Content uploaded by Xander M. Bezuijen
Author content
All content in this area was uploaded by Xander M. Bezuijen on Nov 16, 2017
Content may be subject to copyright.
Howleaders stimulate employeelearning:
Aleader–member exchangeapproach
Xander M. Bezuijen, Karen van Dam*, Peter T. van den Berg
and Henk Thierry
Tilburg University,The Netherlands
This study investigated how leader–member exchange (LMX), goal setting, and
feedback arerelated to employee engagement in learning activities. Tw odifferent
mechanisms wereproposed: amediating mechanism holding that LMX elicits specific
leader behaviours (i.e., goal setting and feedback) which would mediate the LMX-
learning relationship,and amoderating mechanism, holding that LMX would strengthen
the effect of these leader behaviours. Asample of 1,112 employees from 7organizations
completed questionnaires that measured LMX, goal specificity,feedback, and self-
reports of employee engagement in learning activities. The 233 direct leaders of these
employees completed questionnaires that measured goal difficulty and leader ratings of
employee engagement in learning activities. Multi-level analysis showed that goal
difficulty and goal specificity mediated the relationship between LMX and employee
engagement in learning activities, and that LMX moderated the relationship of goal
difficulty with employee engagement in learning activities. With these findings, the
present study contributes to the literatures on LMX, goal setting, and employee
development.
Employee engagement in learning activities is becoming increasingly important for
organizational effectiveness and foremployee success in today’srapidly changing
workplace.Engagement in learning activities refers to employees’ discretionary
behavioursinongoing learning activities to master newknowledge, skills,and abilities
(Bezuijen, van den Berg, van Dam, &Thierry,2009). Theseactivities can take different
formssuch as training assignments on and offthe job, challenging and novel tasks,
special projects, and job transitions (McCauley&Hezlett, 2001).Employee learning is
generally considered an essential prerequisite fororganizational adaptability and
competitiveness (Maurer,Pierce, &Shore, 2002). Despite the importance of employee
learning, littleisknown about how leadersencourageemployees to engageinlearning
activities, and researchershave not delineated the specific behavioursand mechanisms
through which such learning occurs(Ve ra &Crossan, 2004).
*Correspondence should be addressed to Dr Karen van Dam, Tilburg University,FSW,W/O Psychology,POBox 90153,
5000 LE Tilburg, The Netherlands (e-mail: k.vandam@uvt.nl).
The
British
Psychological
Society
673
Journal of Occupational and Organizational Psychology(2010), 83, 673–693
q2010 The British Psychological Society
www.bpsjournals.co.uk
DOI:10.1348/096317909X468099
The objective of the present study was to address this issuebyapplying leader–
member exchange(LMX) theory to employee learning. According to LMXtheory,
leadersformunique relationships with each of their subordinates so that high LMX
relationships are characterized by high levels of mutual support, trust, and loyalty
(Graen &Uhl-Bien, 1995; Sparrowe &Liden, 1997). LMX has been associated with
important outcomes such as job performance, job satisfaction, and retention (Gerstner
&Day,1997).Although the relevance of LMX foremployee learning has been suggested
in the literature (e.g., Dragoni, 2005),studies examining this relationship are scarce.
The present researchcontributes to the literature in three ways. First, we provide
insightinto two different yet related mechanisms through which leaders encourage
employee learning: amediating and amoderating mechanism. We argue that the LMX
relationship elicits specific leader behavioursthat mediate the relationship between
LMX and employee learning. Theoretical and empirical evidence suggests that leaders
are likely to use goal setting and feedback to encourage employee learning (e.g., Lam,
Huang,&Snape, 2007).Wealso argue that LMX acts as amoderatorinthat it increases
the effectiveness of these leaderbehaviours. Whereas amoderating effect of LMXhas
been suggested in the literature (Driver,2002), only afew studies have actually
investigated this effect (Scandura &Graen, 1984; Sparrowe, Soetjipto,&Kraimer,2006).
The second contribution of this study concerns how LMXand goal-setting theories
can be integrated with one another.Whereas these theories have been studied
independently in the past, we propose that LMX is an important antecedent of goal
setting,that is, the extent to which leadersengageingoalsetting dependsonthe quality
of the LMXrelationship.Moreover,byinvestigating the effects of learning goals, as
opposed to performance goals, the present study aims to extendprevious research
aboutlearning goals, atopic that has receivedattention only recently (e.g., Seijts,
Latham,Tasa, &Latham,2004).
Finally, the present study contributes to the employee development literature by
using abroadapproachtoemployee learning and by focusing on abroader arrangement
of learning activities beyond just traditional, formal developmental activities (McCauley
&Hezlett, 2001; Noe, Wilk, Mullen, &Wanek, 1997). This broader approach to
employee learning is in accordance with the literature on ‘the learning organization’
(e.g., Senge, 2006) and ‘organizationallearning’ (e.g., Argyris &Schon, 1978), which
emphasizes the need foramultitudeoflearning activities in ordertocreate aculture of
continuous personal development. Despite the recent emphasis on abroader learning
perspective, researchusing this perspective is scarce and questions remain as to how
learning at workmay be enhanced (Sonnentag,Niessen, &Ohly,2004).The present
study provides insight into how leaders stimulateemployee learning and under what
conditions this maybemost effective.
Employee engagement in learning activities
Employee engagement in learning activities has becomeanincreasingly important topic
in recent years. Demands made on employees to upgrade their job skills and knowledge
are increasing continuously because organizations are faced with markets that are
global, competitive, and technologically based (Maurer &Tarulli, 1994). Therefore,
employees who activelyengageinlearning activities are one of the most important
sources of competitive advantagefor organizations (Senge, 2006; Vera &Crossan, 2004).
Like organizations, leaders, and employees each benefitfrom employee engagement
in learning activities. Leadersbenefit whentheyencourageemployees to undertake
674 Xander M. Bezuijen et al.
tasks that lead to learning because doingsohelps attain strategic team goals (Jansen,
Vera, &Crossan, 2009).Moreover,employee learning is associated with improved job
performance (Arthur,Bennett,Edens,&Bell, 2003), and the acquisition of new skills
that serve as antecedents of job performance (Aguinis &Kraiger,2009). Learning is also
crucial foremployees because theyneedtostay current in alabour market that no
longer guarantees long-termemployment and job security (Arthur &Rousseau, 1996).
Moreover,continuous learning is an important partofmany jobs because job
requirements are undergoing constantchange(Sonnentag et al.,2004). In addition to
being anecessity,engagement in learning activities can be challenging and fun,and
researchhas associated employee learning with positive workattitudes such as job
satisfaction, organizational commitment, and retention (Mikkelsen, Saksvik, Eriksen, &
Ursin, 1999).
There are many different ways foremployees to engageinlearning activities.
Whereas organizations focused on formal training programmes in the past, they
recognize now that valuable learning takes place within the daily worksituation (Poell,
van Dam, &van den Berg, 2004).Muchemployee learning is believed to occur through
assignments that go beyond usual job responsibilities such as new and challenging tasks,
job transitions,task-force assignments, temporary attachments to otherworkunits,and
projectwork(Birdi, Allan, &Warr, 1997; Maurer et al.,2002). Instead of viewing
learning as an occasional training need, employee learning is viewed as acontinuous
process, that may also focus on future assignments and career development (McCauley
&Hezlett, 2001;Noe et al.,1997).
Becausemuch learning occurswithin the dailyworksituation, leadersare generally
considered to be an important force behind employee engagement in learning activities
(Sonnentag et al.,2004).There is clear evidence that supportfrom the leaderenhances
participation in learning activities (Birdi et al.,1997; Colquitt, LePine, &Noe, 2000).
