zhao, H., Co, H.C., 1997. Adoption and implementation of advanced
manufacturing technology in Singapore. International Journal of
Production Economics 48 (1), 7–13.