Tải bản đầy đủ

Do employee ratings of HR effectiveness impact task performance

Master Thesis Human Resource Studies
Do employee ratings of HR effectiveness impact task performance?
The mediating role of intrinsic motivation and the moderating role of leadermember exchange.

Student: Marloes Leeijen
SNR: 1256805
Supervisor: Jeske van Beurden
Second assessor: Marinus Verhagen
January 2018 - August 2018
Tilburg University – School of Social and Behavioral Sciences


Abstract
The purpose of this study was to get more understanding of the establishment of the link
between HRM and performance. In line with the AMO theory (Appelbaum et al., 2001) it was
assumed that the perceptions employees have about the effectiveness of HR practices (i.e.
employee development, career opportunities, performance management, job design,
communication and information sharing, rewards, participation, work-life balance and job
security) positively impacts their task performance. Furthermore, based on the AMO theory
and the self-determination theory (Ryan & Deci, 2000), the expectations were made that
intrinsic motivation was one of the mechanisms through which perceived effectiveness of HR

practices increases task performance and additionally, that Leader-Member Exchange (LMX)
would strengthen the positive relationship between the employees’ HR effectiveness ratings
and intrinsic motivation. A cross-sectional survey study among 464 employees was conducted
in which we, among others, collected effectiveness ratings of nine different HR practices. The
main findings of this study are that the perceived effectiveness of the HR practices do not lead
to higher levels of task performance and for three HR practices it lead to a lower level of task
performance (i.e. career opportunities, performance management and employee development).
In addition, the mediation effect of intrinsic motivation was only found for two of the HR
practices (i.e. performance management and employee development) and LMX was not found
as a moderator, except for one of the practices (i.e. job security). However, it turned out LMX
did have a positive relationship with intrinsic motivation. Furthermore, the results of this study,
the limitations, the suggestions for future research and the practical and theoretical implications
are discussed.

Keywords: perceived HR effectiveness, task performance, intrinsic motivation, LMX, AMO
theory, self-determination theory

2


Introduction
Researchers in Human Resource Management (HRM) generally agree that the
implementation of HR practices leads to higher organizational performance (Den Hartog &
Verburg 2004). These statements resulted in a more positive view on investment in HRM
among managers (Edgar & Geare, 2014). However, there remains ambiguity on how exactly
the HRM-performance link is established (Edgar & Geare, 2014). For example, there often
seems to be a gap between how HR practices were implemented by managers and how workers
perceive the HR practices (Edgar & Geare, 2014). In line with this is the Strategic Human
Resource Management (SHRM) process model, which shows that there is not a direct
relationship between actual HR practices offered by the organization and the reaction of the
employee to these practices, but what accounts for this relationship is how an employee
perceives the HR practices (Nishii & Wright, 2007). In addition, using HR practices does not
imply that employees experience practices as supportive for their work performance (Nishii &
Wright, 2007). Likewise, according to Meyer and Allen (as cited in Edgar & Geare, 2014), the
perceptions employees have about the HR practices has more impact on their performance than
how many or which HR practices are actually implemented in reality. Therefore, this research
investigates employee perceptions about HR practices in terms of HR effectiveness ratings for
their work performance in relation to employee performance. This is in contrast to previous
studies that were focused on perceptions of HR practices in terms of availability or use of HR
practices (see e.g., Boon, Den Hartog, Boselie & Paauwe, 2011; Den Hartog & Verburg, 20


04; Guthrie, 2001). The focus will be on task performance because this form of employee
performance is related to fulfilling the core tasks of your role in a proper way (Griffin, Neal, &
Parker, 2007).

The

task

performance

will

be

self-reported

by

the

employee.

A problem in the HRM literature, is that we do not sufficiently know what is inside the
so called black box that should explain how and why HR practices increase performance
(Boselie, Dietz & Boon, 2005; Chowhan, 2016; Nishii & Wright, 2007). Therefore, by using
the AMO theory (Appelbaum, Bailey, Berg & Kalleberg, 2001) and the self-determination
theory (Ryan & Deci, 2000) this research will focus on intrinsic motivation as a possible
mechanism through which HR practices, and in particular, the employees’ HR effectiveness
ratings, affect employee performance. The AMO theory states that the performance of an
employee is determined by the combination of the employees’ abilities, motivation and
opportunities (AMO) to perform (Gerhart, 2007). According to Jiang et al. (2012) performance
can be improved by using HRM systems, when these systems are enhancing the three factors
(AMO) of performance. Taking these statements into consideration, one would expect that
3


motivation is one of the mechanism through which employees’ HR effectiveness ratings affect
the level of performance. However, it is important to notice that there are different types of
motivation. The most basic distinction that is often been made is between intrinsic and extrinsic
motivation (Ryan & Deci, 2000). In this research the focus will be on intrinsic motivation
because, by following the self-determination theory, it is expected that when HR practices are
perceived as effective, this will help to fulfil people’s psychological needs (i.e. to feel more
competent, to feel related to others and to possess a level of autonomy), which results in people
being intrinsically motivated and therefore perform better (Ryan & Deci, 2000). Besides,
another factor that is stated in previous research to support an individual's feeling of autonomy,
competence and relatedness, and may therefore enhance intrinsic motivation, is the relationship
between a manager and subordinate which is called: Leader-Member Exchange (LMX)
(Martin, Guillaume, Thomas, Lee & Epitropaki, 2016; Ryan & Deci, 2000). In addition, the
quality of the LMX relationship seems to have a strong impact on the perceptions employees
have regarding the quality of HR practices (Sanders, Moorkamp, Torka, Groeneveld, &
Groeneveld, 2010). Accordingly, it is expected that the positive relationship between HR
effectiveness ratings and intrinsic motivation will be stronger when a good relationship between
manager and subordinate is perceived. As a consequence, LMX will be another focus in this
research.
This research will be of scientific relevance for a couple of reasons. First of all,
employees HR effectiveness ratings will be taken into consideration instead of the perceived
use or availability of HR practices, which was the main focus of previous research (see
e.g., Boon, Den Hartog, Boselie & Paauwe, 2011; Den Hartog & Verburg, 2004; Guthrie,
2001). In addition, this research will help identifying whether intrinsic motivation is a
mechanisms that exist inside the black box, and will therefore contribute to specify the process
through which HR and performance impact each other (Nishii & Wright, 2007). Last, when the
effect of LMX on the relationship between HR effectiveness ratings and intrinsic motivation is
clarified, this can give insight in how to enhance one of the mechanism that exists inside the
black box. Furthermore, investments in HR are often expensive for organizations (Sels et al.,
2002). Therefore it is useful for employers to know whether their investments will pay off in
terms of improved performance of the employee and benefits for the organizational
performance, which indicates the practical importance of this study. This research will give
insight to employers how HRM can have a positive impact on the performance of the employees
and to what extent this is through intrinsic motivation. Besides, when LMX turns out to enhance

4


the relationship between HR effectiveness ratings and intrinsic motivation, this can encourage
employers to pay more attention to the relationship between managers and subordinates.
The information above leads to the following research question:
To what extent do employees’ HR effectiveness ratings affect individual task
performance and to what extent is this relationship mediated by intrinsic motivation and
moderated by LMX?

