User Adoption of eHRM - An Empirical

Investigation of Individual Adoption Factors

Using Technology Acceptance Model

Suryanarayan Iyer1(&), Ashis K. Pani2, and L. Gurunathan3

1
XLRI – Xavier School of Management, Jamshedpur, India
surya1968@yahoo.com
2
Information Systems, XLRI - Xavier School of Management,
Jamshedpur, India
3
HRM Area, XLRI – Xavier School of Management, Jamshedpur, India

Abstract. Organizations would reap the intended benefits of Electronic Human

Resource Management (eHRM) implementations through its sustained usage
and adoption by individuals. This study is centered on the view that actual usage
behavior is critical to studying eHRM adoption and needs to be measured in the
context of the intended eHRM outcomes; operational, relational and transfor-
mational. Using a 10-item scale to measure eHRM usage behavior in a research
framework grounded in Technology Acceptance Model (TAM2), this study
investigates the factors influencing eHRM adoption in terms of “intention to
use” and “actual usage behavior”. Results indicate support for most TAM2
hypotheses. The study also enriches our understanding of organizational context
factors; scope of implementation influencing Image-Usefulness relationship and
post implementation stage influencing Ease of use-Intention to use relationship.

Keywords: eHRM adoption
eHRM usage behavior
TAM2

1 Introduction

Electronic Human Resource Management (eHRM), a critical element of digital trans-

formation of the Human Resource (HR) function has been a topic of considerable
interest in recent times. eHRM is the e-enablement of the HR function that is posited to
help organizations achieve cost reduction and efficiency improvement of the HR
function, increased service responsiveness towards employees and improving HR’s
strategic orientation (Lepak and Snell 1998; Lengnick-Hall and Moritz 2003; Ruel
et al. 2004). Extant literature classifies these outcomes into 3 eHRM types; Operational
eHRM (cost reduction), Relational eHRM (increased service responsiveness) and
Transformational eHRM (improving HR’s strategic orientation).
As the world sprints towards organizations digitally connecting their people and
adapting to newer ways of working (remote logins, working from home, virtual
meetings, online content co-creation, etc.), the paradigm for organizations shifts
towards how easy-to-use, intuitive and engaging these digital HR technologies are for
their employees. Consequently, eHRM adoption takes a front seat with both researchers

© IFIP International Federation for Information Processing 2020

Published by Springer Nature Switzerland AG 2020
S. K. Sharma et al. (Eds.): TDIT 2020, IFIP AICT 617, pp. 231–248, 2020.
https://doi.org/10.1007/978-3-030-64849-7_21
232 S. Iyer et al.

and practitioners alike, because it is in the holistic adoption of these technologies lies
the true benefits for organizations As we trace back the last 3 decades, it is important to
recognize those points of inflexion that have shaped the innovations in the eHRM space
and therefore, the optimism, promise and expectations from eHRM in driving an
organization’s HR transformation journey (Iyer 2019). One, the rapid proliferation of
the Internet and its capacity to integrate businesses; internally and externally. This e-
enablement of the HR function has immense potential to help drive cost competi-
tiveness and increase service responsiveness towards employees (Ulrich 1997; Martin
and Reddington 2010). Two, globalization of the world economy wherein market place
imperatives drive organizations to focus on people and their talent – knowledge, skill,
abilities and other attributes as key competitive differentiators (Ulrich 2001). As Lepak
and Snell (1998) say, “Firms compete less on products and markets and more on
competencies, relationships and new ideas”. Therefore, HR innovation in terms of
adapting to new competencies, newer ways of connecting people and newer ways of
working through adoption of digital technologies derive prominence while continu-
ously focusing on people productivity, engagement and their development. Three, the
emergence of collaborative (wikis, blogs, discussion forums – linkedin, facebook,
twitter, etc.) and workplace (Zoom, Slack, Workplace, etc.) technologies. Collaborative
technologies have enabled organizations to socialize, provide avenues for employee
“voice” and access to engagement platforms to share, co-create and collaborate their
learnings (Martin et al. 2009). Workplace technologies on the other hand provide
employee with the necessary productivity tools to connect remotely, work from home
and seamlessly conduct business meetings, attend conference calls/webinars, knowl-
edge sharing/brainstorming sessions and various other day-to-day work related activ-
ities. Four, the advent of Artificial Intelligence (AI) and Machine Learning
(ML) technologies, its increasing applications in HR services and people decision
making. These AI/ML based technologies also known as BOTS or digital assistants
help address employee queries, drive HR process efficiencies by automating repetitive
tasks and are optimistically posited to provide cognitive innovations using Natural
Language Processing like persona based contextual responses to employee queries
(Iyer 2019).

2 eHRM Adoption

Literature suggests that, the performance gains and strategic advantages of techno-
logical innovation are often obstructed by user willingness to accept and use the
systems (Young 1984; Bowen 1986). eHRM is no exception to this challenge and it has
been posited that “user satisfaction” and “actual system usage” provide a more com-
plete picture of the eHRM system’s success than if either measure were applied in
isolation (Haines and Petit 1997). In extant literature, eHRM adoption studies are in 2
categories viz. organizational adoption factors and individual adoption factors. There
are a lot of studies on organizational factors influencing eHRM adoption. On the other
hand, studies on individual eHRM adoption are few. Some significant studies on
individual eHRM adoption have employed Technology Adoption Model (Davis 1989).
Marler, Liang and Dulebohn (2006) studied the effect of “extent of training” and
 User Adoption of eHRM - An Empirical Investigation 233

