Content uploaded by Madison Ngafeeson
Author content
All content in this area was uploaded by Madison Ngafeeson on Jul 19, 2014
Content may be subject to copyright.
Int. J. Electronic Finance, Vol. x, No. x, xxxx 1
Copyright © 200x Inderscience Enterprises Ltd.
An exploratory study of user resistance in
healthcare IT
Madison N. Ngafeeson*
Walker L. Cisler College of Business,
Northern Michigan University,
1401 Presque Isle,
Marquette, MI 49855, USA
E-mail: mngafees@nmu.edu
*Corresponding author
Vishal Midha
Department of Accounting and Business Information System,
Illinois State University, Campus Box 5500, Normal IL 61790, USA
E-mail: vmidha@illinoisstate.edu
Abstract: The US healthcare system is clearly experiencing a major transition.
By 2015, the healthcare sector is expected to have migrated from a paper record
system to a completely electronic health record (EHR) system. The adoption
and use of these systems are expected to increase legibility, reduce costs, limit
medical errors and improve the overall quality of healthcare. Hence, the US
government is investing $70 billion over a ten-year period to facilitate the
transition to an electronic system. However, early reports show that physicians
and nurses among other health professionals continue to resist the full use of
the system. This paper uses the theory of cognitive dissonance to investigate
user resistance in HIT. It builds on a Lapointe and Rivard (2005) framework to
offer an explanation as to why people resist HITs. A conceptual model is
developed and tested. The findings, implications, and limitations of the study
are also discussed.
Keywords: HIT; health information technology; EHRs; electronic health
records; electronic medical records; IT user resistance; change management;
healthcare technology.
Reference to this paper should be made as follows: Ngafeeson, M.N. and
Midha, V. (xxxx) ‘An exploratory study of user resistance in healthcare IT’,
Int. J. Electronic Finance, Vol. x, No. x, pp.xxx–xxx.
Biographical notes: Madison N. Ngafeeson is an Assistant Professor of
Computer Information Systems at the Walker Cisler College of Business,
Northern Michigan University in Marquette, Michigan. He earned his PhD in
Business Administration from the University of Texas-Pan American,
in Edinburg, Texas. His general research interests are in the areas of the
adoption, implementation, diffusion and use of information systems in
organisations; with a special focus in health information systems management.
His works have been published in such outlets as the International Journal of
Electronic Healthcare, the International Journal of Electronic Government
Research, and in conferences such as the European Conference on Information
Systems and the Decision Sciences Institute.
2 M.N. Ngafeeson and V. Midha
Vishal Midha is an Associate Professor in the Department of Accounting
and Business Information Systems at Illinois State University. He received
his PhD in MIS from the University of North Carolina at Greensboro.
His current other research interests include open source software development,
information privacy and security concerns, and knowledge management.
He has published in Decision Support Systems, Communications of AIS,
International Journal of Electronic Commerce, Electronic Markets, Journal of
CIS, and many national and international conferences, including the
International Conference of Information Systems, and Americas Conference on
Information Systems. Presently, he also serves as an Associate Editor in the
International Journal of Information Security and Privacy.
This paper is a revised and expanded version of a paper entitled ‘Making
sense of user resistance in health care organizational change management’
presented at Decision Sciences Institute-Southwest Region, Albuquerque, NM,
March 12–16, 2013.
1 Introduction
“Why do people resist evidence that challenges the validity of long-held
beliefs? And why do they persist in maladaptive behaviour even when
persuasive information or personal experience recommends change?” (Sherman
and Cohen, 2002)
The migration from paper to EHRs in the USA has already begun: signalling a wave of
change in the health sector that has also met resistance. The Meaningful Use mandate of
the Department of Health and Human Services means that all healthcare organisations
must adopt and use EHR ‘meaningfully’. It has now been over two years since this
initiative was launched, and early reports show that the benefits are palpable, but
changing the way business has been done in the healthcare system is still a challenge for
physicians and other practitioners (Buntin et al., 2011).
In a recent study of 20 information technology (IT) and IT-related journals over the
past 25 years, Lapointe and Rivard (2005) found that 43 articles identified resistance
as a key implementation issue. Although these works acknowledged the importance of
resistance, most did not delve into the nature of resistance. Furthermore, Lapointe and
Rivard (2005) point out that only as little as four of these articles attempt to address the
how and why of resistance. Additionally, while many theoretical models have been
proposed so far (see Joshi, 1991; Piderit, 2000; Martinko et al., 1996; Markus, 1983),
there is still a dearth of literature on the subject. There is almost a lack of empirically
tested frameworks. Notable exceptions include Bhattacherjee and Hikmet (2007) and
Kim and Kankanhalli (2009). There is however no doubt that the understanding
of how and why resistance takes place is both important to information system (IS)
researchers, and organisational scientists and managers. From a more practical
standpoint, a shared understanding of the resistance phenomenon among researchers and
managers should help to mitigate resistance to- and increase acceptance of information
technology (Martinko et al., 1996). It is worth mentioning that in this paper, we use the
term ‘IT’ in a narrow sense to refer to just technology, while the term ‘IS’ is used in a
broader sense to refer to the interaction of people, the organisation structure and
technology.