Theoretical framework
The relationship between employees’ HR effectiveness ratings and task performance
“HR practices represent the activities related to the management of work and
employment in organizations” (Beijer, 2014, p. 3). The study of Beijer (2014) emphasizes the
importance of acknowledging the different measures of the HR practices construct. A
distinction should be made between the intended HR practices and how these practices were
implemented (Beijer, 2014). Besides, a difference should be made between the manager
perspective and the employee perspective of the implemented HR practices (Beijer, 2014).
These different constructs were built on the SHRM Process model of Wright and Nishii (2007),
which explains the process through which HR practices affect performance and the variability
that arises in the different steps of that process. First of all, there seems to be a difference
between the actual HR practices offered by the organization (which Beijer (2014)
conceptualizes: the manager perspective of the implemented HR practices) and how the HR
practices were intended by the management. Second, the perceptions employees have of the
HR practices (which Beijer (2014) conceptualizes to as: the employee perspective of the
implemented HR practices) are often not in line with the actual HR practices. Additionally, the
model emphasizes the importance of the employee perceptions of the HR practices when
predicting the reaction of that employee, which in turn has an effect on the performance (Wright
& Nishii, 2007). Besides, according to Wright and Nishii (2007) employees do not all respond
in the same way to similar HR practices offered by an organization and they do not all share
the same perception of these HR practices. As Beijer (2014) acknowledges it is important to
specify which specific construct of HR practices is studied and as the SHRM Process model
(Wright & Nishii, 2007) shows, the construct of HR practices that is the most direct predictor
of employee reactions to HR practices is the perceptions they have of these practices.
Accordingly, in this research the focus will not be on the intended or actual HR practices offered
by the organization but on the perceptions employees have of these practices. In more detail,
the independent variable in this study will be employee’s perceived effectiveness of the HR
5


practices offered by the organization and in particular, how effective the employees perceive
these practices for their performance. The nine HR practices that are examined in this research
belong to the general categories of HR practices that are most used in current research (Boselie
et al., 2005) and involve: employee development, career opportunities, performance
management, job design, communication and information sharing, participation, work-life
balance, job security and rewards. In this research the perceived effectiveness of all nine
practices will be examined separately because the HR practice approach was followed. This
approach sees HRM as: “a collection of multiple, separate practices without any mutually
reinforcing effects” (Veld, 2012, p. 132). This approach is used in order to be able to find out
which practices are most important in developing the desirable outcomes (Veld, 2012) and is
therefore used in this study to be able to give extensive, practical recommendations to
organizations. When the HR practices would be approached as a bundle, detailed advice about
the differences in effect of the distinct HR practices could not be given.
In addition, the dependent variable in this research is task performance. In this research
the performance of individual employees will be measured and this will be done at the
proficiency level. When an employee’s performance is classified to the level of proficiency he
or see “fulfills the prescribed or predictable requirements of the role” (Griffin et al, 2007,
p.330). The concept of the individual level of proficiency is closely related to task performance,
hence, the dependent variable in this research is called task performance.
The AMO theory (Appelbaum et al., 2001) is frequently used to explain the link
between HRM and performance. Whereas other theories are often focused on organizational
performance, the AMO theory takes into account individual features of employees and is
therefore applicable to the performance of the individual employee (Paauwe, 2009). The AMO
theory indicates that whether employees are able to perform well, depends on three factors
(Bos-Nehles, Van Riemsdijk, & Kees Looise, 2013). First of all, an employee should have the
ability to perform, which means he or she should have possessing of the right knowledge and
competences. Second, the employee should have the motivation to perform, because they feel
like they have sufficient reasons and get properly rewarded for it. Last, the employee should
have the opportunity to perform, which depends on how supportive the work environment is
(Bos-Nehles et al., 2013). When the HR practices are deployed in an effective way, they can
enhance employee’s ability, motivation and opportunity, which will affect the individual task
performance in a positive way (Jiang et al., 2012). Therefore, by following the AMO theory,
one would expect that when the HR practices are considered as effective, this enhances the task
performance. Besides, previous research also supports the assumption that (employees’
6


perceptions of) HR practices are positively related to performance and that this may occur by
influencing the abilities, motivation and attitudes of employees (Alfes, Truss, Soane, Rees &
Gatenby, 2013; Huselid, 1995; Parker et al., 2003). In accordance, Wright and Nishi (2007)
state that employee outcomes are influenced by the perceptions employees have of the HR
practices offered by the organization. Therefore, when following the AMO theory and empirical
evidence, a positive relationship would be expected between the perceived effectiveness of HR
practices and task performance
Hypothesis 1: Employees’ HR effectiveness ratings for their work performance is
positively associated with task performance.
The relationship between employees’ HR effectiveness ratings, intrinsic motivation and
task performance
When someone is motivated this means he or she is “moved to do something” (Ryan &
Deci, 2000, p. 54). There is not only a difference in how motivated people are, but also what
the orientation of the motivation is (Ryan & Deci, 2000). A distinction has been made between
two types of motivation, namely, extrinsic motivation and intrinsic motivation. Extrinsic
motivation is defined as: “doing something because it leads to a separable outcome” (Ryan &
Deci, 2000, p. 55). Whereas intrinsic motivation refers to “doing something because it is
inherently interesting or enjoyable” (Ryan & Deci, 2000, p. 55). In this study the focus will be
on intrinsic motivation in relation to HR effectiveness ratings and task performance, which will
be

