7311763 Leadership and Organizational Learnings Role on Innovation and Performance Lessons From...

8/3/2019 7311763 Leadership and Organizational Learnings Role on Innovation and Performance Lessons From Spain

1/23

Industrial Marketing Management 36 (2007) 349 359

Leadership and organizational learnings role on innovation and performance: Lessons from SpainJ. Alberto Aragon-Correa a,b,*, Vctor J. Garca-Morales b, Eulogio Cordon-Pozo b a

Rotterdam-Erasmus University, The Netherlands b University of Granada, Spain

Received 3 January 2005; received in revised form 7 April 2005; accepted 27 September 2005 Available online 15 December 2005

Abstract Leadership style has been traditionally emphasized as one of the most important individual influences on firm innovation. Scholars are now paying growing attention to the possibility that the collective capability of organizationallearning plays a key role in determining innovation. We propose that leadershipstyle, an individual feature, and organizational learning, a collective process, simultaneously and positively affect firm innovation. A structural equation model and data from 408 large firms in four sectors supported our hypotheses. Orga

nizational learning had a stronger direct influence on innovation than CEO transformational leadership for our sample; however, leadership had a strong, significant influence on organizational learning, indirectly affecting firm innovation.Additionally, innovation positively and significantly influenced performance. Organizational learning also positively affected performance, but interestingly mainly through innovation. D 2005 Published by Elsevier Inc.Keywords: Innovation; Organizational learning; Transformational leadership; Performance; Organizational capabilities

1. Introduction Market orientation and subsequent firm innovation are widely recognized to be essential for the survival and growth of organizations (Bello, Lohtia, & Sangtani, 2004; Damanpour & Gopalakrishnan, 2001; Hurley & Hult, 1998). Porter (1990) suggested that by the late twentieth century, most industrial econo

mies had moved to an innovation-driven stage, during which firms competed on how torapidly and profitably innovate. In this context, it is especially important togain a better understanding of factors influencing the successful development offirm innovations. Different definitions of innovation have been proposed (e.g.,Knight, 1967; Zaltman, Duncan, & Holbek, 1973). We accepted for our work the definition of innovation stated by the Product Development and Management Association (PDMA, 2004): A new idea, method, or device. The act of creating a new productor process. The act includes invention* Corresponding author. School of Economics and Business, University of Granada,Campus Cartuja, s.n., Granada 18071, Spain. Tel.: +34 958 24 23 54; fax: +34 958 24 62 22. E-mail addresses: [email protected] (J..A. Aragon-Correa), [email protected] (V.J. Garca-Morales), [email protected] (E. Cordon-Pozo). 0019-8501/$ - see front

matter D 2005 Published by Elsevier Inc. doi:10.1016/j.indmarman.2005.09.006

as well as the work required to bring an idea or concept into final form. Althoughfirm innovation is widely prescribed as a means to improve organizational performance, many firms do not or cannot properly develop it. Researchers have urgedattention to what makes it possible for firms to develop innovation, looking foranswers beyond semiautomatic stimulus-response processes (Zollo & Winter, 2002, p.341). Many authors have focused their attention on analyzing whether specific managerial characteristics influence the generation of innovation in organizations, while others have focused on analysis of organizational factors. We want to highlight the simultaneous influence of both kinds of factors. Leadership style has been emphasized as one of the most important individual influences on firm innovation, because leaders can directly decide to introduce new ideas into an orga

nization, set specific goals, and encourage innovation initiatives from subordinates (Harbone & Johne, 2003; McDonough, 2000; Sethi, 2000). Specifically, several writers have linked transformational leadership to innovation (e.g. Howell & Avoli


2/23


3/23

350

J.A. Aragon-Correa et al. / Industrial Marketing Management 36 (2007) 349 359

inspiring followers to pursue the vision; change or align systems to accommodatetheir vision rather than work within existing systems; and coach followers to take on greater responsibility for both their own and others development (Howell &