An exploratory study of user resistance in healthcare IT 3
Furthermore, earlier research on IS introduction in the workplace has also blamed
system implementation failures on factors that go beyond a mere worker-technology
relationship (Martinko et al., 1996). When workers either avoid, walk-around or overtly
resist the use of a system, implementation goals are undermined and failure of
implementation is possible. As Martinko et al. (1996) have also pointed out, huge losses
in financial investments are often associated with these implementation failures. It is the
view of many researchers that understanding and managing resistance to information
technology in a larger context of organisational change is very critical if IS must support
organisations in achieving desired outcomes (Kim and Kankanhalli, 2009; Coetsee,
1999).
This research focuses on understanding why healthcare IT users resist information
technology. We examine the user resistance phenomenon through the lens of the
cognitive dissonance theory literature. A conceptual model, exploring the antecedents of
user resistance is developed based on a generic resistance model earlier proposed by
Lapointe and Rivard (2005). The proposed model is further tested empirically. The role
and relationship of the key antecedents, namely: perceived loss of control, perceived
dissatisfaction, technology self-efficacy and social enabling effect are especially
highlighted, in the context of this sector-wide organisational change.
In the following section, we review the IT resistance literature: making a special
emphasis on the theory of cognitive dissonance and the IT resistance framework
developed by Lapointe and Rivard (2005). Next, we develop and test the research model
using preliminary data. The findings are then presented while the implications,
limitations and conclusions of the research are discussed in the end.
2 Literature review
Information technology resistance has been defined as behaviours intended to prevent the
implementation or use of a system or to prevent system designers from achieving their
objectives (Markus, 1983). In the context of this research, user resistance is defined as
healthcare IT users’ behaviours intended to oppose and prevent the use of health IT
systems to achieve desired organisational healthcare outcomes following the
implementation of a new health IT system. According to these formulations of the
concept of resistance, three important points are noteworthy:
• resistance is first and foremost a behaviour
• it can be overt or covert
• its effects can hinder system outcomes.
An often common concept that is underlies resistance is the quest of whether it is
negative or positive. Some have viewed resistance as negative – especially from the
perspective of management – however, resistance has also been thought of as a positive
feedback mechanism in which, the user can communicate with the implementer (Waddell
and Sohal, 1998; Piderit, 2000).
Very few researchers have conceptualised resistance; fewer still have attempted to
test their models empirically (Kim and Kankanhalli, 2009; Bhattacherjee and Hikmet,
2007; Lapointe and Rivard, 2005; Joshi, 1991). Generally, four major theories have been
proposed to explain resistance. Joshi (1991) proposed a model based on the equity theory
4 M.N. Ngafeeson and V. Midha
called the equity-implementation model. This model attempted to explain resistance to
change. In it, Joshi (1991) proposed that individuals attempt to evaluate most changes:
both favourable and unfavourable in order to make decisions. Favourable changes
like increased wages and promotions are easily and quickly accepted while changes
considered as unfavourable are resisted. The equity theory is therefore presented as the
evaluative framework through which individuals evaluate options and make their choices.
Markus (1983) also proposed a set of three theories drawn from Kling (1980) in which he
elaborated that resistance theories fall into one of three perspectives namely: factors
completely internal to them, factors inherent in the technology or the interaction of
people and system factors. Of these three theories, Markus (1983) built her interaction
model on the third. Martinko et al. (1996) also proposed the attributional model in which
they posit that an individual’s attributions influence the individual’s expectation with
regard to performance outcomes that in turn drive behavioural behaviours towards
technology. Lastly, Kim and Kankanhalli (2009) use the status quo bias theory to explain
why people may prefer to maintain their current status or situation over change. Most of
these resistance theories have been based on social psychological variables which link the
realms of cognition, affect and behaviour (Piderit, 2000).
A summary of key research in IT resistance is presented in Table 1. Evidently, case
studies and literature reviews have dominated the major approaches to the investigation
of resistance. While both conceptual and empirical frameworks have been proposed,
these frameworks have largely been untested. The current research is similar to
extant studies in that it: builds on Lapointe and Rivard (2005) framework, considers
resistance as an attitudinal outcome, examines the individual resistance to IT, and adopts
a post-implementation perspective. Nevertheless, it departs from previous studies in that
it uses the theory of cognitive dissonance which, so far, has not been leveraged in
resistance literature. Additionally, the study not only proposes a theory-based model of
resistance, but actually tests it too. Previous research has done little or no testing, as can
be seen on Table 1. Lastly, this current study shows that perceived threats, which hitherto
has been considered as a single construct, is in fact two related, but completely unique
phenomena.