explained

using

the

AMO

theory

(Appelbaum

et

al.,

2001) and self-

determination theory (Ryan & Deci, 2000).
In the previously discussed AMO theory, it is explained that motivation is an
important factor for explaining the relationship between HR and performance: employees are
able to perform well, when they feel like they have sufficient reasons and get properly rewarded
for it (Bos-Nehles et al., 2013). To explain the motivational aspect of the model in more detail,
the notion of reciprocity should be taken into consideration. This means that the way an
employee is treated by the organization, will be exchanged by the corresponding behavior of
an employee (Boxall & Purcell, 2016). Accordingly, employees’ willingness to perform
properly for the organization, depends on the way the organization performs towards the
employee (Boxall & Purcell, 2016). Therefore, when employees feel like the organization wants
to invest in them, by means of effective HR practices, they will respond by means of devotion
and exertion, which enhances the performance (Boxall & Purcell, 2016). By following this
7


explanation and the AMO theory, it is expected that motivation is one of the mechanisms
through which perceived effectiveness of HR practices increases task performance. Because
other mechanisms are explained by the AMO theory as well (i.e. ability and opportunity), one
would expect intrinsic motivation only to partially mediate this relationship.
Why specifically intrinsic motivation is of interest in this study can be explained using
the self-determination theory of Ryan and Deci (2000). This theory explains that there are three
basic human needs and by fulfilling these needs, a person’s well-being will be optimized
(Seligman & Csikszentmihalyi, 2014; Schultz & Schultz, 2016). The three basic human needs
are: the need for competence, when one has confidence that he or she can complete difficult
tasks. The need for autonomy: when one has the freedom to decide how he or she does the work,
based on their interests, needs and values. And last, the need for relatedness: when one feels
closely connected to others (Seligman & Csikszentmihalyi, 2014; Schultz & Schultz, 2016).
Ryan and Deci (2000) state that when the three basic human needs are satisfied, people’s
intrinsic motivation will be stimulated. We expect that higher HR effectiveness ratings of the
aforementioned HR practices can help fulfilling the three basic needs. For example, when an
employee believes following a certain training was effective, this can satisfy the need for
competence, or when he or she has the opportunity to have an influence on their job design, this
can satisfy the need for autonomy. Last, Baard, Deci and Ryan, 2004 conducted a study which
turned out that satisfaction of the three basic human needs also leads to higher work
performance. As these needs result in intrinsic motivation, this indicates the relevance of
intrinsic motivation for task performance as well. Based on the literature and theories discussed,
the

following

hypotheses

are

formulated:

Hypothesis 2: Employees’ HR effectiveness ratings are positively associated with
intrinsic motivation.

Hypothesis 3: Intrinsic motivation is positively associated with task performance.
Hypothesis 4: The positive relationship between employees’ HR effectiveness ratings
and task performance is partially mediated by intrinsic motivation.

8


The moderating role of LMX in the relationship between HR effectiveness ratings and
intrinsic motivation
Leader-Member Exchange (LMX) focuses on the dyadic and unique relationship
managers possess with each subordinate (Ilies, Nahrgang, & Morgeson, 2007). The theory
associated with LMX is the LMX theory and gives a broader understanding of the concept. It
explains that leaders (managers) differ in the quality of exchange relationship they have with
each of their subordinates and that this has an influence on the behavior and attitudes of the
manager and subordinate (Ilies et al., 2007). Furthermore, Sanders et al. (2010) state that a
relationship of good quality between a manager and an employee and the employees’
experience of the quality of the HR practices can stimulate each other. Therefore, one might
expect that the relationship between employees’ HR effectiveness ratings and intrinsic
motivation will be stronger when the relationship between manager and subordinate is better.
The aforementioned self-determination theory can help to explain the expected
strengthening effect of LMX, because it is expected that LMX affects the fulfillment of the
three needs explained by this theory (Gagné & Deci, 2005; Graves & Luciano, 2013). An
employee’s level of fulfillment of the need for competence, autonomy and relatedness is
influenced by the perception that the employee has of the quality of the relationship they have
with their manager (Graves & Luciano, 2013). In a high-quality exchange relationship, the
manager will help the employee in developing necessary skills and abilities, the manager will
give the employee freedom and autonomy and last, the manager will make sure an employee
feels valued, encouraged and supported (Erdogan & Enders, 2007; Graves & Luciano, 2013;
Henderson, Wayne, Shore, Bommer & Tetrick, 2008). Therefore, the need for competence,
autonomy and relatedness can be further fulfilled when experiencing a positive relationship
with the manager. Also, as previously explained, we expect that perceived HR effectiveness
can help to fulfil the three basic human needs, which results in an increased level of intrinsic
motivation (Ryan & Deci, 2000). When we suppose LMX also helps to satisfy those needs, we
can expect that when a positive LMX relationship exists, this will strengthen the effect of
perceived HR effectiveness on intrinsic motivation. Besides, as Ryan and Deci (2000) state that
when the three basic human needs are fulfilled, this will stimulate the intrinsic motivation, it is
also expected that LMX has a direct, positive effect on intrinsic motivation. Therefore, we come
to the following hypotheses:

9


Hypothesis 5: Leader-Member Exchange will strengthen the positive relationship
between

employees’

HR

effectiveness

ratings

and

intrinsic

motivation.

Hypothesis 6: Leader-Member Exchange is positively associated with intrinsic
motivation.

Figure 1. Conceptual model
Note. HR effectiveness ratings include the effectiveness ratings of the following HR practices: employee development,
career opportunities, performance management, job design, communication and information sharing, rewards,
participation, work-life balance and job security.

Method
Research design
A quantitative, cross-sectional study has been executed. This explanatory study tested
the conceptual model displayed in Figure 1. 10 master students collaboratively collected the
data for this research and searched for respondents by using their own network. The intention
was that every student would invite 50-100 respondents, which means the total, intended sample
was 500 to 1000 employees.

Population and sample
In the end, the actual number of employee respondents of this study was 464 (N=464),
with a response rate of 62.3%. Additionally, the number of manager respondents was 76
(N=76), with a response rate of 80.9%. The demographic characteristics of this sample can be
found in Table 1 in Appendix A. The population of this study were employees who are working
in organizations in five different industries in the Netherlands. These five industries include
10


business services (30.4%), government (3.6%), healthcare (20.1%), education (16.1%) and
production (8.2%). 41.3% of the respondents were male and 58.7% were female and the average
age was 33.58 years (SD=12.47). The average organizational tenure was 8.45 years
(SD=10.26). A possible problem that could occur in this research was that common method
bias would arise (spurious variance that is attributable to the measurement method rather than
to the constructs the measures are assumed to represent (Organ & Ryan, 1995; Podsakoff &
Todor, 1985)). To reduce the common method bias, employees were requested to ask their
manager to cooperate in this research too by filling in a short questionnaire about the same
subjects as the employee questionnaire. As an example, they were asked to rate the performance
of the employees, in order to be able to compare the performance rated by the employee itself
and the rating of the manager. Because of the small amount of manager responses, it was
decided to not include them in the main analyses of the research.