Avolio, 1993). Attention to the organizational influences on innovation is alsoimportant. Although several such influences have been analyzed, scholars are paying growing attention to the possibility that the collective capability of organizational learning plays a key role in determining innovation (Senge, 1990; Senge, Roberts, Ross, Smith, & Kleiner, 1994; Tushman & Nadler, 1986). Organizational learning has been defined as a collective capability based on experiential and cognitive processes and involving knowledge acquisition, knowledge sharing, and knowledge utilization (e.g., DiBella, Nevis, & Gould, 1996; Zollo & Winter, 2002). We propose that both collective (organizational learning) and individual (transformational leadership) factors influence firms to develop and implement organizational innovation. Many previous studies, although contributing significantly to the understanding of innovation, have not addressed how the efficacy of in

novation may vary with the simultaneous influence of different organizational factors and have not analyzed both direct and indirect influences (Van de Ven, 1993). Identifying and better understanding those influences will complement the general prescription that firms should innovate. Additionally, the ultimate purpose of firm innovation is new knowledge and new applications, especially those connected to organizational improvements, and many researchers have claimed a positive relationship between organizational learning and performance. We sought to reinforce this work by contributing to the analysis of the influence of innovation on performance. Further, we sought to show how the influence of organizationallearning on performance is strengthened by the generation of innovation. In this article, our focus is primarily on research questions that concern firm innovation. We first examine the nature and strength of transformational leadership and organizational learning as antecedents of firm innovation. We then investigate

whether firm innovation, organizational learning and transformational leadershipaffect financial performance. And finally, using these research findings, we develop a model of direct and indirect influences to guide future research in thisarena and offer managerial implications. 2. Framework and hypotheses Capabilities require that multiple characteristics be already embedded in a firm (Grant, 1991). Like any other capability, organizational innovation depends on the presence of capabilities by which firms synthesize and acquire knowledge resources andgenerate new applications from those resources (e.g., Calantone, Cavusgil, & Zhao, 2002; Celuch, Kasouf, & Peruvemba, 2002). All these antecedents have to be analyzed globally and integrated to achieve systemic thinking. In the following sections, we present a model consisting of five hypotheses about how transformational leadership and organizational learning simultaneously condition firm innovation. We also propose an indirect relationship between transformational leadership and innovation through organizational learning. We recognize that other variables might be considered in such a model; however, it was necessary to limit ourmodel to be able to offer empirical evidence for our arguments, and we chose these two factors to represent a focus on individual and on collective explanations for innovation activity, respectively. Our aim here was simultaneous consideration of these relevant antecedents of firm innovation. Additionally, we developed two hypotheses about innovations effect on performance. Fig. 1 illustrates theproposed model. 2.1. The influence of organizational learning on firm innovationMany works in the growing literature on organizational learning have noted a positive relationship between organizational learning and firm innovation (e.g., Calantone et al., 2002; Tushman & Nadler, 1986). Organizational learning supports

creativity (e.g., Sanchez & Mahoney, 1996), inspires new knowledge and ideas (e.g., Damanpour, 1991; Dishman &


4/23

1 Organizational Learning H3(+) 1 Transformational Leaders

ip

H4(+)

H1(+)

3 Performance

H2(+) 2 Innovation

H5(+)

Fig. 1. Hypot

esized model.


5/23


351

Pearson, 2003), and increases ability to understand and apply t

em (e.g., Damanpour, 1991). Generative learning, t

e most advanced form of organizational learning, occurs w

en an organization is willing to question long-

eld assumptions abo

ut its mission, customers, capabilities, or strategy and generate c

anges in itspractices, strategies, and values (e.g., Argyris & Sc

on, 1996; Senge, 1990). T

is kind of learning is a necessary underpinning for radical innovations in products and processes (Senge et al., 1994). T

ese ideas

ave recently begun to receive some empirical attention. Hurley and Hult (1998) focused on a large agency of t

e US federal government to s

ow t

at organizational innovativeness was positively associated wit

a culture t

at emp

asizes adaptation, innovation, and learning. Meeus, Oerlemans, and Hage (2001) analyzed a sample of innovator firms tos

ow t

at more comple innovative activities urged firms to coordinate and e c

ange information between users and producers, w

ic

implies strong interactive learning. T

ese arguments lead to our first

ypot

esis: Hypot

esis 1. Organizational learning positively influences firm innovation. 2.2. T

e influence of transfo

rmational leaders

ip on firm innovation Leaders play a significant role in s

aping firms potential to generate innovations by encouraging an appropriate environment and making decisions t