Table 1 A summary of key research and models in IT resistance
Theoretical perspective/view
Type of study/technology
type Type of model
Markus (1983) Interaction theory
Power and politics dynamics
Neither good nor bad
Case study/Group
analysis
Financial information
systems
Theoretical
Untested
Hirschheim and
Newman (1988)
Resistance as aggression,
projection, avoidance
Case study/Group
analysis
Insurance policy
processing systems
Conceptual
An exploratory study of user resistance in healthcare IT 5
Table 1 A summary of key research and models in IT resistance (continued)
Theoretical perspective/view
Type of study/technology
type Type of model
Joshi (1991) Equity theory
Resistance as a result of gain or
loss of equity status
Case study/Individual
level
Clinical laboratory
system; banking system;
fourth generation
programming language
Theoretical model
Untested
Martinko et al.
(1996)
Attribution theory
Learned helplessness
Literature review/
Individual level
Conceptual
Untested
Lapointe and
Rivard (2005)
Combination of extant theories
Process model
Case study/multi-level –
group level
Electronic medical
records
Theoretical
Untested
Bhattacherjee and
Hikmet (2007)
Dual factor model
Technology acceptance model
Empirical study
Post-implementation of a
clinical system
Theoretical
Empirical test
Kim and
Kankanhalli
(2009)
Integration of technology
acceptance and status quo bias
perspective
Empirical study
Pre-implementation of an
IT enterprise system
Theoretical
Empirical test
2.1 The Lapointe and Rivard (2005) framework
Lapointe and Rivard (2005) proposed a multilevel longitudinal approach to explain the
evolutionary nature of IT user resistance. This process model, based on prior literature,
suggested that people resist IT primarily because of certain threats that they perceive.
According to the Lapointe and Rivard (2005), users of an information technology
constantly make projections about the consequences of the use of a given technology.
If the expected conditions following its use are threatening, they will resist. These threats
could be due to perceived inequities, loss of power, stress or fear (Joshi, 1991; Markus,
1983; Marakas and Hornik, 1996). Their model further suggested that perceived threats
were preceded by certain initial conditions. Initial conditions, according to Lapointe and
Rivard (2005), include habits, routines, social values, and workplace interrelationships
(e.g. distribution of power) within an organisation.
2.2 The theory of cognitive dissonance
The theory of cognitive dissonance was first proposed by Festinger (1957) and has been
used for over 50 years to explain change behaviours. The original theory holds that
“when an individual holds two or more elements of knowledge that are relevant to each
other but inconsistent with one other, a state of discomfort is created” (Harmon-Jones
et al., 2010). The resulting discomforting state is called ‘dissonance’. Because dissonance
originates from the conflicting “things a person knows about himself, about his
behaviour, and about surroundings” (Festinger 1957, p.9), the concept is collectively
known as cognitive dissonance.
6 M.N. Ngafeeson and V. Midha
Generally, there exists some consistency between what a person knows and what he
does. For example, if an individual believes that getting an education is a good idea, they
are likely to encourage their children to get an education. This example captures the idea
of ‘consistency’ in belief and action; and is generally a norm in life. However, there are
exceptions to this rule. An individual may know that stealing is wrong and that it might
constitute an offense against the law; and yet, be involved in a theft. According to the
dissonance theory, this inconsistency or ‘lack-of-fit’ of cognitions motivates the
individual to be involved in a psychological effort to reduce the inconsistency between
the cognitions. Hence, if an individual who holds the belief that stealing is wrong steals,
he would likely be in a state of dissonance. Once this happens, the theory predicts that the
individual is likely to do one of two things. He may either justify his action (“I only stole
because I was hungry”) or could change his initial belief that stealing is wrong (“Stealing
is not that bad, as long as it’s the only option available”), to reduce dissonance. On the
other hand, if his initial beliefs are strong enough, he may decide to hold on
to his primary cognition and discontinue stealing – the dissonant behaviour; thereby
reducing dissonance. Researchers often measure dissonance reduction as attitude change
(Harmon-Jones et al., 2010). Hence, attitude change in response to a dissonant condition
is expected to be in the direction of the cognition that is most resistant to change.