“quality of training” on outcome variables such as “perceived ease of use”, “perceived

usefulness”, “employee resources” and “intention to use” new software. They con-
cluded that employee resources mediated the relationship between the extent of training
and intention to use new technology. Raaij and Schepers (2008) tested the acceptance
of virtual learning environments in China using the extended TAM 2 model. Marler
et al. (2009) added a few additional constructs like perceived organization support
(POS), managerial pressure to the standard TAM constructs; perceived usefulness,
perceived ease of use and attitude towards using the system.
A key point to note in these individual adoption studies is that, the outcome
variable in most cases is “Intention to use” and not “Actual Usage”. Even though a few
studies have incorporated “Actual Use” as the outcome variable, they have measured it
using a single item “On average, how much time do you spend on the system every day.
___hours and ___ minutes”. In the words of Marler et al. (2009), “including an
objective measure of user behavior would enhance the validity of the results and allow
us to directly examine the relationship between perceived resources and system use”.
eHRM is a business application with intended operational, relational and transforma-
tional goals and outcomes at the individual HR process level (e.g. recruiting, employee
and manager self-service, employee portal, leave management, performance manage-
ment, competency management, learning management, career & succession manage-
ment). There would be differing levels of individual motivation and employees’ attitude
toward using the eHRM system depending on the type of eHRM outcome. Therefore,
actual usage of eHRM needs to be measured at the individual submodule level (or HR
process level) corresponding to the respective intended eHRM outcomes. The primary
purpose of this study therefore, is to identify and explain the factors influencing user
adoption of eHRM in terms of user acceptance (intention to use) and actual usage
(Actual Behavior) at granular eHRM outcome level.
This study makes two important contributions; first, we have attempted to measure
eHRM usage behavior in greater depth which is synergistic with the eHRM goals of
cost reduction and productivity (operational eHRM), increasing service responsiveness
(relational eHRM) and improving HR’s strategic orientation (transformational eHRM).
eHRM being a business application solution with multiple intended outcomes (Oper-
ational, Relational, Transformational), there would be differing levels of individual
motivation to use the eHRM system. It is therefore, important to measure eHRM usage
behavior for each type of eHRM outcome rather than measuring it at an overall level.
We have operationalized the eHRM usage behavior construct with 10 items covering
specific employee facing eHRM areas like employee records, employee portal,
employee help desk, social collaboration and networking, goal setting and performance
management, skill and competency management, career planning and training and
development. Second, the research identifies and reinforces that “organizational con-
text” should be considered when studying the factors influencing eHRM adoption
(Marler et al. 2009). This study identifies two such factors relating to organizational
context, “scope of implementation” and “stage of implementation” that is suggested to
be taken up in future research.
234 S. Iyer et al.

3 Proposed Hypothesis

We have adopted one of the most established Technology Acceptance Model (TAM2)
for our study (Venkatesh and Davis 2000). Before testing the eHRM adoption to
additional constructs in TAM 2, it is important to consider the original TAM for testing
eHRM adoption. The original TAM (Davis 1989) has been empirically tested in
multiple settings and applications and found to be relevant thereby establishing the
robustness of its theoretical foundation (Adams et al. 1992; Legris et al. 2003). We
further rely on studies relevant to eHRM field viz., eLearning system adoption (Selim
2003; Ong et al. 2004; Raiij and Schepers 2008) which have been empirically tested for
the constructs in the original TAM viz., Perceived Usefulness (USE) and Perceived
Ease of Use (EASE) which have a significant influence on an individual’s intention to
use (INT) a technology. According to TAM, an individual’s behavioral intention to use
a system is determined by two beliefs; “perceived usefulness” defined as the extent to
which a person believes that using a system will enhance his or her job performance
and “perceived ease of use” defined as the extent to which a person believes that the
system usage be free of effort. TAM also theorizes that ‘perceived usefulness” is
influenced by “perceived ease of use” because all things being equal, the easier the
system to use, and the more useful it will be (Venkatesh and Davis 2000). For example,
Individuals may resist updating their training and certification details if on one hand
they find that the “learning management system” is not “easy to use” (involving too
many clicks, too much mandatory information to be provided, inflexible). On the other
hand, even if the system were made “easy to use”, their “intention to use” and “actual
usage” will be negatively impacted if individuals perceive that their training
history/certification status is not considered during potential assessments, career
planning or creating development plans. In line with TAM, we hypothesize;

H1: Perceived Usefulness of the eHRM system (USE) will have a significant positive
effect on Intention to Use the eHRM system (INT).
H2: Perceived Ease of Use of the eHRM system (EASE) will have a significant
positive effect on Perceived Usefulness of the eHRM system (USE).
H3: Perceived Ease of Use of the eHRM system (EASE) will have a significant
positive effect on Intention to Use the eHRM system (INT).
H4: Intention to Use the eHRM system (INT) will have a significant positive effect
on actual eHRM usage behavior (USAGE).

Social Influence Processes. TAM2 discusses two basic constructs with reference to
social processes influencing behavioral intention viz., Subjective Norm (SN) and
Image.
Subjective Norm. In Theory of Reasoned Action (Fishbein and Ajzen 1975) and
Theory of Planned Behavior (Ajzen 1991), the theoretical underpinning for under-
standing the social influence on behavioral intention, Subjective Norm has been defined
as “a person’s perception that most people who are important to them think that they
should or should not perform a behavior in question. In other words, as stated by
 User Adoption of eHRM - An Empirical Investigation 235

Venkatesh and Davis (2000), “the rationale for a direct effect of subjective norm on
behavioral intention is that people may choose to perform a behavior even if they are
not themselves favorable towards the behavior or its consequences, if they believe that
one or more important referents think that they should and they are sufficiently
motivated to comply with the referents”. We therefore, hypothesize:

H5: Subjective Norm (SN) will have a positive direct effect on perceived usefulness
of the eHRM system (USE)
H6: Subjective Norm (SN) will have a positive direct effect on intention to use the
eHRM system (INT)