Noe and Wilk (1993) found that employee perceptions of their supervisor’s supportfor
development activity influencedemployee engagement in learning activities. Similarly,
Colquittet al.’s (2000) meta-analytic study showed that employee motivation to learn
was related to their leader’ssupportive behaviours.Despite this evidence, surprisingly
fewstudies have examined the underlying process betweenleaders and employees that
results in employee learning. The present study focused on the LMX relationship to
explain how leaderbehaviour and employee learning are related.
Leader–member exchange and employee engagement in learning activities
LMX theorypositsthat leadersand membersengageinarole development process
during which differentiated role definitions developbetween aleader and an individual
employee (Graen &Cashman, 1975).Whereas low-quality relationships involve
rudimentaryexchanges that typify the basic employmentcontract, high-quality
relationships are characterized by mutual trust, respect,and loyalty between leader and
employee. Based on the concepts of social exchange(Blau, 1964) and reciprocity
(Gouldner,1960),LMX theory posits that the exchangerelationship creates afeeling of
obligation in memberstoreciprocate high-quality relationships(Graen &Uhl-Bien,
1995).Assuch, high-LMX employees are sometimes referred to as ‘trusted assistants’
who are committed to the leader and who enhance their leader’seffectiveness (Liden,
Sparrowe, &Wayne, 1997). Researchhas demonstrated that LMXisrelated to important
employee and organizational outcomes such as job performance, organizational
citizenship behaviour,job satisfaction, organizational commitment, retention, and
LMX and employee learning 675
openness to organizational change(e.g., Gerstner &Day,1997; Hofmann, Morgeson, &
Gerras, 2003; van Dam, Oreg, &Schyns, 2008).
It is generallyexpected that high-LMX employees engageinmore learning activities
than do low-LMX employees (e.g., Driver,2002; Maurer et al.,2002; Paparoidamis,
2005). Empiricalevidence hasindicated thatemployees in high-quality LMX
relationships, in contrast to those in low-quality LMX relationships, receive greater
opportunities forpersonal growthand more challenging work assignments (Graen&
Scandura, 1987; Liden et al.,1997), are providedwith higher levels of support (Kraimer,
Wayne, &Jaworski, 2001), and progress more rapidly in their careers(Scandura&
Schriesheim, 1994; Wayne, Liden, Kraimer,&Graf, 1999).The quality of the leader–
member relationship has also been found to affect employee learning goal orientation
(Janssen &Van Yperen,2004).
However,few studies have focused on the specific mechanisms and behaviours
through which LMX impacts employee learning. We arguethat two mechanisms within
the LMX relationship might be at work. The first mechanism, which is related to the role-
making process, refers to amediation effect of leaderbehavioursand holds that leaders
will more activelystimulate and challengehigh-LMX, compared to low-LMX employees,
to engageinlearning activities. The second mechanism, which is related to the normof
reciprocity,refers to amoderating effect and holds that high-LMX employees will
respond more strongly,than will low-LMX employees, to their leader’sgoal setting and
feedback behavioursbecause theywant to reciprocate the leader’sefforts and
expectations. Thesetwo mechanisms are discussed more extensivelybelow.The
researchmodel is presented in Figure 1.
Goal setting and feedbackasmediating leader behaviours
Settinggoals and providing feedback are two of the behavioursthat leaderscan use to
encourage high-LMX employees to learnbecause these behavioursfacilitate the role-
making process of the exchangerelationship (Graen &Scandura, 1987; Graen &
Uhl-Bien, 1995). We arguethat leaders will test their expectations of employees by
setting challenging goals. Each time agoalhas been attained, leaders’ trust, respect,and
expectations will increase and the quality of the LMXrelationship will improve. It is
likely that this role-making process extendstolearning goals. That is, in high-LMX
conditions, leaders not only set morechallenging performance goals, theyset more
challenging learning goals as well. In turn, employees who pursue these learning
goals update their knowledge and skills (Lam et al.,2007) and maintain their roles as
trusted assistants.
Leader-
member
exchange
Goal setting:
-Goal difficulty
-Goal specificity
Feedback
Employee
engagement in
learning activities
Figure1.The research model.
676 Xander M. Bezuijen et al.
Similarly,the exchangerelationship might affect goal specificity and feedback.
Researchhas shownthat leaderscommunicate more frequentlywith high-LMX
employees than with low-LMX employees (Kacmar,Witt, Zivnuska, &Gully,2003),and
that employees moreoften seek feedback from aleader which theybelieve (Fedor,
Rensvold, &Adams, 1992; Va ncouver &Morrison, 1995),trust, and respect (Ashford &
Cummings, 1985; Lam et al.,2007; Va ndeWalle, Ganesan, Challagalla, &Brown, 2000).
Leadersand employees will exchangeinformation with one another about the
employee’slearning activities and goals, and progress towardsthese goals. Through this
information exchange, high-LMX employees will receive greater clarity about their goals
and morefeedback, compared to low-LMX employees. Leader feedback supportsgoal-
setting effortsbecause employees need feedback to adjust the level or direction of their
effortstomatch what the goal requires (Locke &Latham,2002). In sum, leaderswill set
more difficultand more specificlearning goals, and provide more feedback in high-LMX
conditions compared to low-LMX conditions.
In turn, goal setting and feedback will stimulateemployees to engageinlearning
activities (e.g., Maurer,Weiss, &Barbeite, 2003; Noe et al.,1997). Goal setting has
proven to be an effectivemotivational technique forenhancingemployee
performance (Locke &Latham,2002).Strong support has been found forthe
theory’spostulates that difficultgoals lead to higher levels of performance than do
easy goals, and that specific goals produce ahigher level of output than do vague or
‘do your best’ goals (e.g., Latham,Locke, &Fassina,2002; Locke &Latham,1990).
Whereas goal difficulty has been found to affect persistence, goal specificity has been
shown to direct attention and effort towardsgoal-relevant activities (Locke &Latham,
2002).The goal-behaviour relationship is strongest when people are committed to
their goal (Klein, Wesson, Hollenbeck,&Alge, 1999). Meta-analysesreveal that goal
difficulty and goal specificity have shown effect sizes ranging from .42 to .80 for
variousbehaviours(Locke &Latham,1990).
Evidence indicates that goal settingisalso applicabletoemployee learning, and
that setting learning goals is effective (e.g., Payne, Youngcourt, &Beaubien, 2007;
Seijts et al.,2004). Seijts et al. (2004)found that setting specificlearning goals resulted
in more positive outcomes than setting vague learning goals. Moreover,extant
evidence indicates that the pursuit of learning goals is positivelyrelated to learning,
feedback seeking behaviour, and performance (Payne et al.,2007). Consequently,itis
likely that leaderswho set difficultlearning goals and specific learning goals will
encourage employees to engageinlearning activities (e.g., Maureret al.,2003; Noe
et al.,1997).
Like goal setting, feedback can help employees to learnand enhance their work
performance (Goodman, Wood, &Hendrickx, 2004; Kluger &DeNisi, 1996). Feedback
interventionsthat direct attention to appropriate task behaviour have been found to
lead to more rapidlearning,decreased errors during training,and improved
performance (Goodman et al.,2004; Kluger &DeNisi, 1996).Thistypeoffeedback
provides recipients with information about their workbehaviourand performance, and
suggests how theycan makeimprovements (Ashford &Cummings,1985; Chen, Lam, &
Zhong, 2007). Feedback may further affect learning by enhancingthe relative exposure
of recipientstoinstances of good versus bad performance, thus increasing the number
of learning opportunities (Goodman et al.,2004).