Procedure
For selecting respondents, convenience sampling was used. This is a way of sampling
where the respondents are selected because they are easily assessable for the researchers
(Etikan, Musa, & Alkassim, 2016). This way of sampling (in contrast to, for example, random
sampling) can have negative consequences for the generalizability of the sample to the overall
population, however for the feasibility of the research the decision for using convenience
sampling was made. Participants were either contacted personally by one of the researchers or
they were attracted by a flyer (Appendix J) which was distributed on the social media accounts
of the researchers, after which they contacted the researchers themselves. The procedure of data
collection was both on individual level and on manager level, as both managers and employees
were approached. When a whole organization or a whole team was willing to take part in this
research, random sampling has been applied to select for which employees the manager was
asked to fill in the questionnaire (the random sample was based on the initials of the
participants). The data in this research was collected using a questionnaire (see Appendix G).
Respondents who participated in the research received a cover letter and informed consent,
including information before filling in the questionnaire and the purpose of the study and which
assured the confidentiality of the data (see Appendix F). A separate questionnaire and cover
letter was made for the employees and for the managers (see Appendix H and I). Respondents
were able to choose between filling in the questionnaire online or a hard copy version and could
choose between a Dutch or English version. The online version of the questionnaire was created
with the program Qualtrics. The hard copy questionnaire was distributed among the employees
11


at their workplace by one of the researchers. The employees could hand in their finished
questionnaire to the manager in a closed envelope.

Instruments
Although most of the scales used in this research were already validated in previous
research, the validity of all the scales has been checked again by doing a factor analysis
(Principal Component Analysis). Two requirements have been taken into account when
examining the output. First, the Kaiser-Meyer-Olkin (KMO) measure needs to be higher than
0.6. Second, Barlett’s Test of Sphericity has to be significant (p<0.05). Besides, to test the
reliability of the scales, a reliability analysis was conducted which presented the Cronbach’s
Alpha (α). When Cronbach’s alpha (α) has a score of 0.7 or higher, it is considered as ‘good’,
when it is between 0.6 and 0.7 it is ‘sufficient’ and when it is lower than 0.6 the reliability of
the scale is ‘insufficient’ (Evers, van Vliet-Mulder, & Groot, 2000). The results of the analyses
can be found in appendix A, Table 2.
Task performance. Task performance has been measured with the three items of the
individual task proficiency scale created by Griffin et al. (2007). The respondents had to rate
how often they participated in certain behavior, this was done on a 7-point Likert scale, ranging
from 1 (never) to 7 (always). An example of an item of this scale is: “carried out the core parts
of your job well”. This scale was filled in by the employee and the manager, both rating the
employee’s performance. A principle component analysis was carried out to check the validity
of the scale. The Kaiser-Meyer-Olkin (KMO) showed a score of .761, which is higher than the
norm of .6 and also Bartlett’s Test of Sphericity is significant, therefore we can conclude that
all the items belong to one factor. Next, a reliability analysis was conducted to check the
reliability of the scale. The Cronbach’s alpha of the scale was 0.812 (α), which is above 0.7, so
the reliability of the scale is considered as good (Evers et al., 2000).
Leader-Member Exchange. Leader-Member Exchange was measured by using the
LMX-7 scale of Graen and Uhl-Bien (1995). This scale consists of seven items which were
measured on a 5-point Likert scale, ranging for 1 (totally disagree) to 5 (totally agree). An
example of an item of this scale is: “My manager understands my problems and needs”. Again
a principle component analysis was carried out, the results showed a KMO value of .874 (>.6)
and Bartlett’s Test of Sphericity is significant. Accordingly, all items belong to one factor. The
Cronbach’s alpha of the scale is .863 (>.7) and therefore the scale’s reliability is assessed as
good (Evers et al., 2000).

12


Intrinsic motivation. Intrinsic motivation will be measured using the scale of Gagné et
al. (2015). The intrinsic motivation scale consists of three items measured on a 7-point Likert
scale ranging from 1 (not at all) to 7 (completely). The stem of the items is: “Why do you or
would you put efforts into your current job?” and an example of an item is: “Because I have
fun doing my job”. The results of the principle component analysis showed that the items were
loaded on one factor because the KMO value is .743 and again Bartlett’s test of Sphericity is
significant. Furthermore the Cronbach’s alpha of the scale is .90 (>.7), thus the scale is highly
reliable (Evers et al., 2000).
HR effectiveness ratings. HR effectiveness ratings was measured using a scale which is
based on the scales created by Veld, Paauwe, and Boselie (2010) and Boon, Den Hartog,
Boselie, and Paauwe (2010). In this study, items of nine HR practices were included,
namely: employee development, career opportunities, performance management, job design,
communication and information sharing, participation, work-life balance, job security and
rewards. The total scale consisted of 32 items. In the analyses, the effectiveness ratings of the
different HR practices will be examined as nine different variables. The questions were
answered on a 5-point Likert scale ranging from 1 (totally disagree) to 5 (totally agree).
However, to provide a more simplistic way for interpreting the results, the scores were recoded
before doing the analyses, whereby a score of 1 or 2 was coded as -1, a score of 3 was coded as
0 and a score of 4 or 5 was coded as 1. First the respondent received a question whether or not
they make use of a certain practice, based on their answer they received a follow-up question
concerning how effective they perceive the HR practice is for their work performance. First of
all, when the respondent did use a certain practice, an example of an item of HR effectiveness
ratings is: “My work results have improved in the last year as a result of making my own
decisions in work.” When the respondent did not use the practice, an example of a follow-up
question of HR effectiveness ratings is: “My work results would have improved in the last year
if I would have made my own decisions in work.” For the analyses, the answers of the group
who had used the practice and the group who had not used the practice were merged to one
scale which presents the overall HR effectiveness rating.
A confirmatory factor analysis (CFA) was carried out to assure construct validity of the
use of HR practices scale. To do this analysis the program Mplus (Muthén, 2011) has been used
since this program is able to deal with dichotomous variables. For evaluating the model fit, the
recommendation of Hu and Bentler (1998) were followed by using multiple indices of fit, which
includes the chi square statistic (χ2), the comparative fit index (CFI; acceptable >.9 and good
>.95), the Tucker-Lewis index (TLI; acceptable >.9 and good >.95) and the root mean square
13