at promote successful generation and implementation of knowledge (Kanter, 1983; Van de Ven, 1993). T

e analysis of firm leaders c

aracteristics (e.g., education, background, personality, or attitudes)

as generatedwide attention from innovation researc

ers (Storey, 2000; Tus

man & Nadler, 1986). Style of leaders

ip

as been

ig

lig

ted as an especially important influence on innovation (Harbone & Jo

ne, 2003; Kanter, 1983; McDonoug

, 2000; Set

i, 2000). Transformational leaders

ip, w

ic

as been contrasted wit

Ftraditional_ or Ftransactional_ leaders

ip, includes a wide strategic vision about t

e advantages of c

ange and adaptation (Dess & Picken, 2000), significant interest in a communicative culture (Hult, Ferrell, Hurley, & Giunipero, 2000), attention to t

edevelopment of people (Barczak & Wilemon, 1992), and acceptance of mistakes (Sn

ell, 2001). It is important to

ig

lig

t t

at managers perceptions about t

eir own roles in t

eir organizations strongly influence t

eir capability to promote t

is kind of leaders

ip in an organization. Several features of transformational leaders

ip are relevant for firm innovation. Transformational leaders

ave an interactive vision, paying ma imum attention to effective communication and s

aringvalues (e.g., Adair, 1990; Quinn, 1988) and encouraging an appropriate environment for innovative teams (Tus

man & Nadler, 1986). T

ey support collective processes of organizational learning (Manz, Barstein, Hostager, & S

apiro, 1989), reciprocal trust between organization members and leaders (Scott & Bruce, 1994), and favorable attitudes toward

proactivity and risk (Lefebvre & Lefebvre, 1992). Transformational leaders perceive t

eir role more as coordination t

an as command and control (Barczak & Wilemon, 1992). All t

ese features toget

er allow a better understanding of t

e strong relations

ips between collaborative, innovative transformational leaders

ip and factors positively influencing organizational innovation (e.g., Farr & Ford, 1990; Kanter, 1983). Transformational leaders

ip is more often linked to successful innovation t

an is transactional leaders

ip (Dess & Picken, 2000; Manz et al., 1989). T

ese arguments lead to t

e ne t

ypot

esis: Hypot

esis 2. Transformational leaders

ip positively influences firm innovation. Simultaneously, it is important for our work to

ig

lig

t t

at transformational leaders

ip and organizational learning are also related. T

is circumstance implies indirect influences onorganizational innovation, influences t

at

ave usually been absent from previous researc

analysis. Many aut

ors

ave asserted relations

ips between leaders

ip and organizational learning (e.g., Senge, 1990; Senge et al., 1994; Tus

man &

Nadler, 1986). Traditional leaders

ip

as been c

aracterized as

ig

ly individualistic and asystematic and as making t

e learning of organizational teams difficult;

owever, transformational leaders

ip is focused on active promotion of empl


6/23

oyees participation in collective decisions and activities (Adair, 1990; Bass, 1991). Transformational leaders s

ould be able to build teams and provide t

em wit

direction, energy, and support for processes of c

ange and organizational learning (Blackler & McDonald, 2000; McDonoug

, 2000; Nadler & Tus

man, 1990). Morespecifically, transformational leaders

ip fuels organizational learning by promoting intellectual stimulation, inspirational motivation, and self-confidence among organization members (Coad & Berry, 1998). A capability for transformational

leaders

ip

as been even described as one of t

e most important means of developing learning organizations (e.g., Maani & Benton, 1999; Slater & Narver, 1995; Snell, 2001). T

us, we predict: Hypot

esis 3a. Transformational leaders

ip positively influences organizational learning. Hypot

esis 3b. Transformational leaders

ip positively and indirectly influences firm innovation t

roug

organizationallearning. 2.3. T

e influence on performance 2.3.1. Effects of organizational learning on performance T

e importance of organizational learning for a companys survival and effective performance