In conclusion, we argue that membersinhigher-quality LMX relationshipswill
engage in learning activitiesmoreoften than membersinlower-quality LMX
relationships because their leaders providethem with more difficultand more specific
LMX and employee learning 677
learning goals, and with more feedback. Theseleader behavioursare thought to mediate
the relationship between LMX and employee learning.
Hypothesis 1: Goal difficulty mediates the relationship between LMX and employee
engagement in learning activities.
Hypothesis 2: Goal specificity mediates the relationship between LMX and employee
engagement in learning activities.
Hypothesis 3: Feedback mediates the relationship between LMX and employee engagement
in learning activities.
LMX as amoderating condition
We also posit, in addition to the aforementioned mediation effect,that LMX moderates
the relationshipsofgoal difficulty,goalspecificity,and feedback, respectively, with
employee engagement in learning activities. Ahigh-LMX relationship is an optimal
condition forsocial exchange andreciprocity.Insucharelationship,the
communication is much more intense(Kacmar et al.,2003),and characterized by
mutual interest (Uhl-Bien &Maslyn, 2003) and feedback seeking (VandeWalle, Ganesan,
Challagalla, &Brown, 2000).Wearguethat these LMX-relationship characteristics affect
employees’ responses to leader behaviours(cf. Maurer et al.,2002).
Although veryfew studieshave investigated the moderating effect of LMX on the
effectiveness of leader behaviours(e.g., Scandura &Graen, 1984; Sparrowe et al.,2006),
several argumentssupportthis assumption. First, high-LMX employees will tryharderto
fulfil the leader’srole expectations as expressed in their leader’sgoal setting and
feedback, because theycan benefit more from the relationship than low-LMX employees
(Lam et al.,2007; Maslyn &Uhl-Bien, 2001). Second,previous studies have found that
feedback is moreeasily accepted within ahigh-trust relationship (Chen et al.,2007).
Third, ahigh-quality LMX relationship enhancesgoalcommitment (Klein et al.,1999),
which in turnstrengthens the relationship between goal setting and performance
(Locke &Latham,1990).Finally, being atrusted assistant might increase high-LMX
employees’ self-expectations to act successfully on the feedback provided by the leader,
which will have apositive effect on their learning behaviors(Bandura, 1997).
The reverse canbeargued forlow-LMX conditions. In negative reciprocal
relationships, self-interestinstead of mutual interest (Uhl-Bien &Maslyn, 2003), and
resistance to leader influence (Tepper,Duffy,&Shaw, 2001) are more common. As a
consequence,leaders’ behaviourswill not be as effectiveasinhigh-LMX relationships.
That is, uponfeedback and goal setting, low-LMX members will not feel astrong
inclination to invest in the relationship, or pay back the leader by engaginginlearning
activities. Becausetrust is low, employees in low LMX-relationships will not readily
accept the leader’sfeedback and goals. Finally, being on the outside circle might have a
negative effect on employees’ self-expectations and,therefore, will not stimulatelow-
LMX memberstoengageinchallenging learning opportunities.
Hypothesis 4: LMX moderates therelationship between goal difficulty and employee
engagement in learning activities, such that this relationship is more positiveunder high-quality
LMX conditions than under low-quality LMX conditions.
Hypothesis 5: LMX moderates the relationship between goal specificity and employee
engagement in learning activities, such that this relationship is more positiveunder high-quality
LMX conditions than under low-quality LMX conditions.
678 Xander M. Bezuijen et al.
Hypothesis 6: LMX moderates the relationship between feedback and employee engagement in
learning activities, such that this relationship is more positiveunder high-quality LMX conditions
than under low-quality LMX conditions.
Method
Sample and procedure
Atotal of 1,112 employees and 233 leadersfrom 7organizationsinThe Netherlands
participated in the present study.Differenttypes of workwere represented: health care
( N¼302), police ( N¼188), penitentiary(N¼156), social services(N¼102),
security services(N¼94), high-tech ( N¼208), and vocational training ( N¼62).
Employees performed various job duties within these sectorssuch as those of nurse,
physician, police officer,security guard,counsellor,office worker,social worker,day
care worker,R&D specialist, factoryworker,and teacher.
One questionnaire wasdeveloped foremployees and aseparate questionnaire was
developed forleaders(see Appendix). The employee questionnaire included aself-
assessment of their engagement in learning activities, LMX, goal specificity,and
feedback from their leader. The leader questionnaire constituted leader ratings of
employee engagement in learning activities, and goal difficulty.Atotal of 3,295
questionnaireswere sent to employees, of which 1,546(47%) were returned. The
leadersofthese employees wereapproached after the employee questionnaires were
returnedand wereasked to complete the leader questionnaire. Of the 1,546 employee
questionnaires, 1,112 (34%) could be matched to their leaders’ questionnaires. Chi-
square tests fornon-response bias indicated that there werenodifferences between
employee respondents and non-respondents forage,gender,and years of education.
Leadersrated about 5employees ( M¼4: 8) on average, with the rangebeing from alow
of 1toahigh of 15 employees.
The employees ranged in agefrom 17 to 65 years(M¼40:5, SD ¼9 :5) and had an
averageof14.9 years ( SD ¼2:3) of education. Of the employees, 47% were male.
Employees worked 2.4 years(SD ¼2:2) forthe same leaderonaverage.
Measures
Employee engagement in learningactivities
An eight-item scale that addressed arangeofrelevant learning activities was developed
based on workfrom previous studies (Birdiet al.,1997; Maurer &Tarulli, 1994; Maurer
et al.,2003; Noe &Wilk, 1993). Respondents were asked to indicate on afive-point scale
from 1(never)to5(very often)how frequently the employees performed the
behavioursdescribed in the items.Employees provided self-reportoftheir behaviours
whereas leadersprovided ratings of their employees’ behaviours. Cronbach’salpha was
.85 and .91 forthe employee and leader scales, respectively.
Becausethese scales weredeveloped forthis study,theywere tested first in apilot
study using four organizations: an electricity company ( N¼350), an information
technology service organization ( N¼91), atax office ( N¼207), and apurchase
organization (N¼135). Exploratoryfactor analysis (principal axis factoring and oblimin
rotation) supported the measurement model and indicated that the scale foremployee
engagement in learning activities measured adistinct construct. Cronbach’salpha was
.78 and .88 foremployee and leader perceptions of learning activities, respectively.
LMX and employee learning 679
Moreover,inthis sample, the scales were significantly related to self-efficacy,willingness
to engageinlearning activities, and personal initiative.
LMX relationship
Graen and Uhl-Bien’s(1995) seven-item LMXscale was usedtomeasure employee
perceptions of the LMX relationship. Responses weremade on afive-point Likertscale
that ranged from 1(not good )to5( very good ). Cronbach’salpha was .92.
Goal difficulty
1
Asix-item scale (Bezuijen et al.,2009) was used to measure learning goal difficulty.The
goals referred to activities that involved learning such as extending knowledge and
skills.Leaders, and not employees, were asked to rate the difficulty of the goals that were
set foreach employee, because studies have shownthat employee perceptions of goal
difficulty areconfounded with self-efficacy beliefs (Latham &Locke, 1991; Locke &
Latham,1990). Leadersindicated goal difficulty on afive-point Likertscale that ranged
from 1(very easy goals)to5(very difficult goals). Cronbach’salpha was .87.
Goal specificity
We used Bezuijen et al.’s (2009) six-item scale to measure learning goal specificity.The
items addressed the same learning goals as the items of the goal difficulty scale.
Employees responded to the items on afive-point Likertscale that ranged from 1(very
vague goals)to5( very specific goals). Cronbach’salpha was.93.