error of approximation (RMSEA; acceptable <.08, however preferably close to .06).
The CFA which forced all the 32 items to load on one factor resulted in fit indices which
are considered as low (see Appendix A, Table 3). The 9-factor model that was expected showed
sufficient model fit (see Appendix A, Table 3). However, three items (one item of job design,
one of performance management and one of rewards scale, see Appendix A, Table 3) had a low
score on the factor that was expected and had little variance, that is the reason why these items
were removed. The 9-factor-II model for use of HR practices showed good fit indices when
the three items were removed (9-factor-II (χ²(341) = 574.314, Comparative Fit Index (CFI) =
.954, Tucker Lewis Index (TLI) = .946, Root Mean Square Error of Approximation (RMSEA)
= .038 (see Appendix A, Table 3)). The above mentioned CFA was conducted for the use of
HR practices items. As the HR effectiveness rating items were the follow-up questions to these
use items, the corresponding HR effectiveness items were removed from the scales as well.
Next, a principle component analysis and reliability analysis in SPSS were conducted for the
HR effectiveness ratings scales of all the nine different practices. The results can be found in
appendix A, Table 2. From the results it can be concluded that the validity and reliability of the
scales can be considered as sufficient.
Control variables. For testing spuriousness in the relationships between the different
variables, in the analyses of this study control variables were added. First of
all, gender (0=male, 1=female) is included because previous research has shown that women
often underestimate their abilities and therefore tend to score lower on self-evaluations of
performance (Beyer, 1990; Carr, Thomas, & Mednick, 1985). Second, organizational
tenure (measured in years) will be included because evidence found that tenure is positively
related to in-role performance and to intrinsic motivation (Kuvaas, 2006; Ng & Feldman, 2010).
Also previous research suggests that age can have both a negative and a positive effect on job
performance, which seems to differ for different professions and the effect that was found
differs per research (Rhodes, 1983), therefore this variable will also be included as a control
variable in this research.

Data analysis
The program IBM SPSS Statistics version 24 was used to test the model and
corresponding hypotheses of this research. First of all, the data was checked for errors and
missing data. To coop with the missing data, list wise deletion has been applied while doing the
analyses.
The Hayes PROCESS macro in SPSS (Hayes, 2013) was used to test the hypotheses of
14


the conceptual model of this study. The PROCESS Macro model 7 was used to check the
relationships present in the model of this research. Nine separate analyses were conducted
concerning the effectiveness ratings of the nine different practices. In particular, Model 7 has
been used to test the direct effect between HR effectiveness ratings of the different HR practices
and task performance (H1), the direct effect of the HR effectiveness ratings on intrinsic
motivation (H2), the direct effect of intrinsic motivation on task performance (H3), whether the
direct effect of HR effectiveness ratings and task performance is (partially) mediated by
intrinsic motivation (H4) and whether the direct relationship between HR effectiveness ratings
and intrinsic motivation is moderated by LMX (H5) and therefore whether a moderated
mediation exists in this model. Last the direct effect of LMX on Intrinsic motivation (H6) was
tested by using model 7. All the control variables were added to this model as well. It was
decided to not include the effectiveness ratings of the other HR practices as control variables in
the analyses, but to analyze all practices individually (this decision was based on the HR
practice approach explained earlier in this study (Veld et al., 2010; Veld, 2012)). When model
7 gives insignificant relationships, we are able to test the separate relationships using different
PROCESS models. We will test the separate models to see if there are any differences in terms
of significance, regarding the separate relationships in the conceptual model. The moderation
effect of LMX on the relationship between HR effectiveness ratings and intrinsic motivation
can be tested using PROCESS macro model 1. Besides, the mediation effect of intrinsic
motivation in the relationship between HR effectiveness ratings and task performance can be
checked using PROCESS macro model 4.

Results
Descriptive statistics
In Table 1 the means (M), standard deviations (SD) and correlations between the
variables can be found. The significant correlations between the main variables of the
conceptual model will be discussed. The exact numbers and the correlations of the control
variables can be found in Table 1.
First of all intrinsic motivation is positively related to LMX and task performance.
Besides, LMX is also positively related to task performance. For the HR effectiveness ratings
of the separate practices intrinsic motivation is positively related to performance management,
job design, communication and information sharing and participation and negatively to
rewards.

LMX shows both negative (career opportunities and rewards) and positive

correlations (performance management, job design and communication and information
15


sharing) to the effectiveness ratings. And last, for task performance only three significant
correlations with HR practices were found, which were all negative (employee development,
career opportunities and performance management). What is notable, is that the correlations
between the different variables are low (>.29) (Cohen, as cited in Pallant, 2010).

16


Table 1. Means (M), standard deviations (SD) and correlations (N=464)
Variable

M

SD

1.

1. Age
2. Gender

2.

3.

4.

5.

6.

7.

8.

9.

10.

11.

12.

13.

34.78

12.93

1

1.54

.50

-.16**

1

3. Tenure

8.68

10.34

.74**

-.10*

1

4. LMX

3.68

.60

.02

.11*

.02

1

5. Intrinsic

5.17

1.16

.16**

.08

.10*

.29**

1

5.66

.75

.07

.08

.06

.13**

.23**

1

.53

.51

-.20**

.01

-.14**

.04

.17*

-.15**

1

-.00

.61

-.18**

-.13**

-.11*

-.12**

-.04

-.18**

.36**

1

.09

.63

-.25**

.00

-.18**

.14**

.10*

-.14**

.40**

.30**

1

.62

.47

-.08

.01

-.11*

.12**

.21**

.02

.36**

.22**

.22**

1

.16

.60

-.04

.02

-.08

.13**

.11*

-.02

.16**

.14**

.20**

.29**

1

.12

.59

.10*

-.14**

.03

.03

.17**

-.02

.23**

.27**

.17**

.35**

.58**

1

.15

.62

-.08

.09

-.07

.03

.06

-.07

.15**

.14**

.19**

.28**

.28**

.23**

1

.04

.70

-.02

-.04

-.01

-.03

.05

.01

.16**

.17**

.15**

.24**

.32**

.27**

.28**

-.15

.66

-.26**

-.14**

-.18**

-.16**

-.17**

-.05

.10*

.17**

.06

.08

.14**

.14**

.17**

.17

.33

-.20**

-.07

-.17**

.02

.11*

-.12*

.55**

.55**

.53**

.56**

.62**

.64**

.55**

14.