as been

ig

ly emp

asized in t

e literature (e.g., Argyris & Sc

on, 1996; Huber, 1991; Senge, 1990; Za

ay & Handfield, 2004). However, empirical analysis of t

is relations

ip

as been limited. Some recent works

ave begun to verify t

is positive relations

ip. Sc

roeder, Bates, and Junttila (2002) developed


7/23

352


resource-based

ypot

eses and s

owed a positive relations

ip between internal and e ternal learning and organizational performance in 164 manufacturing plants from si different countries. Bontis, Crossan, and Hulland (2002) used respondent

s in 32 Canadian funds to s

ow t

at Fstocks of learning_ at all organizational levels

ad a positive relations

ip wit

business performance. Finally, Za

ra, Ireland, and Hitt (2000) s

owed a strong relations

ip between international diversity and mode of market entry and t

e breadt

, dept

, and speed of a new venture firms tec

nological learning, especially w

en a firm undertook formal knowledge integration. T

us: Hypot

esis 4. Organizational learning will be positively associated wit

performance. 2.3.2. Effects of firm innovation on performance Firm innovation

as been traditionally focused on t

e improvement of organizational performance (Damanpour, 1991; Zaltman et al., 1973). Some previous work

as noted t

at only certain c

aracteristics of an innovation and not t

e innovation itself is positively linked to organizational performance (e.g., Danneels & Kleinsc

midt, 2001; Gopalakris

nan, 2000). In any case, t

e vast majority of previous public

ations agree t

at organizational innovation positively influences performance. For illustration, we now provide some e amples. Irwin, Hoffman, and Lamont (1998)used a resource-based view to s

ow t

e positive relations

ip between tec

nological innovations and organizational performance and stated t

at t

e innovation c

aracteristics of rarity, value, and inimitability moderated t

is relations

ip. Hurley and Hult (1998) demonstrated positive relations

ips between organizationalinnovation, a market orientation, and organizational learning and s

owed t

at all t

ese elements toget

er influenced t

e potential for good performance. Capron(1999) s

owed similar relation s

ips after mergers and acquisitions and, finally,Loof and Hes

mati (2002) s

owed t

e negative impact of avoiding innovations. Inview of t

e positive relations

ips seen in previous researc

, our

ypot

esis is: Hypot

esis 5. Firm innovation will positively influence performance. 3. Met

odology 3.1. Sample and procedures T

e sample of firms was randomly selected from

t

e Dun and Bradstreet 2001 database, w

ic

includes t

e 50,000 biggest companies operating in Spain. T

e final sample contained 900 firms in four wide categories: farming, manufacturing, construction, and services. We defined t

ese categories to control for confounding effects. C

oosing a sample of firms located in arelatively

omogeneous geograp

ic, cultural, legal, and political space enabledus to minimize t

e impact of variables t

at could not be controlled (Hofstede, 1980). T

e Spanis

market is relatively well developed, is w

olly integrated intot

e European Union, and

as

ad a

slig

tly better rate of growt

in recent years t

an t

e European market overall.However, Spain

as received relatively little attention from organizational researc

ers. Drawing on our interviews wit

five managers and si academics interested in t

e topic and familiar wit

t

e Spanis

market, we developed a structuredquestionnaire to investigate

ow organizations face learning and innovation issues. T

ese developmental interviewees did not provide data for t

e empirical investigation. We decided to use CEOs as our key informants since t

ey receive information from a wide range of departments and, t

erefore, are a very valuable source for evaluating aspects of organizations. T

ey also play a major role in forming and molding organizational c

aracteristics by determining t

e types of be

avior t

at are e pected and supported (Baer & Frese, 2003). In addition, use of CEOs meant t

at informants were similar across organizations, and t

us t

at t

eirlevels of influence in t

eir organizations was constant, w

ic

increased t

e validity of t

e variables measurements (Glick, 1985). Because t

e vast majority of t

e CEOs were native Spanis

speakers, t

e questionnaire was written in Spanis

to avoid any problem wit

t

e language. T

e questionnaires were mailed to t

e CEO

s of t

e 900 randomly selected firms along wit

a cover letter. We used t

is metod rat

er t

an interviews because a mailed survey enabled us to reac

a greaternumber of firms at a lower cost, put less pressure for an immediate response on