Feedback
Leader feedback concerning employee learning was measured with Bezuijen et al.
(2009)four-item scale. This measure was based on Kluger and DeNisi’s(1996) concept
of task-detail feedback and focused on learning processes concerning the task.
Employees rated their leader’sfeedback on afive-point Likertscale that ranged from 1
( strongly disagree)to5( strongly agree). Cronbach’salpha was .87.
Control variables
Gender, age, and years of education wereincludedascontrol variables because they
have been associated with learning activities (Maurer et al.,2003).Wealso controlled for
the organization in which the employees worked by introducing dummy variables for
each of the organizations. Organizationalmembership indeed affected the extent to
which employees engaged in learning activities. Employees whoworked in the social
service organizations, or forthe vocational training school, showed the least amount of
learning, and those who worked forthe police showed the most. Because organizational
1
Because scales forlearning goal difficulty,learning goal specificity, and learning feedbackdid not exist, we developed new
scales for these variables which can be found in Bezuijen et al. (2009). Whereas we needed ascale that would measure
learning goals as set by the leader,goal setting, goal difficulty,and goal specificity have been manipulatedinmost previous
studies.Similarly,feedbackstudies usually focus on feedback seeking behavior,oronfeedbackprovided in experimental
settings.Asfar as we know, the extent to which leaders provide feedbackonlearning issues has not been measured before.
680 Xander M. Bezuijen et al.
membership did not affect the regression coefficientsofthe variables in the model, the
dummy variables were not includedinthe tables.
Factor analyses
Exploratoryand confirmatoryfactor analyses were used to test whether the scales
measured distinct constructs. The exploratory factor analyses consisted of principal axis
factoring with oblimin rotation. The analyses were conducted twice. Model Aused
employee self-reportratings of employee engagement in learning activities, whereas
Model Bused leader ratings of employee engagement in learning activities. The
expected five-factor model appeared which explained 64 and 67% of the variance for
Models Aand B, respectively.
The confirmatoryanalysis was also conducted twice. Chi-square difference tests
were performed to investigate the assumption that five variables would account forthe
relationships between the responses (Anderson &Gerbing, 1988).The results of the
confirmatoryfactor analyses are presentedinTable 1.
The fit indices of the confirmatoryfactor analysessuggested that the alternative
one-factor and three-factor models didnot fit the data well, whereas the five-factor model
fit the data reasonably well (Model A: x 2ð421;N¼1;112Þ¼2;033:4, SRMR ¼:05,
CFI ¼:93, RMSEA ¼:06; ModelB:x2ð421;N¼1 ;112 Þ¼2 ;212:1, SRMR ¼:04,
CFI ¼:93, RMSEA ¼:06).When the five-factor model was compared to the three-
factormodel, the resulting difference in chi-square (Dx2ð7 ;N¼1 ;112Þ¼3;366:5,
D df ¼7, p,.001) was highly significant, suggestingthat the hypothesized five-factor
model fit the data better than the three-factor model. All items loaded significantly on
the intended factor,with factor loadings ranging from b¼0:48 to 0.92 forModel A, and
b¼0:58 to 0.92 forModel B. In sum, the exploratoryand confirmatoryfactor analyses
each supported the hypothesized measurementmodel.
Analyses
The data were hierarchically nested, with the individual leader–employee relationship at
the first-level, and with the group (of employees whoworked under the sameleader) at
the second level. Nested data may create problems in standardregression procedures
because thestandardassumption of independent andidenticallydistributed
observationsisgenerally not valid. To control forinterdependence among variables
that concernemployees whoworkunder the sameleader,atwo-level (i.e., multi-level)
hierarchical regression analysis with full maximum likelihoodestimation was conducted
using the procedure described by Hox (2002). We used the HLM program version 5.04.
The intercept-only model was assessed to determine the intra-classcorrelations (ICC) in
the first step of this procedure. The ICC indicates how much of the total variance in the
dependent variable, that is explained by the two levels, is accounted forbythe second
level (which is the group in this instance). The findings showed that ICCð1Þ¼:26,
x 2ð230;N¼1 ;112Þ¼580:49, p,.001 foremployee self-reports of employee learning,
andICCð1Þ¼:24, x 2ð230;N¼1 ;112 Þ¼569:69, p,:000 forleaderratings of
employee learning. These findings indicated that working under the same leader
affected employee engagement in learning activities to aconsiderable extent; therefore
multi-level analysis was the appropriate technique foranalysing our data.
We examined whether Baron and Kenny’s(1986) three conditions formediation
were met when testing the mediating effect of leader behaviour in our model (see also
LMX and employee learning 681
Ta ble 1. Results of confirmatoryfactor analyses
Model Factor SRMR TLI CFI RMSEA df x
2
Dx
2
p
One-factor All variables .14 (.19) .43 (.37) .47 (.41) .15 (.17) 431 (431) 9,670.0 (11,107.7)
Three-factor Employeeengagement
in learning activities
.11 (.15) .70 (.71) .72 (.73) .11 (.12) 428 (428) 6,303.5 (6,871.9) 3,366.5 (4,235.8) ***
Feedback, goal difficulty,
goal specificity
LMX
Five-factor Employeeengagement in
learning activities
.05 (.04) .92 (.92) .92 (.93) .06 (.06) 421 (421) 2,033.4 (2,212.1) 2,902.2 (3,284.7) ***
LMX
Feedback
Goal difficulty
Goal specificity
Note.The outcomes for Model Bare reported in parentheses; ***p,.001.
682 Xander M. Bezuijen et al.
Mathieu &Taylor,2006). Baron and Kennynotethat: (a)asignificant relationship should
exist between the independent variable and the dependent variable; (b) asignificant
relationship should exist between the independent variable and the mediator; and (c) a
significant relationship should exist between the mediator and the dependent variable
while holding the independent variable constant. In the case of afully mediating effect,
the independent variable will have no significant relationship with the dependent
variable when the mediator is added to the analysis. Sobel’s(1982) test formediation
was used to establish the significance of the indirecteffect of the independent variable
on the dependent variable.
Results
The means, standard deviations, and intercorrelations of all variables are presented in
Table 2. The correlation of .40 between leader ratings and employee self-reports of
employee engagement in learning activities indicated that leadersand employees agreed
moderately about employee learning behaviour. Moderatecorrelations between self-
ratings and leader ratings are common as can be seen in the averageintercorrelation of
.35 that has been reported in the meta-analysis by Harrisand Schaubroeck (1988).
The results of the multi-level regression models are presented in Table 3. The first
step of the analyses revealed that gender,age,and education were significantly
correlated with employee engagement in learning activities, with male, younger,and
higher-educated employees reporting morelearning activities.
As Table 3shows, the first condition formediation was met. There were significant,
positive relationships between LMXand bothemployee and leader ratings of
engagement in learning activities when gender,age,and education were controlled.
To examine the second conditionfor mediation, we investigatedwhether
relationships existed between LMXand the mediating variables.Multiple regression
analyses, in which the control variables were included, showed significant relationships
with LMX forall threemediators, namely, goal difficulty (b¼0: 14, p,.001), goal
specificity ( b¼0 :51, p,.001), and feedback ( b¼0 :58, p,.001). Therefore, the
secondcondition formediation was met as well.
To examinethe third condition formediation, the relationships of goal difficulty,goal
specificity,and feedback, respectively, with engagementinlearning activities were
examinedwhile holding LMXconstant. As shown in Ta ble 3(step 2), goal difficulty was
positivelyrelated to engagement in learning activities. Sobel’s(1982) test formediation
showed that this mediation effect was significant. The zvalues of the indirect effect of
LMX through goal difficulty on engagement in learning activities was5.52 ( p,:001) for
employee self-report ratingsand 4.63 (p,:001) forleader ratings.Therefore,
Hypothesis 1was supported.