Image and Social Influence. “In a typical work environment with a high degree of
interdependence with other social actors in carrying out ones duties, increased status
within the group is a basis of power and influence via processes such as social
exchange, coalition formation and resource allocation” (Blau 1964; Pfeffer 1981,
1982). “By performing behaviors that are consistent with group norms, an individual
achieves membership and the associated social support. This results in increased power
and influence that provides a general basis for greater productivity” (Pfeffer 1982). As a
result, an individual may perceive that using the eHRM system will lead to perfor-
mance improvements indirectly due to image enhancement over and above the per-
formance benefits attributable to the system itself. For example, individuals who are a
part of specialist communities (compensation and rewards, innovation, research &
development, etc.) would perceive themselves as having a better image than others for
being a part of an elite group. The member individuals in addition would also perceive
benefits at the prospect of collaborative learning from the elite community thereby;
forming positive perceptions related to image and improved job performance (per-
ceived usefulness). We therefore, hypothesize that:

H7: Subjective Norm (SN) will have a positive effect on image perception towards
the eHRM system (IMAGE)
H8: Image will have a positive effect on perceived usefulness of the eHRM system
(USE).

Cognitive Instrumental Processes. TAM 2 theorizes four cognitive instrumental

determinants of perceived usefulness viz., job relevance (JR), output quality (OQ),
result demonstrability (RD) and perceived ease of use. These cognitive instrumental
constructs have been theorized by Venkatesh and Davis (2000) after carefully studying
multiple theoretical foundations viz. Work Motivation theory (Vroom 1964), Action
theory from social psychology (Fishbein and Ajzen 1975) and Task-contingent deci-
sion making from behavioral decision theory (Beach and Mitchell 1978). Based on
their extensive study of the above theories, they posited that “people use mental rep-
resentation for assessing the match between important work goals and consequences of
performing the act of using a system as a basis for forming judgments about the use-
performance contingency. We therefore, hypothesize;
236 S. Iyer et al.

H9: Job Relevance (JR) will have a positive effect on the perceived usefulness of
the eHRM system (USE).
H10: Output Quality (OQ) will have a positive effect on perceived usefulness of the
eHRM system (USE).
H11: Result Demonstrability (RD) will have a positive effect on perceived usefulness
of the eHRM system (USE)

4 Research Methods

4.1 Measures – TAM 2

For the present study, all the nine constructs of the TAM2 (Venkatesh and Davis 2000)
viz., Subjective Norm, Image, Job Relevance, Output Quality, Result Demonstrability,
Perceived Usefulness, Perceived Ease of Use, Intention to Use and Actual Usage
Behavior were considered as a part of the eHRM adoption model to be tested and were
measured using 26 items as proposed in the original model by Venkatesh and Davis
(2000). The TAM 2 measures used in the study are shown in Appendix 1.

4.2 Measures –eHRM Usage Behaviour

The final outcome construct of TAM 2 viz., usage behaviour was specifically devel-
oped for the current research. Though Venkatesh and Davis (2000) have indicated that
they measured “usage behaviour” by asking a single question, “On average, how much
time do you spend on the system everyday?-___hours and ___ minutes”, it was felt
during the design phase for the current research that a mere duration of use does not
capture the entire domain of eHRM actual usage behaviour. eHRM is a business
application with intended operational, relational and transformational goals and out-
comes at the individual HR process level (e.g. recruiting, employee and manager self-
service, employee portal, leave management, performance management, competency
management, learning management, career & succession management). There would
be differing levels of individual motivation and employees’ attitude toward using the
eHRM system depending on the type of eHRM outcome. We therefore, argue that
actual usage of eHRM needs to be measured at the individual submodule level (or HR
process level) corresponding to the respective intended eHRM outcomes.
Accordingly, 10 items were created for the 3 attributes of eHRM usage behavior
viz., operational (2 items), relational (4 items) and transformational (4 items). These 10
items representing eHRM usage is provided in Table 4. The combinations of items
borrowed from the original TAM and the 10 items created for “eHRM usage behavior”
were subject to scale validity check viz., face validity and construct validity. We also
subjected the scale items to other forms of robust validation checks such as discrimi-
nant, convergent and predictive validity in order to confirm construct validity (Hardesty
and Bearden 2004).
 User Adoption of eHRM - An Empirical Investigation 237

4.3 Data Collection and Sampling

Primary data was collected through structured undisguised questionnaires administered
to the respondents. The questionnaires were administered through individual face to
face meetings or through e-mail as per the convenience of the respondents at home or
in office. The 26 items from the TAM 2 scale were measured on a 5-point Likert scale
(5 - Strongly Agree to 1 - Strongly Disagree) while the 10 items pertaining to eHRM
usage behaviour were measured on a 4 - point scale (0 - Don’t use, 1 - Low usage, 2 -
Medium usage and 3 - High usage). A total of 249 correct responses were received and
considered for further analysis. It was found from the initial review of the responses
that most of the organizations had implemented the following eHRM processes viz.,
HR automation of employee records, leave, payroll, time accounting and performance
management. Therefore, in order to measure eHRM usage behavior correctly, we
selected only the response of three items (u1, u2 and u8 – refer Table 4) of eHRM
usage behavior corresponding to the above implementation scope for inclusion in the
structural model and further analysis.

4.4 Response Bias

The existence of response bias for the present research was assessed by performing a t-
test on the responses collected from 2 modes of survey viz. personal administration and
email. The t-test results indicate that there is no significant difference between the
responses obtained by these 2 modes of administration indicating that the responses
obtained are not susceptible to the “response bias” owing to different data collection
methods adopted except for 2 constructs viz., OQ and Image.