15.

16.

motivation
6. Task
performance
7. Emp. Dev.
Effectiveness
8. Career Opp.
Effectiveness
9. Perf. Man.
Effectiveness
10. Job design
Effectiveness
11. Com/ Inf
Effectiveness
12. Participation
Effectiveness
13. Work-life
Effectiveness
14. Job security

1

Effectiveness
15. Rewards

.25**

1

Effectiveness
16. Bundle HR

.59**

.44**

effectiveness

Note. ** p < .01. * p < .05, min/max scores: Age (years); Gender (1=male, 2=female); Tenure (years); LMX (1-5); Intrinsic motivation (1-7); Task performance (1-7); HR Effectiveness scales
(1-5) = recoded (1=-1, 2=-1, 3=0, 4=1, 5=1)

17

1


Testing the conceptual model
Hypothesis 1 stated that HR effectiveness ratings is positively associated with task
performance. The results from process model 4 show that for three HR practices the opposite
effect is found (career opportunities (B=-.20, p<.01, see Table 1, model 2, Appendix B),
performance management (B=-.20, p<.01, see Table 2, model 2, Appendix B) and employee
development (B=-.27, p<.001, see Table 3, model 2, Appendix B)), namely, a negative
relationship between the effectiveness ratings of the HR practices and task performance. This
means that employees who rate these HR practices as more effective, score lower on task
performance. For the other HR practices, no significant relationships with task performance
were found (see Table 4, 5, 6, 7, 8 and 9, model 2, Appendix B). To conclude, hypothesis 1 was
rejected.
Hypothesis 2 predicted a positive relationship between HR effectiveness ratings and
intrinsic motivation. For the effectiveness ratings of the following HR practices the expected
relationship was found: performance management (B=.21, p<.05, see Table 2, model 1,
Appendix B), employee development (B=.33, p<.01, see Table 3, model 1, Appendix B), job
design (B=.47, p<.001, see Table 4, model 1, Appendix B) and participation (B=.31, p<.001,
see Table 5, model 1, Appendix B). The more effective these practices are considered, the
higher the level of intrinsic motivation of the respondent. For the other HR practices no
significant relationship was found with intrinsic motivation (see Table 1, 6, 7, 8 and 9, model
1, Appendix B). As a consequence hypothesis 2 was partially accepted.
Hypothesis 3 stated that intrinsic motivation is positively associated with task
performance. The Process analyses (model 4) confirm this hypothesis (e.g. B=.14, p<.01, see
Table 1, model 2, Appendix B). This means that the higher an employee’s intrinsic motivation
is, the higher their task performance will be. Therefore this hypothesis was accepted.
Hypothesis 4 stated that the positive relationship between employee’s HR effectiveness
ratings and task performance is partially mediated by intrinsic motivation. Results showed no
positive relationship between employee’s HR effectiveness ratings and task performance.
Besides for most of the HR practices the total effect from HR effectiveness ratings on task
performance was insignificant (see Table 4, 5, 6, 7, 8 and 9, Appendix B). One of the conditions
for a mediation effect to exist, is a significant total effect between the independent and
dependent variable. As this is not the case for these variables, we can conclude that intrinsic
motivation is not a mediator in these relationships. However intrinsic motivation does partially
mediate the negative effect of the effectiveness ratings of the performance management practice
(B=.03, LLCI=.01, ULCI=.06) and the employee development practice (B=.05, LLCI=.02,
18


ULCI=.09) on task performance. For the negative relationship between the effectiveness ratings
of career opportunities and task performance no evidence was found that intrinsic motivation
mediates this relationship, because the indirect effect is insignificant (see Table 1, Appendix
B). Given these findings, the hypothesis was partly accepted.
Hypothesis 5 stated that Leader-Member Exchange will strengthen the positive
relationship between employee’s HR effectiveness ratings and intrinsic motivation. Solely for
the HR practice job security a significant interaction effect was found (B=-.38, p<.05, see Table
8, model 3, Appendix B). However, the interaction effect was only found to be significant at
low levels of LMX (B=.34, LLCI=.08, ULCI=.61). Therefore the hypothesis was only accepted
for this specific effect. However for the other HR practices, insignificant interaction effects
were given (see Table 1, 2, 3, 4, 5, 6, 7 and 9, model 3, Appendix B). Accordingly, for the HR
practice job security at medium and high levels of LMX and for the other HR practices, the
hypothesis was rejected.
Last, hypothesis 6 predicts that Leader-Member Exchange is positively associated with
intrinsic motivation. Results of the Process analyses show that there is indeed a positive
relationship between LMX and intrinsic motivation (e.g. B=.51, p<.001, see Table 1, model 3,
Appendix B). Therefore, the results confirm that when employees experience higher levels of
LMX, their intrinsic motivation increases. As a result, hypothesis 6 was accepted.
In table 2 an overview of the findings per hypothesis can be found.

Table 2. Overview of findings per hypothesis.
Hypothesis
H1

Main findings
Opposite confirmed
Rejected

H2

Confirmed

H3
H4

Rejected
Confirmed
Confirmed
Rejected

H5

Confirmed
Rejected

H6

Confirmed

career opportunities, performance management, employee
development
job design, participation, communication and information sharing,
work-life balance, job security, rewards
performance management, employee development, job design,
participation
career opportunities, communication and information sharing,
work-life balance, job security, rewards
all practices
performance management, employee development
career opportunities, job design, participation, communication and
information sharing, work-life balance, job security, rewards
job security (only at low levels of LMX)
career opportunities, performance management, employee
development, job design, participation, communication and
information sharing, work-life balance, job security (at medium and
high levels of LMX), rewards
all practices