Similarly,goal specificity was significantly and positivelyrelated to employee
engagement in learning activities. Sobel’s(1982) test formediation showed that this
mediation effect was significant. The zvalues of the indirect effect of LMXthrough goal
specificity on learning was 8.78 ( p,: 001) foremployee self-ratings and 2.26 (p,.05)
forleader ratings.Consequently,Hypothesis 2was supported as well. Although Sobel’s
test showed that the relationship betweenLMX and learning significantly declined in
strength, this relationship remained significant, indicating that goal difficulty and goal
specificity only partially mediated this relationship.
Contrarytoexpectations, feedback was negativelyrelated to employee engagement
in learning activities. Therefore, Hypothesis 3was rejected. It should be noted, however,
LMX and employee learning 683
Ta ble 2. Means, standarddeviations, intercorrelations, and reliabilities (in parentheses) of measures
MSD123456789
1. Gender
a
0.47 0.50 –
2. Age 40.46 9.47 2.04 –
3. Ye ars of education 14.94 2.28 2.04 2.04 –
4. Leader–member exchange 3.35 0.82 2.06* 2.04 2.01 (.92)
5. Goal difficulty 2.57 0.81 .00 2.12*** .12*** .15*** (.87)
6. Goal specificity 2.78 1.06 2.11*** 2.09** .01 .51*** .22*** (.93)
7. Feedback 3.06 0.89 .03 2.13*** 2.15*** .59*** .11*** .57*** (.87)
8. Employeeengagement in learning activities
(employeeself-reports)
2.84 0.65 2.14*** 2.25*** .14*** .23*** .23*** .38*** .17*** (.85)
9. Employeeengagement in learning activities
(leader ratings)
2.76 0.78 2.08** 2.24*** .11*** .22*** .61*** .23*** .11*** .40*** (.91)
Note.N¼1 ;112 supervisor–subordinate dyads. Forr$: 06,p,: 05; for r$: 08,p,: 01; for r$:10,p,:001; *p,:05; **p,: 01; ***p,:001:
a
Female ¼1; male ¼0.
684 Xander M. Bezuijen et al.
that whereas the regression coefficients were negative, the zero-order correlations
between feedback and bothmeasures of learning werepositive. These different
outcomes might indicate the presence of aconfoundingorsuppressing variable, or a
spurious correlation.
With respect to the moderating effects of LMX (see Ta ble 3, step 3), the results for
both self-reportand leader ratings showed significant interactions forgoaldifficulty and
feedback, respectively. Accordingly,Hypotheses 4and 6weresupported. The
interaction of employee learning and goal specificity wasnot significant foreither of
the ratings, therefore Hypothesis 5was rejected.
We looked at the combined effects of the main variables and the interaction terms to
learnmoreabout the interaction effects. The relationship between goal difficulty and
Ta ble 3. Multi-level model with dependent variable employee engagement in learning activities
Employee engagement in learning
activities (employee self-reports)
Employee engagement in learning
activities (leader ratings)
Variable BSEbBSEb
Step 1
Gender
a
20.15 .04 20.12 *** 20.06 .05 20.04
Age 20.02 .00 20.22 *** 20.02 .00 20.23 ***
Education 0.04 .01 0.12 *** 0.04 .01 0.12 ***
LMX 0.17 .03 0.22 *** 0.19 .03 0.20 ***
Deviance 1,960.0 2,360.8
Pseudo R
2
.10 .12
Step 2
Gender
a
20.13 .04 20.10 *** 20.05 .04 20.03
Age 20.01 .00 20.18 *** 20.01 .00 20.15 ***
Education 0.03 .01 0.10 *** 0.02 .01 0.06 *
LMX 0.06 .03 0.07 *0.13 .03 0.14 ***
Goal difficulty 0.12 .02 0.14 *** 0.55 .03 0.57 ***
Goal specificity 0.20 .02 0.33 *** 0.05 .02 0.07 *
Feedback 20.05 .03 20.07 *20.08 .03 20.09 **
Deviance 1,820.4 1,876.1
Dx
2
(df ¼3) 139.6*** (df ¼3) 463.9***
Pseudo R
2
.21 .42
Step 3
Gender
a
20.14 .04 20.10 *** 20.06 .04 20.04
Age 20.01 .00 20.19 *** 20.01 .00 20.16 ***
Education 0.03 .01 0.10 *** 0.02 .01 0.06 **
LMX 0.08 .03 0.11 ** 0.14 .03 0.15 ***
Goal difficulty 0.11 .02 0.14 *** 0.55 .03 0.57 ***
Goal specificity 0.20 .02 0.33 *** 0.05 .02 0.07 **
Feedback 20.06 .03 20.08 *20.08 .03 20.09 ***
LMX £goal difficulty 0.06 .02 0.06 ** 0.09 .03 0.08 **
LMX £goal specificity 0.01 .03 0.01 20.03 .03 20.04
LMX £feedback 0.08 .03 0.11 ** 0.06 .03 0.07 *
Deviance 1,792.4 1,859.4
Dx
2
(df ¼3) 28.0*** (df ¼3) 16.7**
Pseudo R
2
.23 .44
Note.
a
Female ¼1; male ¼0; * p,: 05; **p,:01; ***p,: 001:
LMX and employee learning 685
employee engagement in learning activities was stronger in high-LMX than in low-LMX
conditions (see Figure 2). This relationship varied from B¼0:06 at one standard
deviationbelow the mean value of LMX, to B¼0 :16 at one standard deviation above the
mean value of LMX foremployee self-reports, and from B¼0:49 to 0.60 forleader
ratings (see Figure 2). The relationship between feedback and employee engagement in
learning activities varied from B¼20 :13 at one standard deviation below the mean
value of LMX, to B¼0:01 at one standarddeviation above the mean value of LMXfor
employee self-reports, and from B¼20 :13 to 20.02 forleader ratings.
Because the ICC values weresizable, post hoc analyses wereconducted to test
whether employee engagement in learning activities was related to group-level
characteristics. The findings showed that, at the group level,employee engagement in
learning activities was unrelated to LMX, specific goals, difficult goals, and feedback. Also,
the LMX group-level variable did not moderate the relationships of specific goals, difficult
goals, and feedback with employee engagement in learning activities.The post hoc
analyses wererun using employee self-reports as well as leader ratings of employee
engagement in learning activities;however,source of data had no effect on the outcomes.
1
2
3
4
5
12345
12345
Goal difficulty
12345
12345
Goal difficulty
Engagement in learning activities
(employees)
LowLMX
High LMX
(a) (b)
(c) (d)
1
2
3
4
5
Engagement in learning activities
(leaders)
LowLMX
High LMX
1
2
3
4
5
Feedback
Engagement in learning activities
(employees)
LowLMX
High LMX
1
2
3
4
5
Feedback
Engagement in learning activities
(leaders)
LowLMX
High LMX
Figure2.Interactions between LMX and goal difficulty and feedback on employee engagement in
learning activities as perceived by employees and leaders. Note.The regression lines aredrawn from one
standarddeviation below the mean of the moderator (LMX) to one standarddeviation above.