8/23

t

e potential informants, and gave respondents a greater feeling of autonomy. To reduce possible desirability bias, we promised t

at we would keep all individual responses completely confidential and confirmed t

at our analyses would be restricted to an aggregated level t

at would prevent t

e identification of any organization. We mailed eac

CEO w

o

ad not yet responded t

ree reminders. Four

undred twenty-t

ree CEOs finally answered t

e questionnaire but, because of missing values, only 408 questionnaires were included in t

e researc

. T

e response r

ate was 45.33%. We did not find significant differences in type of business or number of employees between t

e respondents and t

e sample or between early and late responders. Furt

ermore, since all measures were collected wit

t

e same survey instrument, we tested for t

e possibility of common met

od bias using Harmansone-factor test (e.g., Scott & Bruce, 1994). A principal components factor analysis on t

e questionnaire measurement items yielded four factors wit

eigenvalues greater t

an 1.0 t

at accounted for 66% of t

e total variance. Since several factors, as opposed to one factor, were identified, and since t

e first factor did not account for most of t

e variance, common met

od variance did not appear tobe present (Podsakoff & Organ, 1986). 3.2. Measures Scales are important in designing a survey instrument in management researc

. As no single measure can precisely capture be

avior, researc

ers usually combine two or more measures into a

scale to gauge eac

variable. Given t

at developing new scales is a comple

task, w

erever possible we


9/23


353

used pretested scales from past empirical studies to ensure t

eir validity and reliability. 3.2.1. Transformational leaders

ip T

e strategic literature includesresearc

t

at measures and evaluates transformational leaders

ip (e.g., Coad &

Berry, 1998; Hult et al., 2000; Podsakoff, Mackenzie, & Bommer, 1996). Style ofleaders

ip is broadly based on a managers assumption about

is/

er role in an organization. T

erefore, we drew five items from t

e scale designed by Podsakoff etal. (1996) to assess aspects of transformational leaders

ip. We asked CEOs w

et

er t

ey gave priority to seeking new opportunities for t

eir organizations; tried to develop a clear common view of final aims more t

an s

ort-term objectives;emp

asized motivating t

e rest of t

e company more t

an controlling; acted as t

e organizations leading force more t

an as supervisor; and, finally, coordinatedt

eir colleagues on t

e job. All t

e items in t

is paper used a Likert-type 7-point scale (1, totally disagree to 7, totally agree). A confirmatory 2 factor analv 5 = 14.25, normed fit inde [NFI] = 0.97, nonnormed fit inde [NNFI] = 0.96, goodness-of-fit inde [GFI] = 0.99, comparative fit inde [CFI] = 0.98, adjusted

goodness-of-fit inde

[AGFI] = 0.98) subsequently verified t

e scales unidimensionality and its

ig

validity and reliability (a = 0.850). 3.2.2. Organizationallearning T

e capability of organizational learning

as received muc

more t

eoretical t

an empirical attention. Additionally, t

ere are wide differences among t

e assumptions, procedures, and objectives of previous measures. We took measures from two previous scales t

at

ad close conceptual links wit

our researc

, reflected prior trends well, and

ad been verified in detail. We used t

e first two items of Kale, Sing

, and Perlmutters (2000) scale and added two items based onEdmondsons (1999) work. T

is four-item organizational learning scale asked respondents w

et

er, over t

e last 3 years, t

eir organizations

ad acquired muc

newand relevant knowledge, if organizational members

ad acquired critical capacities and skills, if organizational improvements

ad been influenced by t

e entryof new knowledge, and if t

eir organizations were learning organizations. T

is s

cale was similar to ot

er recently proposed measures of e

ternal and internal learning (e.g., Sc

roeder et al., 2002). We conducted a 2 confirmatory factor analysis to validate our scales (v 2 = 2.40, NFI = 0.99, NNFI = 0.99, GFI = 0.99, CFI = 0.99, AGFI = 0.99). Results s

owed t

at final scale was unidimensional and

ad

ig

reliability (a = 0.919). 3.2.3. Firm innovation Numerous researc

ers

ave analyzed organizations innovation using reliable and valid scales. We based ourscale on Miller and Friesens (1983) work. We first defined firm innovation (in contrast to industry or market innovation) for respondents and t