4.5 Analysis
We have applied Structural Equation Modeling (SEM), a second-generation analytical
technique (Iyer and Israel 2012) as the analytical tool of analysis and maximum
likelihood estimation method for examining the proposed hypothesis. Taking support
from the studies by various researchers who have used similar analytical tool (e.g.,
Anderson and Gerbing 1988), we chose a two-step analysis method. In the first step, we
conduct a Confirmatory Factor Analysis (CFA) of the overall measurement model
based on the correlation coefficient matrix of each factor of eHRM adoption in TAM2
and usage behaviour developed exclusively for the current study. This helps in con-
firming the fitness of the 9 constructs in terms of composite reliability, convergent and
discriminant validity. In the second step, the structural model is further examined based
on the covariance matrix and the hypotheses of the study tested for the entire eHRM
adoption/usage behaviour model.
238 S. Iyer et al.

5 Results
5.1 Exploratory Factor Analysis and Reliability
Before proceeding with assessment of model of the present study, it was decided to test
the univariate normality pattern of the study variables. Accordingly, we computed
skewness and kurtosis for each of the variables of the TAM 2 model. All the variables
have passed the stipulated criterion of ±2 value for skewness and kurtosis (Hair et al.
2010). Having assessed that the data collected do not indicate any sign of non-
normality, we proceeded with performing an Exploratory Factor Analysis (EFA). The
factor analysis (varimax rotated) results of all the 26 items for a 9-factor solution
extracted 9 factors with the respective items clearly loading onto their corresponding
factors. The KMO measure of sampling adequacy was 0.825 and the Bartlet’s test of
sphericity (v2 value of 4765.5, df = 406) was significant at 0.001 level. All the 9
factors extracted resulted in a total variance of 80.327. Upon confirming the existence
of clear dimensions of TAM2 components in the EFA study, we moved on to assess the
consistency of the data pertaining to each of the 9 components extracted separately.
The Cronbach a reliability coefficients were greater than the acceptable limit of 0.7 for
each of the TAM2 components (dimensions) as prescribed by Nunnally (1978).

5.2 Common Method Variance

Common Method Variance was not found to be a significant threat in the present
research because the results of the EFA of all the study variables brought out 9 factors
with 80% total variance explained with the first factor explaining 27% of the total
variance (unrotated) and 12.14% of total variance (varimax rotated) fulfilling Harman’s
single factor solution criteria.

5.3 TAM 2 Measurement Model for EHRM Adoption

A first order Confirmatory Factor Analysis (CFA) was conducted to confirm the fitness
of the measurement model. The measurement model test results shown in Table 1
indicate that the factor loadings of most of the items were above 0.7 (except four items
where the factor loading was marginally less at 0.65 and 0.68). These items were
retained because they were very critical to the study and the judges who validated the
scales confirmed their relevance and importance for the study. One usage behavior
item, “I use the eHRM system for my goal/key result area (KRA) setting process and
entering my self-evaluation, achievements and training needs” had a factor loading of
0.59. This was deliberated upon by the researchers and primary analysis showed that
this item is critical to the study since majority of the organizations who were covered in
the study had implemented performance management process in their organizations
and it was felt important to capture the actual usage behavior for this vital HR process.
Further, removal of the item was also not significantly improving the model fit.
Therefore, we retained this item in the final analysis. The critical ratio (CR) values of
all the items reached the significant level (p < .001). We further analyzed the com-
posite reliability for each of the constructs and found them all to be greater than 0.7
 User Adoption of eHRM - An Empirical Investigation 239

Table 1. Constructs & Indicators of measurement items of eHRM adoption – First Order
Measurement Item Factor loading Standard error Error variance CR value Construct Alpha* AVE
USE 0.93 0.76
(a1) 0.91 0.10 0.82 14.84
(a2) 0.93 0.09 0.86 15.19
(a3) 0.90 0.10 0.80 14.68
(a4) 0.74 0.05 0.55
EASE 0.86 0.61
(a5) 0.73 0.07 0.54 11.56
(a6) 0.82 0.07 0.67 12.95
(a7) 0.79 0.07 0.62 12.52
(a8) 0.78 0.09 0.61
SN 0.92 0.85
(a9) 0.88 0.06 0.77 15.15
(a10) 0.96 0.07 0.93
IMAGE 0.87 0.70
(a14) 0.68 0.07 0.47 12.24
(a15) 0.93 0.06 0.86 17.81
(a16) 0.87 0.05 0.76
JR 0.88 0.79
(a17) 0.91 0.08 0.82 15.29
(a18) 0.87 0.06 0.75
OQ 0.69 0.53
(a19) 0.80 0.12 0.63 8.77
(a20) 0.65 0.11 0.43
RD 0.88 0.66
(a21) 0.78 0.05 0.61 16.01
(a22) 0.83 0.05 0.69 17.93
(a23) 0.92 0.11 0.85
(a24R) 0.68 0.06 0.46 12.73
INT 0.89 0.81
(a25) 0.88 0.77
(a26) 0.92 0.08 0.84 14.51
USAGE 0.72 0.47
(u1) 0.77 0.13 0.59
(u2) 0.68 0.13 0.46 7.79
(u8) 0.59 0.11 0.35 7.28
Fit Index 623.39 263.00
Chi-Square (df)
CFI 0.914
RMSEA 0.074
(*) Composite Reliability
240 S. Iyer et al.

except for one factor, “Output Quality” which was marginally lower at 0.69. This
implies that all the components exhibit a good internal consistency as the composite
reliability coefficients are greater than 0.6 as recommended by Bagozzi and Yi (1990).
We calculated the Average Variance Extracted (AVE) estimates for each of the factors
and all of them are nearly equal to or greater than 0.5 (Hair et al. 2010) thereby
confirming the convergent validity of the factors in the model. The AVE of each
construct was further compared with the squared correlations of all other construct
combinations (Table 2). The AVE of all the constructs was found to be greater than the
squared correlations of the other construct combinations revealing the independent
nature of each of the TAM2 constructs. Overall, the above results provide adequate
evidence of convergent validity, discriminant validity and composite reliability of the
model adopted for the study. The other key model fit statistics and values of the CFA,
CFI = 0.914 and RMSEA = 0.074 indicate an acceptable level of model fit.