686 Xander M. Bezuijen et al.
Discussion
The findings provide evidencefor two differentmechanisms involving LMX, namely,a
mediating effect as well as amoderating effect. The first mechanism refers to how the
leader behavioursofgoal difficulty and goal specificity mediate the relationship between
LMX and employee learning. Leadersset more difficult, and more specificlearning goals
forhigh-LMX membersthan theydid forlow-LMX members. By setting such goals,
leadersencourage high-LMXemployees to develop themselves and to live up to their
leader’s expectations. However,goalsetting didnot account completely forthe
relationship betweenLMX and employee learning, and it is possible that other leader
behaviourssuch as modelling (Bandura, 1997) or providing learning opportunities(Noe
&Wilk, 1993), servedasmediatorsaswell. Alternatively, the direct relationship between
LMX and learning might indicate that high-LMX employees engageinlearning activities
more frequently to show their loyalty and earntheir leaders’trust.
2
Future researchis
needed to addressadditional potential mediatorsand consideralternative explanations.
The second mechanism concerns the moderating effect of the LMX relationship on
employee responses to leader behaviours. As predicted, LMXquality strengthened the
effect of setting difficultgoals. Compared to employees in low-quality relationships,
employees in high-quality LMX relationships appearedmore eagertosucceed with
difficultlearning goals by engaginginlearning activities. Thisfinding supportsMaurer
et al.’s (2002) proposition that LMXmay be an important condition forthe effectiveness
of leadership behaviours. LMX did not act as amoderatorfor goal specificity,however.
Settingspecific goals appeared an effectivetool forencouraging learning activities
regardless of the exchangerelationship. These differential findings forgoal difficulty and
goal specificity might be explained by the fact that performing difficult tasks, more so
than performing specific tasks, requires greater goal commitment (Locke &Latham,
1990),which is enhanced by ahigh-quality LMXrelationship (Klein et al.,1999).
Feedback wassimilarly expected to mediate the relationship between LMXand
learning behaviour,and the findings indeed showed that leaders providedmore
feedback to membersinhigh-quality LMX relationships. In contrast to our prediction,
however,the relationship betweenfeedback and engagement in learning activities was
negative,suggestingthat employees engageless frequently in learning activities upon
receiving their leader’sfeedback. Negative effects of feedback are not uncommon.
Ameta-analytic study by Kluger and DeNisi (1996) revealed that feedback had anegative
outcome in 38% of the studies. As Kluger and DeNisi note, feedback might hamper
subsequent behaviour when it has anegative impact on the receiver’sself-efficacy.
In the present study,the zero-order correlation betweenfeedback and learning was
significant and positive; therefore, the negative regression coefficient might be a
statistical artifact caused by aspurious relationship or it might indicate the presence of a
suppressor (Bollen, 1989).
Although the lack of amediating effect forfeedback wasunexpected, the findings do
indicate that LMX moderated the relationship between feedback and learning activities
(see Figure 2). The moderating effect is in line with the hypothesized beneficial effects
of LMX, and suggests that LMXmay neutralize apossible negative effect of feedback on
learning activities. Future researchisneeded to delineate the relationship between
feedback and learning activities more accurately.
2
We would liketothank an anonymous reviewer for pointing this possibility out to us.
LMX and employee learning 687
This study has some limitations that should be acknowledged. Although great effort
was madetocollect data from two separate sources, it is still possible to have common-
method bias within asingle source. However,itappearsthat common-methodbias is
not an important issue in our study given that the confirmatoryfactor analyses and
regression analysesyielded similar results forleader and employee ratings of employee
learning.
Asecondlimitation concerns the cross-sectionaldesign, which limits the degree to
which we can makecausal inferences. Longitudinal researchisneeded to address the
issue of causality (Podsakoff, MacKenzie, Lee, &Podsakoff, 2003) because it provides
additional opportunities to examine patterns of changeand reciprocal relationships
(Williams &Podsakoff, 1989).
This study has implications forpractice. Leadersare expected to create learning
organizations by stimulating employee learning (London &Smither,1999; Senge, 2006).
Our findings reveal that leadersare inclined to treat subordinates differently,and that
theyare more active and more effectiveinencouraging learning activities when they
have trusting, respectful, and reciprocal exchangerelationships with their members.
Leadersshould becomeaware of this tendency, and trytostimulateall employees to
engageinlearning activities, regardless of the exchangerelationship. Furthermore,
leaderscan trytodevelop social bonds with more employees, and learnabout the
development needs and expectations of low-LMX employees as well. LMX training
programmes are effective in helping leadersbuild better relationships with employees
(Scandura &Graen, 1984). Moreover, simply putting more effort into the relationship by
having regular contact with employees and showing interest,has also showntobe
effectivefor improving the quality of LMX relationships(Maslyn &Uhl-Bien, 2001).
This study has implications fortheoryaswell. First, this study contributes to the LMX
literature by highlighting two mechanisms that underlie the relationship between LMX
and employee learning, namely amediating and amoderating mechanism. Although
these mechanisms appear to workindifferentways, theyare inextricably partofthe
reciprocal and dynamic nature of the exchangerelationship (Graen &Scandura, 1987).
Leadersinvest more in employees whom theyvalueand trust by setting difficult and
specific learning goals and providing feedback; in turn, high-LMX employees are more
receptive to their leader’sgoal setting and feedback, and show morelearning activities.
Our findings are in accordance with the existence of areciprocal interaction that may
lead to an upward or downward spiral. Whether the relationship creates atrusted
assistant, or escalates into anumberofunhealthy conditions such as negative reciprocity
(Uhl-Bien &Maslyn, 2003), supervisor-targeted aggression (Hershcovis et al.,2007),or
employee retaliation (Townsend, Phillips, &Elkins,2000), may depend upon boththe
leader and the employee. Future researchcan address several important issues: whether
the two mechanisms discovered in our researchextend to other areas such as job
performance, contextual behaviour,and workattitudes; whether additional leader
behaviourscan be found that mayexplain the relationshipsbetween LMX and employee
engagement in learning activities; and finally,whether the differentdimensions of LMX,
as proposed by Liden and Maslyn (1998), relate to different leader behavioursand
employee responses.
Second, the present study contributes to the goal-setting literature by investigating
the role of goal setting within the LMX relationship. Our observation that leadersset
more difficultand more specific goals foremployees in high-LMX conditions confirms
that goal setting is an important factor within the role-making process. Furthermore, our
study supports recent claims that goal setting is avaluable tool forenhancing learning
688 Xander M. Bezuijen et al.
(e.g., Goodman et al.,2004; Seijts et al.,2004; Wayne, Shore, &Liden,1997).Our
findings indicate that setting learning goals that are difficultand specific enhances the
extent to which employees engageinlearning activities. Futureresearchmight explore
whethergoal setting relates to work behavioursother than performance and learning
such as initiative, altruism,organizationcitizenship behaviour, andinnovative
behaviour.
Finally, this study contributes to the employee development literature by using a
broadapproachtoemployee learning and by showing that leadersplay an important
roleindevelopingalearning organization.Althoughsomepublicationshave
emphasized the need foremployees to take responsibility fortheir own development
(e.g., Arthur &Rousseau, 1996),the present study shows that engagement in learning
activities is more than an individual act. Employee learning is in fact fostered by the
relationship that employees have with their leaderand is ashared responsibility that
involves the employee, the leader,and the organization as awhole. Organizations
that strive forcontinuous learning should develop an organization-wide concernthat
acquiring and applying newskills and knowledge is important forall employees
(London&Smither, 1999). Whereasemployees areultimatelyresponsible for
recognizing and realizing their own development needs, the organization in general,
and the leader more specifically,need to provide the resources and support that enable
employee learning.
References
Aguinis, H., &Kraiger,K.(2009). Benefits of training and development forindividuals and teams,
organizations and society.Annual Review of Psychology ,60,451–474.