en asked t

em toevaluate

ow

ig

, relative to competitors, t

eir firms rates of new product/service introduction and c

anges in internal operating practices were

for t

e last 3 years. A confirmatory factor analysis s

owed t

at our scale was unidimensional and reliable (a = 0.777). We also included questions allowing t

eCEO respondents to offer precise quantitative data on organizational innovationand innovation radicality (e.g. number of new products in t

e last 3 years). Weincluded questions tapping bot

types of assessment in our interviews (subjective evaluation and quantitative data), but t

e managers were more open to offeringt

eir perceptions t

an to offering precise quantitative data (only 61 offered quantitative data). T

erefore, we tested t

e model using a perceptual measure offirm innovation in w

ic

eac

respondent rated

is or

er organizations innovation relative to t

at of ot

er firms in t

e. W

ere possible, we calculated t

e correlation between t

e objective and subjective data. T

ese were

ig

and statistically significant (0.76, p < 0.01). 3.2.4. Performance Having reviewed

ow performance was measured in different works of strategic researc

(e.g., Venkatraman &Ramanujan, 1986), we drew up an eig

t-item scale to measure organizational perf

ormance. T

e CEOs were asked to evaluate t

eir firms performance for t

e last 3 years, measured as return on assets, return on internal resources, and sales growt

in t

eir main products or services and markets. T

ey were also asked to compa


10/23

re t

ese measures wit

t

eir principal competitors performance, noting w

ic

wereabove t

e mean. T

e use of scales evaluating performance in comparison wit

main competitors is one of t

e practices most widely used in recent studies to provide an objective reference for sampled managers (Steensman & Corley, 2000). Manyresearc

ers

ave used managers subjective perceptions to measure beneficial outcomes for firms. Ot

ers

ave preferred objective data, suc

as return on assets.Sc

olars

ave widely establis

ed t

at t

ere is a

ig

correlation and concurrent

validity between objective and subjective data on performance, w

ic

implies t

at bot

are valid w

en calculating a firms performance (e.g., Dess & Robinson, 1984; Venkatraman & Ramanujan, 1986). We included questions involving bot

types of assessment in our interviews, but t

e managers were more open to offering t

eir general views t

an to offering precise quantitative data; t

erefore, we testedt

e model using a perceptual measure of financial performance (t

ree items, seven-point scale). W

en possible, we calculated t

e correlation between objectiveand subjective data, and t

ese were

ig

and significant. A confirmatory factoranalysis 2 (v 20 = 285.95, NFI = 0.92, NNFI = 0.90, GFI = 0.96, CFI = 0.93, AGFI= 0.92) s

owed t

at t

e scale was unidimensional and

ig

ly reliable (a = 0.889). 3.2.5. Control variables Size may affect an organizations ability to learn (DiBella et al., 1996; Tsang, 1997) or to innovate (Damanpour, 1992; Vossen, 1998).

T

e size indicators initially used for t

is researc

were firm income and number of employees. Information for t

ese variables was gat

ered t

roug

t

e surveyand validated using Dun and Bradstreet; correlation coefficients between t

ese sources were strong and significant.


11/23

354


Table 1 Means, standard deviations and correlation Variable 1. 2. 3. 4. 5. Transformational Leaders

ip Organizational Learning Innovation Performance Size Mean5.22 5.37 4.67 4.83 3.4 S.D. 0.94 1.14 1.19 1.02 1.68 1 1.000 0.473*** 0.387***

0.456*** 0.068 2 1.000 0.587*** 0.488*** 0.010 3 4 5

1.000 0.509*** 0.071

1.000 0.009

1.000

*** Significant at p < 0.001 (two-tailed); n = 408.fs

In addition, because size and income were

ig

ly correlated, we used number of employees only in our model (Weaver, Trevino, & Coc

ran, 1999). Major industry ty

pe was measured at t

e two-digit SIC code level and t

en aggregated to four widecategories, as described under Sample and procedures above. T