19


Post-hoc analyses
As can be read above, the main results of the analyses are not always in line with the
expectations. What is most notable, are the negative and insignificant effects between the
effectiveness ratings of the HR practices and task performance. Because this is not in
correspondence with the literature, some post-hoc analyses were conducted to find out whether
there are factors that might change the results. In order to easily compare the results of the posthoc analyses to original analyses, the effectiveness ratings of the different HR practices were
combined in a bundle and a process analysis was conducted with the bundle of HR practices as
the independent variable. The results show that the effectiveness ratings of all HR practices
together have a negative effect on task performance (B=-.32, p<.01, see Table 1, model 2,
Appendix C).
Furthermore, a post-hoc analysis was performed using the manager ratings of the task
performance of the employees, to compare these outcomes to the outcomes in the regular
analyses where the employees own ratings of their task performance were used. The results of
this process analysis can be found in Table 2. Unfortunately, this analysis does not show any
significant results that are related to the relationship between HR effectiveness ratings and task
performance (see Table 2, model 2, Appendix C). In addition, the number of manager responses
was quite low (N=81 in comparison to N=446 for the employee responses), which means there
is a possibility that these responses are not representative for the whole population, that is the
reason why the decision has been made to not include the manager data in this research.
In addition, Griffin et al. (2007) make a distinction between three levels of performance,
namely Proficiency, Adaptivity and Proactivity. Task performance is related to the proficiency
level of performance and therefore the other forms of performance were not taken into account
in the main model of this research. However, data was also collected for adaptivity and
proactivity. Because a negative effect was found between the HR effectiveness ratings and task
performance, which is the opposite of what was expected when taking earlier research into
account, post-hoc analyses were performed to check whether the effect differs for the other
forms of performance. More information about the scales that have been used and the reliability
and validity measurements of these scales can be found in appendix E. For the first analysis,
task adaptivity was used as the dependent variable. What is notable, is that in the results of this
process analysis can be seen that a positive relationship exists between the overall (bundle) HR
effectiveness ratings and task adaptivity (B=.26, p<.05, see Table 3, model 2, Appendix C).
When conducting the same analysis with task proactivity as the dependent variable, a positive
relationship is found as well (B=.79, p<.001, see Table 4, model 2, Appendix C). To conclude,
20


when the bundle of HR practices is assessed as more effective, the task performance of the
employees will decrease, but the task adaptivity and task proactivity of the employees will
increase.

Discussion
The purpose of this study was to get more insight in the establishment of the HRMperformance link. It was assumed that the perceptions employees have about the effectiveness
of HR practices has an influence on their level of performance. However, the results of this
study show that for most of the HR practices, no relationship was found between employee’s
HR effectiveness ratings and the employee’s task performance. For three of the HR practices
that were included in this research a negative relationship was found between the HR
effectiveness ratings and task performance. These practices were career opportunities,
performance management and employee development. The findings are not in line with the
expectation, which was based on the AMO theory (Appelbaum et al., 2001), that when the
organization offers effective HR practices to employees, the abilities, motivation and
opportunities of these employees will be enhanced, which results in a higher level of task
performance (Jiang et al., 2012). The results of the post-hoc analyses, show that on the other
two forms of performance (task adaptivity and task proactivity), HR effectiveness ratings did
have the expected effect. Thus, when employees perceive the HR practices offered by the
organization as effective, they will be better able to adjust themselves to changing factors at the
workplace and they show more initiative in stimulating and facilitating changes when needed
(Griffin et al, 2007).
A cause of the unexpected findings might be the way effectiveness was measured in
this research. The respondents first got a question whether they use a certain practice or not and
depending on the answer, they got a follow-up question whether the practice had been effective
for enhancing their performance or, when they indicated they had not used the practice, whether
they thought the practice might have been effective. In the analyses, no difference was made
between respondents who had used the practices or who had not used the practices. However,
the scores of the two groups of respondents do in fact not mean the same. The scores of the
group of respondents who had used the practices indicate how effective they rate the practices,
whereas the scores of the other group indicate the expected effectiveness as they do not have
experience in using the practice. Table 1 (Appendix K) shows an overview of the distribution
of respondents who did use a certain practice or did not use a certain practice. The respondents

21


who did not use a practice could only make an estimation whether using a certain practice would
have been effective for their performance, instead of rating how effective they really perceived
it was. Gardner and Wright (2009) explain in their study that measurement errors can occur
when the reasoning of respondents taking part in the research is (unconsciously) biased by
implicit theories. In more detail, this means the answers given in the survey are not fully based
on proper, valid information, but also using mental models that exist because of the way
information is processed and stored within the brain (Gardner & Wright, 2009). The result is
that people make assumptions based on earlier experiences or existing information, which is
not always accurate in the given situation. For the current study, it could be possible that
respondents that indicated they had not used the practices, made use of implicit theories to make
an estimation of the effectiveness of the HR practices, as they do not have the assess to valid
information to base their opinion on. This way of reasoning might have affected the results of
the perceived effectiveness variable, which might have affected the measured relationship
between perceived HR effectiveness and task performance, which resulted in insignificant
relationships.
For three of the HR practices that were included in this research a negative relationship
was found between HR effectiveness ratings and task performance, namely for career
opportunities, performance management and employee development. A possible explanation
for these unexpected results might be that they do not affect performance in the short term, but
only after a given amount of time. The expectation might be that these practices help to enhance
performance, as they help employees to develop and learn and to achieve their work goals. This
could be the cause for positive effectiveness ratings of respondents. However these three
practices might ask for effort at short notice, which might be the cause the performance is not
influenced instantly, but will raise in the long-term (e.g. following a training might help to more
easily achieve work goals in the future, however employees first need to put effort and time in
it in the short term). Whereas the other practices (such as for example work-life balance,
communication and information sharing and job design) might affect the task performance
directly.
To explain the relationship between intrinsic motivation and task performance, the
AMO theory (Appelbaum et al., 2001) was used which suggests that whether employees are
able to perform well, partly depends on their level of motivation. Besides, according to Jiang et
al. (2012) when HR practices are deployed in an effective way, they can enhance the level of
motivation. That is why it was expected that perceived HR effectiveness would enhance
intrinsic motivation, which would again enhance task performance. The results of this study
22