5.4 Testing TAM 2 Structural Model for EHRM Adoption – I

The results of the Structural Model for testing eHRM adoption are shown in Table 3.
The TAM 2 model for eHRM adoption was tested using Structural Equation Modeling
(SEM). In this study, we have simultaneously tested all the nine constructs of the TAM
2 in our structural model. The results are shown in Table 3. Looking at the results
therein, we find that the overall fit using CFI criteria was 0.828, RMSEA = 0.10,
v2/df = 3.493, indicating the poor fit of the model. A perusal of the structure coeffi-
cients and their corresponding significance in Table 3 indicated that paths OQ ! USE,
Image ! USE and Ease ! INT were not significant. This being the case, a scrutiny of
factor loadings (lambda coefficients) of the corresponding items on their respective
factors indicated that only item a19, The quality of the output I get from the eHRM
system is high” belonging to OQ had a non-significant and poor loading of 0.07 and
warranted removal from the overall model. Hence, it was decided to discard the same
and re-run the model. Although we felt the need to retain atleast one item related to
OQ, modification indices indicated that v2 value would be reduced by 21.838 if the
item a20 of OQ is merged with RD. A discussion with the experts also indicated that
clubbing of item a20 of OQ was meaningful as the content of a20 was similar to the
contents of RD. Further discussion with experts also revealed that the perceived
closeness of meaning by the respondents can be attributed to the words “output” (for
items in OQ) and “results (for items in RD) and probably this is the reason why
modification indices of SEM indicated the need for linking item a20 with RD. Further,
this study pertains to user adoption of eHRM systems and OQ discusses that “given a
choice set of multiple relevant systems, an individual would be inclined to choose a
system that delivers highest quality”. In the context of eHRM systems, individuals
seldom have a choice of multiple eHRM systems to base a comparative preference with
regard to OQ. Hence, there is a good possibility that the respondents perceived the
items with reference to increasing their job performance rather than the overall quality
of the system itself. We therefore, decided to remove OQ and its item a19 from the
model and club item a20 with RD for the next SEM run.
 User Adoption of eHRM - An Empirical Investigation 241

Table 2. Mean, Standard Deviation, Squared Correlations and AVE for the model constructs
compared with squared correlations of other constructs
Mean SD USE EASE SN IMAGE JR OQ RD INT USAGE
USE 4.12 0.704 0.757
EASE 4.05 0.644 0.338 0.608
SN 3.55 0.860 0.151 0.089 0.846
IMAGE 2.37 1.028 0.034 0.010 0.225 0.695
JR 3.90 0.848 0.286 0.124 0.095 0.078 0.788
OQ 3.62 0.784 0.229 0.287 0.062 0.001 0.348 0.531
RD 3.73 0.720 0.104 0.090 0.040 0.110 0.362 0.371 0.656
INT 3.93 0.708 0.113 0.061 0.172 0.264 0.194 0.028 0.206 0.807
USAGE 2.72 0.664 0.000 0.001 0.003 0.099 0.030 0.000 0.126 0.032 0.466
Values in bold shown diagonally in the table indicate the Average Variance Extracted.
Values shown below the diagonal are squared correlations.

5.5 Testing TAM 2 Structural Model for EHRM Adoption - II

Results of TAM2 after removal of OQ and addition of a20 with RD in the model
yielded model fit indices of CFI = 0.857, RMSEA = 0.093, v2 = 835.93 (df = 265),
v2/df = 3.154. All this indicated that removal of OQ as a separate component of TAM2
has significantly improved the model fit (D v2 = 162.92, D df = 23). However, a
perusal of structural path coefficients between Image ! USE, Ease ! INT and RD
USE were found insignificant. A scrutiny of modification indices revealed that the v2
statistic can be reduced by 69.97 by linking RD to JR. Discussion on the relevance of
connecting the path from RD ! JR led us to the reasoning that since eHRM is a
business application and its adoption and usage by individuals could be purely for the
purpose of performing in their current and potential jobs, users would consider as
important and project only those results that are relevant to their current and future jobs.
This reasoning corroborates with Venkatesh and Davis (2000) view that: “If a system
produces effective job-relevant results desired by the user, but does so in an obscure
fashion, users of the system are unlikely to understand how useful such a system really
is (p. 192).” This led us to the reasoning that the results from the eHRM system would
benefit individuals in either reducing their administrative burden or lead to higher
productivity and flexibility due to presence of automated workflow driven self-service.
Further, individuals would benefit from competency building, greater knowledge
exchange and thereby advance themselves towards career growth in their respective
organizations. It was hence, reasoned that Results Demonstrability has an influence on
“perceived usefulness” for individuals, if realized in the context of their performance in
their current jobs and building competencies towards advancement into future jobs
(“job relevance”). Hence, it was decided to test the model of TAM2 with RD connected
to JR as it was convincing as stated above.
242 S. Iyer et al.