is variable controls t

e potential influence of industry on learning (Li, 1995) and profitability. T

e survey asked managers to name t

e industry from w

ic

t

e company generated most of its sales. 3.3. Model and analysis T

e LISREL 8.30 program was used totest t

e t

eoretical model. Fig. 1 s

ows t

e basis of t

e model proposed, toget

er wit

t

e

ypot

eses to be tested. We used a recursive nonsaturated model, taking transformational leaders

ip (n 1) as t

e e ogenous latent variable, organizational learning (g 1) as a first-grade endogenous latent variable, and innovation (g 2)1 OL1 y11=.94 2 OL2 y12=.92

and organizationap

rformanc

(g 3) as s

cond-grad

ndog

nousat

nt variab

s

. Through f

xib

int

rp ay b

tw

n th

ory and data, this structura

quation mod

approach bridg

s th

or

tica

and

mpirica

know

dg

to a

ow a b

tt

r und

rstanding of th

r

awor

d. Such ana

ysis a

ows for mod

ing bas

d on bothat

nt and manif

st variab

s, a prop

rty w

suit

d to th

hypoth

siz

d mod

, wh

r

most of th

r

pr

s

nt

d constructs ar

abstractions of unobs

rvab

ph

nom

na. Furth

r, structura

quation mod

ing tak

s into account

rrors in m

asur

m

nt, variab

s with mu tip

indicators, and mu tip

-group comparisons. 4. Ana ysis and r

suts In this s

ction w

pr

s

nt th

main r

s

arch r

suts. First, Tab

1 shows th

m

ans and standard d

viations as w

as th

int

rfactor corr

ati

on matrix for th

study variab

s. Th

r

ar

significant and positiv

corr

atio

ns among transformationa 3 OL3 4 OL4 y14=.86 1 PERFOR1 y38=.94 PERFOR2 y39=.96 31 = .24** PERFOR3 y310=.914 y311=.86 PERFOR5 y312=.93 32 = .73*** 8 9 10 11 12 13 14 15

y13=.86

1 2 3 4 5

LEADER1 x11=.73 LEADER2 x12=.87 LEADER3 x13=.81 LEADER4 x14=.78 LEADER5

1 Organiationa

11= .81*** Learnin

1 Transformational Leadership 21= .37*** x15=.81

21 = .56***

3 P rformanc


12/23

2 Innovation2 y25=.73 INNOVA1 5 y26=.75 INNOVA2 6 y27=.87 INNOVA3 7

3

PERFOR6 y313=.97 PERFOR7 y314=.88

y315=.94 PERFOR8

p < 0.1; * p < 0.05; ** p < 0.01; *** p < 0.001 (two-tai

).Fig. 2. R su

ts of structura

quation mo

.


13/23

J.A. Aragon-Corr a t a. / In

ustriaMark ting Manag m nt 36 (2007) 349 359

355

a

rs

ip, organiationa

arning, innovation an

p rformanc . A s ri s of t sts ( .g., to

ranc , an

varianc inf

ation factor) s

ow

t

a

s nc of mu

tico

in arity (Hair, An

rson, Tat

am, & Back, 1999). W p rform

structura qua

tion mo

ing (Bo

n, 1989) to

stimat

ir

ct an

in

ir

ct

ff

cts, using LISREL an

using t

corration matrix an

asymptotic covarianc matrix as input. T

is typ of ana

ysis

as t

a

vantag of corr cting for unr

ia

i

ity of m asur s an

aso giv s information on t

ir ct an

in

ir ct pat

s

tw n mutip

constructs aft r pot ntia

y confoun

ing varia

s ar contro

for. Fig. 2 s

ows t

stan

ar

i

structuraco ffici nts. T

is

iagram s

ows ony pat

s t

atar significant at t

0.5

v

. T

magnitu

of t

co ffici nts of t

varia

s r f

cts t

ir rativ importanc . Wit

r sp ct to t

quaity of t

m asur m nt mo

for t

samp

, t

constructs

ispay satisfactory r

ia

iity,

as in

icat

y composit ria

iiti s ranging from 0.78 to 0.92 an

s

ar

varianc co ffici nts ranging from 0.62 to 0.85 (Ta

2). Conv rg nt va

i

ity t

xt nt to w

ic

maxima y

iff r nt att mpts to m asur t

sam conc pt agr can

ju

g

y

ooking at

ot

t

significanc

of factor

oa

ings an

s

ar

varianc . T

amount of varianc s

ar

or captur

y a construct s

ou

gr at

r t

an t

amount of m asur m nt rror (s

ar

varianc >0.50). A

t

muti-it

m constructs m t t

is crit rion, ac

oa

ing (k)