show that for a couple of HR practices (performance management, employee development, job
design and participation) there indeed exists a positive relationship between the perceived level
of effectiveness and intrinsic motivation. However, for the other practices (career opportunities,
communication and information sharing, work-life balance, job security and rewards) no effect
was found. Besides, the expectation that intrinsic motivation enhances task performance was
confirmed by the results.
An explanation for the variance in effect of the effectiveness ratings of the different
practices on intrinsic motivation could be found in the self-determination theory (Ryan & Deci,
2000) that was used to explain why HR effectiveness ratings especially have an effect on this
form of motivation. This theory explains that when three basic human needs are fulfilled,
intrinsic motivation will be enhanced (Ryan & Deci, 2000). The three basic human needs are:
the need for competence, when one has confidence that he or she can complete difficult tasks.
The need for autonomy: when one has the freedom to decide how he or she does the work, based
on their interests, needs and values. And last, the need for relatedness: when one feels closely
connected to others (Seligman & Csikszentmihalyi, 2014; Schultz & Schultz, 2016). The
practices that were found to have a positive effect on intrinsic motivation can easily be linked
to one of the human needs. Effective performance management and employee development can
help to accomplish the need for competence, because people get guidance or training to be
better able to complete difficult tasks. The need for autonomy can be fulfilled by means of
effective job design and participation because these practices make employees able to have an
influence on how the work should be done. Last, the practice participation might also help
achieving the need of relatedness, because being able to participate in the decision making of
the organization might make the employee feel more connected to the other people in the
organization. It can be possible that high HR effectiveness ratings indeed enhance the level of
motivation, however some of the practices (career opportunities, communication and
information sharing, work-life balance, job security and rewards) might have a smaller impact
on the fulfilment of the three basic human needs and as a consequence the effectiveness ratings
of these practices do not have an influence on this specific form of motivation. For example,
work-life balance might have an influence on the need for autonomy as it gives more freedom
when and where the work can be done, however it has less effect on the actual job characteristics
than the practice job design. Also communication and information sharing is closely related to
participation and therefore one might think it affects the need for relatedness, however
communication and information sharing is mostly focussed on receiving information from the
organisation, whereas for participation the employee has a more active role itself. Next, for job
23


security and career opportunities a relationship with the need for competence can be expected.
When an organization gives the employee the opportunity to stay inside the organization or to
get a different or better function, the employee might expect he is able to complete his tasks
well. However it does not directly help to get better at completing those tasks and therefore the
effect on this need might not be as relevant as the effect of, for example, employee development.
Last, for rewards no clear relationship with one of the needs can be made, accordingly for this
practice it is most clear a relationship with intrinsic motivation is not found. To conclude, for
the aforementioned HR practices, the perceived effectiveness might also have an influence on
the fulfilment of the basic human needs, however they are less related to these needs than
performance management, employee development, job design and participation and therefore
the effect might be too small to have a significant effect on intrinsic motivation.
Based on the previously mentioned AMO-theory and expected, corresponding
relationships, it was predicted that intrinsic motivation would mediate the positive relationship
between HR effectiveness ratings and task performance. First of all, as explained before, no
positive relationship was found between HR effectiveness ratings and task performance (see
hypothesis 1). Besides, the results showed that no mediation effect between HR effectiveness
and task performance was found for most of the HR practices (job design, participation,
communication and information sharing, work-life balance, job security and rewards).
Nevertheless, a mediation effect of intrinsic motivation was found for the negative effect of the
effectiveness ratings of two of the practices on task performance, namely for performance
management and employee development. The relationship between HR effectiveness ratings
(for the HR practices performance management and employee development) and task
performance partly goes via intrinsic motivation. In more detail, HR effectiveness ratings
positively affect intrinsic motivation and subsequently, intrinsic motivation positively affects
task performance, in this way the relationship between HR effectiveness ratings and task
performance is partly explained. However the effect between HR effectiveness ratings and task
performance is negative, whereas the relationships via intrinsic motivation are positive, as
intrinsic motivation only party mediates the relationship between HR effectiveness ratings and
task performance, there are other factors through which this relationship becomes negative (e.g.
work intensification which increases the level of stress (Ramsay, Scholarios & Harley, 2000)).
Finally, it was hypothesized that LMX would strengthen the positive relationship
between HR effectiveness ratings and intrinsic motivation. This assumption was based on the
fulfilment of the three basic human needs belonging to the self-determination theory (Ryan &
Deci, 2000). As explained before, by fulfilling these needs, intrinsic motivation is enhanced
24


(Ryan & Deci, 2000). It was explained that effective HR practices can help to fulfil the three
based human needs. Besides, a good leader-member relationship enhances the fulfilment of
these needs and therefore the level of intrinsic motivation will raise even more when a good
LMX relationship exists next to effective HR practices. The results of this research show there
is only one significant moderation effect of LMX, for the HR practice job security (solely on
low levels of LMX). Except for this specific effect, for all HR practices no significant
moderation effect of LMX was found. This is not in line with the expectations and previous
research which suggests that a relationship of good quality between a manager and an employee
and the employees’ experience of the quality of the HR practices can stimulate each other
(Sanders et al., 2010). However it does show a positive, direct relationship between LMX and
intrinsic motivation. Therefore, the conclusion might be that a good relationship between
manager and subordinate, does enhance the level of intrinsic motivation, but this does not
happen through the relationship between HR effectiveness ratings and intrinsic motivation, but
as a direct effect. As explained before, effective HR practices and a good leader-member
relationship both enhance the fulfilment of the three basic human needs, which enhances the
level of intrinsic motivation (Graves & Luciano, 2012, Ryan & Deci, 2000). However, because
no moderation effect was found, these factors probably both influence the fulfilment of these
needs independently of each other.

Limitations and suggestions for future research
Although this study shows some interesting findings, there are also some limitations
that should be mentioned. First of all, the coding of the effectiveness scores can be seen as a
limitation of this study. As explained before, respondents first got a question whether or not
they use a certain HR practice at their work. Based on their answer they received a follow-up
question. When a respondent did use a certain practice, they received a question whether they
thought the practice was effective for improving their performance. While they received a
question whether they thought the HR practices would have been effective to enhance their
performance, when they indicated they had not used the practice. Both groups had to score the
level of effectiveness on a 1-5 scale. Subsequently, no distinction has been made between the
two groups while coding the scores. Yet, a score of 5 for the group who did use the practice
means they perceived the practice as effective, whereas a score of 5 for the group who did not
use the practice only indicates they expect the practice to be effective. This means that while
measuring the perceived level of effectiveness, actually two different variables have been
measured. The question is whether employees who did not use a certain practice, are properly
25


Tài liệu bạn tìm kiếm đã sẵn sàng tải về

Tải bản đầy đủ ngay

×