Table 3. Structural Model for testing eHRM Adoption using TAM2

Model Standardized Coefficient v2 df v2/df CFI RMSEA
Model – 1 1005.855 288 3.493 0.828 0.1
SN ! Image 0.476 ***
Image ! Use −0.083
SN ! Use 0.187**
SN ! INT 0.331 ***
JR ! Use 0.343***
OQ ! Use −0.027
RD ! Use 0.111**
Ease ! Use 0.477***
Use ! INT 0.198 **
INT ! Usage 0.16 **
Ease ! INT 0.031
Model - 2 Standardized Coefficient v2 df v2/df CFI RMSEA
835.93 265 3.154 0.857 0.093
SN ! Image 0.477***
Image ! Use −0.031
SN ! Use 0.197**
SN ! INT 0.334***
JR ! Use 0.397***
RD ! Use −0.25
Ease ! Use 0.466***
Use ! INT 0.192**
INT ! Usage 0.161**
Ease ! INT 0.037
Model - 3 Standardized Coefficient v2 df v2/df CFI RMSEA
644.133 252 2.556 0.902 0.079
SN ! Image 0.483***
Image ! Use −0.046
SN ! Use 0.220***
SN ! INT 0.332***
JR ! Use 0.422***
RD ! Use −0.049
Ease ! Use 0.471***
Use ! INT 0.182**
INT ! Usage 0.161**
Ease ! INT 0.052
RD ! JR 0.601***
*** p < .001 ** p < .01 * p < .05
 User Adoption of eHRM - An Empirical Investigation 243

5.6 Testing TAM 2 Structural Model for EHRM Adoption - III

In the third SEM run, as anticipated, we found significant impact of RD on JR with a
structure coefficient of 0.6, while the relationship between RD to Perceived Usefulness
was found to be insignificant. The model fit statistics also improved considerably viz.,
CFI = 0.902, RMSEA = 0.079, v2 = 644.13 (df = 252), v2/df = 2.556 (Table 3). The
modification indices also did not indicate any major reduction in v2 for further mod-
ifications to the model and hence is validated in the current study. Nonetheless, the
relationship between TAM2 components viz., Image and Perceived Usefulness was
again found insignificant. We analyzed this also closely and felt that since the sample
population comes from organizations that have implemented the basic eHRM system
components viz., employee records, employee lifecycle events, payroll, benefits, time
and attendance and performance management the sample population could not perceive
these transactional process as being a source of status or prestige in the organization.
Probably, these results would have been different if the sample population included
organizations that had implemented Integrated Talent Management Systems (especially
Career and Succession Planning, Competency Management, Learning and Knowledge
Management) or Web 2.0 collaboration based Social Networking initiatives in the
organization. Usage of these systems are either related to an individuals’ career growth
(in the case of Talent Management Systems) or related to contemporary technologies
(Web 2.0 collaboration and social networking) both of which could potentially drive
“internalization” and “individualization” amongst individuals. The potential of these
systems, especially, Web 2.0 and social networking in the context of social exchange
and social interdependence is perceived as a source of power, influence and enhanced
status by individuals. We therefore, argue that image will positively impact perceived
usefulness for individuals from organizations that have implemented talent manage-
ment systems and social networking processes.
Another TAM 2 relationship that was found non-significant was “perceived ease of
use” and “intention to use”. This finding was consistent with the results from recent
research (Szajna 1996; Chau and Hu 2002; Selim 2003; Wu and Wang 2005). A very
significant reasoning by Raiij and Schepers (2008) explains this phenomenon of non-
significance fundamentally. They state that TAM 2 (Venkatesh and Davis 2000) did
not find any direct post-implementation effects of “perceived ease of use” on “intention
to use”. On the contrary, the direct effect posited in TAM 2 was in the context of pre-
implementation stage. Therefore, in accordance with Raiij and Schepers (2008), we
also reason that “as users gain experience with a new eHRM system, perceived ease of
use becomes less profound since instrumentality overshadows concerns about the
eHRM system’s ease of use (Adams et al. 1992; Straub et al. 1997). All the other
hypotheses of the study were supported and found significant. The overall model fit
statistics viz., CFI = 0.902, RMSEA = 0.079, v2 = 644.13 (df = 252), v2/df = 2.556
(Table 3) indicates an acceptable level of model fit (Hair et al. 2010).
244 S. Iyer et al.

Table 4. Items for measuring Actual eHRM Usage Behavior

# eHRM process
u1 I use the eHRM system to view and update my personnel records like change of O
address, marital status, contact details, dependent/beneficiary information and tax
declarations
u2 I use the eHRM system to apply for leave, compensatory off, applying for change O
of department or location or any other job related attribute
u3 I use the eHRM system to review and apply for jobs through Internal Job R
postings and refer job postings to my friends under the company’s employee
referral scheme
u4 I use the eHRM – employee portal to check the company directory and R
organization structure and to update myself on company communication,
policies & guidelines
u5 I use the eHRM – Help desk functionality to log in requests for resolution of any R
job related issues, for e.g. Leave regularization, payroll corrections, performance
rating related, or any other personal or job related issues
u6 I participate in “in-company” eHRM Web 2.0 collaboration initiatives (e.g. R
blogging, discussion boards) by actively voicing my opinions and sharing my
knowledge and experience.
u7 I use the eHRM system to update my skills and competencies and review my T
competency improvement areas
u8 I use the eHRM system for my Goal/KRA setting process and entering my self- T
evaluation, achievements and training needs
u9 I carefully review the career paths available on the eHRM system and work with T
my manager to design development plans to help me grow up the career ladder
u10 I use the eHRM system for all my training needs e.g. Enrolling into a course, T
launching and attending the course, entering feedback and cost details, etc.
O – Operational eHRM, R – Relational eHRM, T – Transformational eHRM.