Anderson, J. C., &Gerbing, D. W. (1988). Structural equation modeling in practice: Areview and
recommended two-step approach. Psychological Bulletin,103,411–423.
Argyris, C., &Schon, D. A. (1978). Organizational learning: Atheoryofaction perspective.
Reading, MA: Addison-Wesley.
Arthur,M.B., &Rousseau, D. M. (1996). Theboundaryless career: Anew employment principle
foranew organizational era.New Yo rk: Oxford University Press.
Arthur,W., Jr., Bennett, W. ,Jr.,Edens, P. S., &Bell, S. T. (2003). Effectiveness of training in
organizations: Ameta-analysis of design and evaluation features. Journal of Applied
Psychology ,88,234–245.
Ashford, S. J.,&Cummings, L. L. (1985). Proactive feedback seeking: The instrumental use of the
information environment. Journal of Occupational Psychology ,58,67–79.
Bandura, A. (1997). Self-efficacy: Theexercise of control.New York: Freeman.
Baron, R. M., &Kenny,D.A.(1986). The moderator–mediator variable distinction in social
psychological research: Conceptual, strategic, and statistical considerations. Journal of
Personality and Social Psychology ,51,1173–1182.
Bezuijen, X. M., van den Berg, P. T. ,van Dam, K., &Thierry,H.(2009). Pygmalion and
employee learning: The role of leader behaviors. Journal of Management,35,1248–1267.
doi:10.1177/0149206308329966
Birdi, K., Allan, C., &Warr, P. (1997). Correlates and perceived outcomes of four types of
employee development activity.Journal of Applied Psychology ,82,845–857.
Blau, P. M. (1964). Exchange and power in social life.New Yo rk: Wiley.
Bollen, K. A. (1989). Structural equations with latent variables.New Yo rk:Wiley.
Chen, Z., Lam, W. ,&Zhong, J. A. (2007). Leader–member exchange and member performance:
Anew look at individual-level negativefeedback-seeking behavior andteam-level
empowerment climate. Journal of Applied Psychology ,92,202–212.
LMX and employee learning 689
Colquitt, J.,LePine, J.,&Noe, R. (2000). Toward an integrativetheoryoftraining motivation:
Ameta-analytic path analysis of 20 yearsofresearch. Journal of Applied Psychology ,85,
678–707.
Dragoni, L. (2005). Understanding the emergence of state goal orientation in organizational work
groups: The role of leadership and multilevel climate perceptions. Journal of Applied
Psychology ,90,1084–1095.
Driver,M.(2002). Learning and leadership in organizations: Toward complementarycommunities
of practice. Management Learning,33,99–126.
Fedor,D.B., Rensvold, R. B., &Adams, S. M. (1992). An investigation of factors expected to affect
feedback seeking: Alongitudinal field study.Pe rsonnel Psychology ,45,779–805.
Gerstner,C.R., &Day,D.V.(1997). Meta-analytic review of leader–member exchangetheory:
Correlates and construct issues. Journal of Applied Psychology ,82,827–844.
Goodman, J. S., Wood, R. E., &Hendrickx, M. (2004). Feedback specificity,exploration, and
learning. Journal of Applied Psychology ,89,248–262.
Gouldner,A.W.(1960). The norm of reciprocity: Apreliminarystatement. American Sociological
Review,25,161–177.
Graen, G.,&Cashman, J. F. (1975). Arole-making model of leadership in formal organizations:
Adevelopment approach. In J. G. Hunt &L.L.Larson (Eds.), Leadership frontiers
(pp. 143–165). Kent, OH: Kent State University Press.
Graen, G. B., &Scandura, T. A. (1987). Toward apsychology of dyadic organizing. Research in
Organizational Behavior,9,175–208.
Graen, G. B., &Uhl-Bien, M. (1995). Relationship-based approach to leadership: Development of
leader–member exchange(LMX) theoryofleadership over 25 years: Applying amulti-level
multi-domain perspective. Leadership Quarterly,6,219–247.
Harris, M. M., &Schaubroeck, J. (1988). Ameta-analysis of self-supervisor,self-peer,and peer-
supervisor ratings. Personnel Psychology ,41,43–62.
Hershcovis, M. S., Turner,N., Barling, J.,Arnold, K. A., Dupre
´,K.E., Inness, M., et al. (2007).
Predicting workplace aggression:Ameta-analysis. Journal of Applied Psychology ,92,
228–238.
Hofmann, D. A., Morgeson, F. P. ,&Gerras, S. J. (2003). Climate as amoderator of the relationship
between leader–member exchangeand content specific citizenship: Safety climate as an
exemplar.Journal of Applied Psychology ,88,170–178.
Hox, J. (2002). Multilevel analysis; techniques and applications.Mahwah, NJ: Erlbaum.
Jansen, J. J. P. ,Vera, D.,&Crossan, M. (2009). Strategic leadership forexploration and exploitation:
The moderating role of environmental dynamism. Leadership Quarterly,20,5–18.
Janssen, O.,&VanYperen, N. W. (2004). Employees’ goal orientations, the quality of leader–
member exchange, and the outcomes of job performance and job satisfaction. Academy of
Management Journal,47,368–384.
Kacmar,K.M., Witt, L. A., Zivnuska, S., &Gully,S.M.(2003). The interactiveeffectofleader–
member exchangeand communication frequency on performance ratings. Journal of Applied
Psychology ,88,764–772.
Klein, H. J.,Wesson, M. J.,Hollenbeck, J. R., &Alge, B. J. (1999). Goal commitment and the
goal-setting process: Conceptual clarification and empirical synthesis. Journal of Applied
Psychology ,84,885–896.
Kluger,A.N., &DeNisi, A. (1996). Effects of feedback intervention on performance: Ahistorical
review,ameta-analysis,and apreliminaryfeedback intervention theory. Psychological
Bulletin,119,254–284.
Kraimer,M.L., Wayne, S. J.,&Jaworski, R. A. (2001). Sources of supportand expatriate
performance: The mediating role of expatriate adjustment. Personnel Psychology ,54,71–99.
Lam, W. ,Huang, X., &Snape, E. (2007). Feedback-seeking behavior and leader–member exchange:
Do supervisor-attributed motives matter? Academy of Management Journal,50,348–363.
Latham, G. P. ,&Locke, E. A. (1991). Self-regulation through goal setting. Organizational
Behavior and Human Decision Processes,50,212–247.
690 Xander M. Bezuijen et al.
Latham, G. P. ,Locke, E. A., &Fassina, N. E. (2002). The high-performance cycle: Standing the test
of time. In S. Sonnentag (Ed.), Psychological management in individual performance
(pp. 201–228). Chichester: Wiley.
Liden, R. C., &Maslyn, J. M. (1998). Multidimensionality of leader–member exchange: An
empirical assessment through scale development. Journal of Management,24,43–72.
Liden, R. C., Sparrowe, R. T. ,&Wa yne,S.J.(1997). Leader–member exchangetheory: The past
and potential forthe future. Research in Pe rsonnel and Human Resources Management,15,
47–119.
Locke, E. A., &Latham, G. P. (1990). Atheoryofgoal setting and task performance.Upper Saddle
River,NJ: Prentice-Hall.
Locke, E. A., &Latham, G. P. (2002). Building apracticallyuseful theoryofgoal setting and task
motivation: A35-year Odyssey.American Psychologist,57,705–717.
London, M., &Smither,J.W.(1999). Empowered self-development and continuous learning.
Human Resource Management,38,3–15.
Maslyn,J.M., &Uhl-Bien, M. (2001). Leader–member exchangeand its dimensions: Effects of self-
effort and other’seffortonrelationship quality.Journal of Applied Psychology ,86,697–708.