ing significant y r at

to its un

r ying factor (t-va u s gr at r t

an 28.29). Lik wis , a s ri s of c

i-squar

iff r nc t sts on t

factor corrations s

ow

t

at

iscriminant vai

ity t

gr to w

ic

a construct

iff rs from ot

rs is ac

i v

among a

constructs (An

rson & G r

ing, 1988). InTa

2 Va i

ity, r ia

i ity an

int rna consist ncy Varia

It m Param t r Va

i

ity, ria

iity an

int rnaconsist ncy k* Transformationa

LEADER1 L a

rs

ip LEADER2 LEADER3 LEADER4 LEADER5 Organiationa

OL1 L arning OL2 OL3 OL4 Inn

ovation INNOVA1 INNOVA2 INNOVA3 P rformanc PERFOR1 PERFOR2 PERFOR3 PERFOR4 PERFOR5 PERFOR6 PERFOR7 PERFOR8 kx 11 kx 12 kx 13 kx 14 kx 15 k 11 k 12 ky 13 ky 14

ky 25 ky 26 ky 27 ky 38 ky 39 ky 310 ky 311 ky 312 ky 313 ky 314 ky 315 0.73 (f.p.) 0.87*** 0.81*** (32.60) 0.78*** (30.07) 0.81*** (30.37) 0.94 (f.p.) 0.92***(67.04) 0.86*** (60.53) 0.86*** (44.93) 0.73 (f.p.) 0.75*** (28.29) 0.87*** (31.18) 0.94 (f.p.) 0.96*** (70.13) 0.91*** (55.33) 0.86*** (59.09) 0.93*** (68.24)0.97*** (70.06) 0.88*** (63.71) 0.94*** (66.86) R2 0.53 0.76 0.65 0.61 0.66 0.880.84 0.74 0.74 0.53 0.56 0.76 0.87 0.91 0.83 0.74 0.87 0.94 0.78 0.88 A.M. a =0.850 C.R. = 0.899 S.V. = 0.642

a = 0.919 C.R. = 0.941 S.V. = 0.800 a = 0.777 C.R. = 0.827 S.V. = 0.616 a = 0.889 C.R. =0.978 S.V. = 0.853

k* = stan

ar

i

structuraco ffici nt; R 2 = r

ia

iity; a = a

p

a Cron

ac

;C.R. = compoun

ria

iity; S.V. = s

ar

varianc ; f.p. = fix

param t r; A.M. = a

justm nt m asur m nt. *** Significanc at p < 0.001 (two-tai

).

particuar, w sta

is

iscriminant vai

ity

tw n ac

pair ofat nt var

ia

s

y constraining t

stimat

corration param t r

tw n t

m to 1.0 an

t

n p rforming a c

isquar

iff r nc t st on t

vau s o

tain

for t

constrain

an

unconstrain

mo

s (s An

rson & G r

ing, 1988). T

r su tingsignificant

iff r nc s in c

i-squar in

icat t

at t

constructs ar not p rf ct

y corr

at

an

t

at

iscriminat vai

ity is ac

i v

. T

ov ra

fit m asur s, t

mutip

squar

corration co ffici nts of t

varia

s, an

t

signs an

significanc v s of t

pat

co ffici nts a in

icat t

at t

mo

fits t

2

ata w

(v 165 = 0.34, p < 0.001; NFI = 0.93; NNFI = 0.93; GFI =95; CFI = 0.94; AGFI = 0.94). T

ypot

si

mo

2 was a significant

y

tt

r fit t

an t

nu mo

(v 190 = 2 14953.36, p < 0.001; Dv 25 = 13939.02, p

7311763 Leadership and Organizational Learnings Role on Innovation and Performance Lessons From...

Documents

Transcript of 7311763 Leadership and Organizational Learnings Role on Innovation and Performance Lessons From...