6 Discussion

Through an in-depth study of a large sample population covering individuals from

organizations across all major industries, we have understood the factors affecting
eHRM adoption through one of the most accepted Technology Acceptance Model –
TAM2 (Venkatesh and Davis 2000). Overall, the hypothesized model fit the data well
and most of the TAM2 hypotheses were supported in the study. We did not find
support for hypotheses H3, H8 and H11 though. However, a more detailed analysis led
us to enrich our understanding of the role played by organizational context factors in
explaining the relationships mentioned in H3, H8 and H11.
Hypotheses H3, the effect of eHRM’s perceived ease of use on intention to use was
not supported. On careful analysis, we realized that most of the respondents belonged
to organizations where eHRM was in the post-implementation phase. Raiij and
Schepers (2008) also found a similar phenomenon in their study of acceptance of
virtual learning environments and reasoned that TAM2 did not find any direct post-
implementation effects of “perceived ease of use” on “intention to use”. Marler et al.
(2009) also found similar results in their study of Employee Self Service (ESS) tech-
nology adoption wherein they observed a marked change in the relationship between
 User Adoption of eHRM - An Empirical Investigation 245

ease of use and attitude in the post implementation scenario and suggested that ease of
use becomes redundant with perceived usefulness, post implementation. We make an
important inference that as individuals gain experience of continuously using eHRM
systems, post implementation, instrumentality overshadows concerns about ease of use.
Therefore, unless new eHRM system alternatives offer significantly better ease of use
as compared to what the users are already comfortable with and habituated to –“in-
tention to use” will not be significantly impacted by “perceived ease of use”. Future
research should consider “stage of implementation” – an organization context while
studying the relationship between ease of use and intention to use.
Hypothesis H8, the effect of “image” on “perceived usefulness” was not supported.
On careful analysis, we realized that the sample population belonged to organizations
who had implemented only the core eHRM components like employee records, life-
cycle events, payroll, time and attendance and performance management. These
organizations had not implemented the other eHRM processes like integrated talent
management (career and succession planning, learning and development) and Web 2.0
technologies (social networking, collaboration). Though we had not included this
aspect in our main hypothesis, we probably would have got support for the “image” to
“perceived usefulness” relationship if the sample organizations had implemented
integrated talent management, social networking and web 2.0 collaboration. The
potential of these systems, especially, in the context of social exchange and social
interdependence is perceived as a source of power, influence and status by individuals.
We therefore, suggest that future research focus on the relationship between “image”
and “perceived usefulness” in the context of “implementation scope and coverage”.
Hypothesis H11, the effect of “results demonstrability” to “perceived usefulness”
was also not supported. We carefully, analyzed the results and found a link between
“results demonstrability” and “job relevance”. This led us to probe further into its
significance, though we had not hypothesized the relationship. We infer that eHRM is a
business application and individuals’ perception of its usefulness would largely depend
on its ability to demonstrate results that are relevant in the context of an individual’s
performance in their current or potential jobs. Therefore, results demonstrability in the
context of “job relevance” makes more meaning in the study of its relationship with
perceived usefulness. This probably explains our finding of significant support for the
“results demonstrability” linkage with “job relevance”.

7 Contributions and Implications

This study makes two significant contributions. First, this research reinforces the
organizational context angle to be considered when studying the factors influencing
eHRM adoption (Marler et al. 2009). In this study, we have identified two such factors
of organizational context – “scope of implementation” (effect of Image on perceived
usefulness) and “stage of implementation” (effect of ease of use on intention to use).
Second, we have called for measuring “eHRM usage behavior” in a detailed manner
rather than measuring it at an overall level. We argue that eHRM is a business
application with differing outcomes at operational, relational and transformational
246 S. Iyer et al.

stages of eHRM evolution. Organizations who implement eHRM are at various stages
of maturity in terms of eHRM adoption and hence, measuring eHRM usage behavior at
an overall level does not seem logical. There seems to be a clear gap in literature in the
way eHRM usage is measured. As a first step, therefore, we have attempted to create a
scale to measure eHRM usage behavior with a 10 point scale spanning the three eHRM
types stated in literature – operational, relational and transformational). Our 10 point
eHRM usage scale focuses on individuals (in effect we can say Employee Self-Service
- ESS) and covers employee records, leave, mobility, internal job postings (Operational
eHRM), employee portal, helpdesk, Web 2.0 collaboration, social networks (Relational
eHRM) and competencies, goal setting, performance reviews, career planning, learning
and development (Transformational eHRM). Though we created this 10-point scale
and administered it all the respondents, we got responses only for 3 items which
covered employee records, leave and performance management. This was because
most of the organizations from where we got individual responses had not implemented
the other eHRM processes – a clear case of organizational context of “implementation
scope”. Therefore, though we could not test the entire 10-point scale, we believe we
have initiated a serious discussion on measuring eHRM usage behavior more objec-
tively and in greater detail. Future research should identify a more comprehensive and
innovative sampling procedure to cover those organizations which have implemented
all the eHRM types to test eHRM adoption and in particular, eHRM usage behavior.

Appendix 1. TAM 2 Measures Used in the Study

Item Description
USE
(a1) Using the eHRM system improves my performance in my job
(a2) Using the eHRM system in my job increases my productivity
(a3) Using the eHRM system enhances my effectiveness in my job
(a4) I find the eHRM system to useful in my job
EASE
(a5) My interaction with the eHRM system is clear and understandable
(a6) Interacting with the eHRM system does not require a lot of my mental effort
(a7) I find the eHRM system easy to use
(a8) I find it easy to get the eHRM system to do what I want it to do
SN
(a9) People who influence my behavior think that I should use the eHRM system
(a10) People who are important to me think that I should use the eHRM system
IMAGE
(a14) People in my organization who use the eHRM system have more prestige than
those who do not
(a15) People in my organization who use the eHRM system have a high profile
(a16) Using the eHRM system is a status symbol in my organization
(continued)
 User Adoption of eHRM - An Empirical Investigation 247

(continued)
Item Description
JR
(a17) In my job, usage of the eHRM system is important
(a18) In my job, usage of the eHRM system is relevant
OQ
(a19) The quality of the output I get from the eHRM system is high
(a20) I have no problem with the quality of the eHRM system’s output
RD
(a21) I have no difficulty telling others about the results of using the eHRM system
(a22) I believe I could communicate to others the consequences of using the eHRM
system
(a23) The results of using the eHRM system are apparent to me
(a24R) I would have difficulty explaining why using the eHRM system may or may not
be beneficial
INT
(a25) Assuming I have access to the eHRM system, I intend to use it
(a26) Given that I have access to the eHRM system, I predict that I would use it