Mathieu, J. E., &Taylor, S. R. (2006). Clarifying conditions and decision points formediational type
inferences in organizational behavior.Journal of Organizational Behavior,27,1031–1056.
Maurer,T.J., Pierce, H. R., &Shore, L. M. (2002). Perceived beneficiaryofemployee development
activity: Athree-dimensional social exchangemodel. Academy of Management Review,27,
432–444.
Maurer,T.J., &Tarulli, B. A. (1994). Investigation of perceived environment, perceived outcome,
and person variables in relationship to voluntarydevelopment activity by employees. Journal
of Applied Psychology ,79,3–14.
Maurer,T.J., Weiss, E. M., &Barbeite, F. G. (2003). Amodel of involvement in work-related learning
and development activity: The effects of individual, situational, motivational, and agevariables.
Journal of Applied Psychology ,88,707–724.
McCauley, C. D.,&Hezlett, S. A. (2001). Individual development in the workplace. In N. Anderson,
D. S. Ones, H. K. Sinangil, &C.Viswesvaran (Eds.), Industrial, work and organizational
psychology(Vol. 1, pp. 313–335). Thousand Oaks, CA/London: Sage.
Mikkelsen, A., Saksvik, P. ,Eriksen, H., &Ursin, H. (1999). The impact of learning opportunities
and decision authority on occupational health. Wo rk Stress,13,20–31.
Noe, R. A., &Wilk, S. L. (1993). Investigation of the factorsthat influence employees’ participation
in development activities. Journal of Applied Psychology ,78,291–302.
Noe, R. A., Wilk, S. L., Mullen, E. J.,&Wanek, J. E. (1997). Employee development: Issues in
construct definition and investigation of antecedents. In J. K. Ford, S. W. Ko zlowski, K. Kraiger,
E. Salas, &M.S.Teachout (Eds.), Improving training effectiveness in workorganizations
(pp. 153–188). Mahwah, NJ: Erlbaum.
Paparoidamis, N. G. (2005). Learning orientation and leadership quality: Their impact on
salespersons’ performance. Management Decision,43,1054–1063.
Payne, S. C., Yo ungcourt, S. S., &Beaubien, J. M. (2007). Ameta-analytic examination of the goal
orientation nomological net. Journal of Applied Psychology ,92,128–150.
Podsakoff, P. M., MacKenzie, S. B., Lee, J.-Y., &Podsakoff, N. P. (2003). Common method biases in
behavioral research: Acritical review of the literature and recommended remedies. Journal of
Applied Psychology ,88,879–903.
Poell, R. F. ,van Dam, K., &van den Berg, P. T. (2004). Organising learning in workcontexts.
Applied Psychology: An International Review,53,529–540.
Scandura, T. A., &Graen, G. B. (1984). Moderating effects of initial leader–member exchange
status on the effects of aleadership intervention. Journal of Applied Psychology ,69,428–436.
Scandura, T. A., &Schriesheim, C. A. (1994). Leader–member exchange and supervisor career
mentoring as complementaryconstructs in leadership research. Academy of Management
Journal,37,1588–1602.
LMX and employee learning 691
Seijts, G. H., Latham, G. P. ,Tasa, K., &Latham, B. W. (2004). Goal setting and goal orientation: An
integration of two different yetrelated literatures. AcademyofManagement Journal,47,
227–239.
Senge, P. M. (2006). Thefifth discipline: Theart and practice of the learning organization
(Rev.and upd. ed.). New York: Currency Doubleday.
Sobel, M. E. (1982). Asymptotic confidence intervals forindirect effects in structural equation
models. In S. Leinhart(Ed.), Sociological methodology1982 (pp. 290–312). San Francisco,
CA: Jossey-Bass.
Sonnentag, S., Niessen, C., &Ohly,S.(2004). Learning at work: Training and development.
International Review of Industrial and Organizational Psychology ,19,249–289.
Sparrowe, R. T. ,&Liden, R. C. (1997). Process and structure in leader–member exchange.
Academy of Management Review,22,522–552.
Sparrowe, R. T. ,Soetjipto, B. L., &Kraimer,M.L.(2006). Do leaders’ influence tactics relate to
members’ helping behavior? It depends on the quality of the relationship. Academy of
Management Journal,49,1149–1208.
Tepper,B.J., Duffy, M. K., &Shaw, J. D. (2001). Personality moderatorsofthe relationship between
abusive supervision and subordinates’ resistance. Journal of Applied Psychology ,86,
974–983.
Townsend, J.,Phillips, J. S., &Elkins, T. J. (2000). Employee retaliation: The neglected
consequence of poor leader–member exchangerelations. Journal of Occupational Health
Psychology ,5,457–463.
Uhl-Bien, M., &Maslyn, J. M. (2003). Reciprocity in manager-subordinate relationships:
Components, configurations, and outcomes. Journal of Management,29,511–532.
Vancouver, J. B., &Morrison, E. W. (1995). Feedback inquiry: The effect of source attributes and
individual differences. Organizational Behavior and Human Decision Processes,62,
276–285.
van Dam, K., Oreg, S., &Schyns, B. (2008). Daily workcontexts and resistance to organizational
change:The role of leader–member exchange, development climate, and changeprocess
characteristics. Applied Psychology: An International Review,57,313–334.
VandeWalle, D.,Ganesan, S., Challagalla, G. M., &Brown, S. P. (2000). An integrated model of
feedback-seekingbehavior: Disposition,context,and cognition. Journal of Applied
Psychology ,85,996–1003.
Vera, D.,&Crossan, M. (2004). Strategic leadership and organizational learning. Academy of
Management Review,29,222–240.
Wayne, S. J.,Liden, R. C., Kraimer,M.L., &Graf,I.K.(1999). The role of human capital, motivation
and supervisor sponsorship in predicting career success. Journal of Organizational
Behavior,20,577–595.
Wayne, S. J.,Shore, L. M., &Liden, R. C. (1997). Perceived organizational supportand leader–
member exchange: Asocial exchangeperspective. Academy of Management Journal,40,
82–111.
Williams, L. J.,&Podsakoff, P. M. (1989). Longitudinal field methods forstudying reciprocal
relationships in organizational behavior research: Toward improved causal analysis. Research
in Organizational Behavior,11,247–292.
Received 27 May 2008; revised version received 13 July 2009
692 Xander M. Bezuijen et al.
Appendix
Scale items
Employee engagement in learningactivities
Ispendtime following acourse or educational program.
Iamworking to extend my knowledgeand skills.
Iperform learning tasks that are not partofmyjob.
Ispendtime planning and realizing my career.
Igotomysupervisor to discuss how Ican make progress.
Within my task responsibilities, Iactivelylookfor methodstoimprovemywork.
Within my job, Ilook foractivities from which Ican learn.
Icontinually learnnew skills formyjob.
Goal difficulty
How difficultare the goals youhave set with this specific employee?
goals forperformance levels in employee’scurrent job.
goals forpersonal development.
goals forextension of knowledge and skills.
goals forparticipation in an educational program or course.
goals forthe performance of learning tasks within the function.
goals forworking towardsanother job.
Goal specificity
Have you set clear goals, together with your supervisor, for:::
your performance levels in your current job?
your personal development?
your extension of knowledgeand skills?
your participation in an educational program or course?
your performance of learning tasks within the function?
working towards another job.
Feedback
My supervisor :::
informs me of how Ishould performspecific tasks if something goes wrong.
informs me of whetheritwill benefit my career to follow aspecific course or
training program.
informs me of how Ishould undertakenew tasks.
informs me of which skills Ican improve.
LMX and employee learning 693