References
Adams, D.A., Nelson, R.R., Todd, P.A.: Perceived usefulness, ease of use and usage of
information technology. MIS Q. 16, 227–248 (1992)
Anderson, J.C., Gerbing, D.W.: Structured Equation Modeling in practice: a review and
recommended two-step approach. Psychol. Bull. 103(3), 411–423 (1988)
Azjen, I.: The theory of planned behavior. Organ. Behav. Hum. Decis. Process. 50, 179–211
(1991)
Bagozzi, R.P., Yi, Y.: Assessing method variance in multitrait–multimethod matrices: the case of
self-reported affect and perceptions at work. J. Appl. Psychol. 75, 547–560 (1990)
Beach, L.R., Mitchell, T.R.: A contingency model for selection of decision strategies. Acad.
Manag. Rev. 3(3), 439–449 (1978)
Blau, P.M.: Exchange and Power in Social Life. Wiley, New York (1964)
Bowen, W.: The puny payoff from office computers. Fortune, pp. 20–24 (1986)
Chau, P.Y.K., Hu, P.J.H.: Investigating healthcare professionals’ decisions to accept
telemedicine technology: an empirical test of competing theories. Inf. Manag. 39, 297–311
(2002)
Davis, F.D.: Perceived usefulness, perceived ease of use and user acceptance of technology.
MIS Q. 13, 319–339 (1989)
Fishbein, M., Ajzen, I.: Belief, Attitude, Intention and Behavior: An Introduction to Theory and
Research. Addison-Wesley, Reading (1975)
Hair, F.H., Black, W.C., Babin, B.J., Anderson, R.E.: Multivariate Data Analysis, 7th edn.
Prentice-Hall, Upper Saddle River (2010)
Haines, V.Y., Petit, A.: Conditions for successful human resource information systems. Hum.
Resour. Manag. 36(2), 261–275 (1997)
248 S. Iyer et al.

Hardesty, D.M., Bearden, W.O.: The use of expert judges in scale development. Implications for
improving face validity of measure of unobservable constructs. J. Bus. Res. 57, 98–107
(2004)
Iyer, S.: Understanding the eHRM promise and adoption imperatives. NHRD Netw. J., 1–10
(2019)
Iyer, S., Israel, D.: Structural equation modeling to test the impact of organization communi-
cation satisfaction on employee engagement. South Asian J. Manag. 19(1), 51–81 (2012)
Legris, P., Ingham, J., Collerette, P.: Why do people use information technology? A critical
review of the technology acceptance model. Inf. Manag. 40, 191–204 (2003)
Lepak, D.P., Snell, S.A.: Virtual HR: strategic human resource management in the 21st century.
Hum. Resour. Manag. Rev. 8(3), 215–234 (1998)
Lengnick-Hall, M.L., Moritz, S.: The impact of e-HR on the human resource management
function. J. Labor Res. 24(3), 365–379 (2003)
Marler, J.H., Fisher, S.L., Ke, W.: Employee self-service technology acceptance: a comparison of
pre-implementation and post-implementation relationships. Pers. Psychol. 62(2), 327–358
(2009)
Marler, J.H., Liang, X., Dulebohn, J.H.: Training and effective employee information technology
use. J. Manag. 32(5), 721–743 (2006)
Martin, G., Reddington, M., Kneafsey, M.B.: Web 2.0 and Human Resources: Groundswell or
Hype? Research Report. Chartered Institute of Personnel and Development, London (2009)
Martin, G., Reddington, M.: Theorizing the links between e-HR and strategic HRM: a model,
case illustration and reflections. Int. J. Hum. Resour. Manag. 21, 1553–1574 (2010)
Nunnally, J.C.: Psychometric Theory, 2nd edn. McGraw-Hill, New York (1978)
Ong, C.S., Lai, J.-Y., Wang, Y.-S.: Factors affecting engineers’ acceptance of asynchronous e-
Learning systems in high-tech companies. Inf. Manag. 41, 795–804 (2004)
Pfeffer, J.: Power in Organizations. Ballinger, Cambridge, MA (1981)
Pfeffer, J.: Organizations and Organizations Theory. Pitman, Marshfield (1982)
Raiij, E.M.V., Schepers, J.J.L.: The acceptance and use of a virtual learning environment in
China. Comput. Educ. 50, 838–852 (2008)
Ruel, H.J.M., Bondarouk, T.V., Looise, J.K.: eHRM-innovation or irritation: an explorative
empirical study in five large companies on web-based HRM. Manag. Rev. 15(3), 364–381
(2004)
Szajna, B.: Empirical evaluation of the revised technology acceptance model. Manag. Sci. 42,
85–92 (1996)
Selim, H.M.: An empirical investigation of student acceptance of course websites. Comput.
Educ. 40, 343–360 (2003)
Straub, D.W., Keil, M., Brenner, W.H.: Testing the technology acceptance model across cultures:
a three country study. Inf. Manag. 33, 1–11 (1997)
Ulrich, D.: The changing nature of human resources: a model for multiple roles. In: Ulrich, D.
(ed.) Human Resource Champions. Harvard Business School Press, Boston, MA (1997)
Ulrich, D.: From eBusiness to e-HR. Hum. Resour. Plann. J. 5, 90–97 (2001). International
Association for Human Resources Information Management (IHRIM)
Venkatesh, V., Davis, F.D.: A theoretical extension of the technology acceptance model. Four
Longitudinal field studies. Manage. Sci. 45(2), 186–204 (2000)
Vroom, V.H.: Work and Motivation. Wiley, New York (1964)
Wu, J.H., Wang, S.C.: What drives mobile commerce?: an empirical investigation of the revised
technology acceptance model. Inf. Manag. 42, 719–729 (2005)
Young, T.R.: The lonely micro. Datamation 30, 100–114 (1984)