Questions as to why people experience dissonance and why they are motivated to
reduce it have spun several streams of research in social-psychology and has given birth
to several mini-theories in the area of cognitive dissonance. Among these, the most
popular are: the self-consistency theory (Aronson, 1969); self-affirmation theory (Steele,
1988); self-standards model (Stone and Cooper, 2003); aversive consequences
perspective (Cooper and Fazio, 1984) and the action-based model (Harmon-Jones et al.,
2010). The difference in these theories rests essentially in the attribution of the role
of ‘self’ in the cognitive dissonance process. As has been argued by Harmon-Jones
and Harmon-Jones (2002), Festinger (1957) theory stopped short of explaining why
individuals find cognitive inconsistency aversive.
Each approach, therefore, makes different predictions regarding the role of cognitions
in the dissonance process by assuming that different types of information are regularly
brought to the mind when people assess their behaviour and then attempt to cope with
their discomfort (Stone and Cooper, 2001).
Having discussed the different perspectives of the dissonance theory, it would be
necessary to summarise the fundamental claims of the theory. This theory simply holds
that:
• discrepancies may exist between cognitions, leading to dissonance
• the existence of dissonance leads to motivations within the individual to reduce or
even avoid this increase in dissonance
• the manifestation of these pressures include changes in behaviour, cognition, and a
cautious exposure to new information and opinions.
3 Model development and hypotheses
In this study, resistance and its antecedents are considered in the light of the Lapointe and
Rivard (2005) process framework. As shown in Figure 1, each of the model constructs,
An exploratory study of user resistance in healthcare IT 7
derived from the cognitive dissonance theory (CDT), was mapped to the Lapointe and
Rivard (2005) initial conditions – perceived threats – resistance framework. Summarily,
the initial conditions constructs considered in this model were technology self-efficacy
and social enabling effect; the perceived threats constructs were perceived loss of control
and perceived dissatisfaction; while the user resistance construct was maintained as such.
Hence, this model posits that IT user resistance is predicted by perceived loss of control
and perceived dissatisfaction which are each further predicted by technology self-efficacy
and social enabling effect respectively. The research model is illustrated in Figure 1.
Figure 1 Research model
3.1 IT user resistance
Human attitudes have often been thought of to be best conceptualised in terms of
cognitions, emotions and behavioural intensions (Ajzen, 1984). It is therefore not strange
that a behaviour like resistance – to information technology, in our case – would be
perceived in like manner (Piderit, 2000). Resistance in information systems literature has
often been characterised as an adverse reaction that is detrimental to the organisation
(see Kim and Kankanhalli, 2009). This view sees resistance as a negative reaction that
needs to be dealt with. Very often, it pitches employer vs. employee, or administration vs.
the staff. In other words, the employees are perceived as those resisting the changes from
the employer or the administration. This view seems to be popular given the fact that the
introduction of technology is generally considered as an enabler of positive outcomes in
the workplace. Hence, resisting a technology is generally perceived to be negative in
nature given the fact that it prevents the ‘positive’ outcomes intended by the
administration or employer.
8 M.N. Ngafeeson and V. Midha
It is argued that in the healthcare sector, HIT in general and EHRs in particular will
lead to lowered costs, increased legibility, reduced errors, and improved healthcare
quality delivery (Jha et al., 2009; Blumenthal and Tavenner, 2010). Actions and
behaviours deterrent to the achievement of this purpose may therefore be considered as
resistant behaviours. Consistent with the cognitive dissonance theory described above,
user resistance in this study is defined as follows. When an individual’s intention or
action to reduce dissonance or inconsistency is to rationalise or support his present
state of cognition or belief: such that a ‘new knowledge’ is considered as dissonant or
inconsistent to with the individual’s present cognition or beliefs, the consequent
behaviour can be described as resistance. Simply put, resistance is an implicit or explicit
intension that results in a behaviour that opposes change towards a particular ‘new’
attitude or behaviour. Consequently, this research maintains that user resistance of an
information system is a covert or overt intension that opposes change towards the use of
an information system. This definition has three implications:
• resistance is first and foremost a behaviour
• it can be overt or covert
• its goal is to hinder system outcomes.
In summary, user is the subject; the object is the information technology (EHR); and the
attitude is resistance.
In the sections following, we develop the research model and the hypothesised
relationships based on the Festinger (1957) cognitive dissonance theory while using the
Lapointe and Rivard (2005) generic model as a basis.
3.2 Perceived loss of control
Festinger (1957) questioned: “what… are the circumstances that make it difficult for the
person to change his actions” (p.25)? In response to his own question, Festinger (1957)
provides insight to the resistance concept in many ways. First, he suggests that people
resist change because, it is ‘painful’, or may ‘involve loss’. Furthermore, he asserts
“the magnitude of this resistance to change will be determined by the extent of pain or
loss which must be endured” (p.25). Perceived loss of control refers to an individual’s
perception that carrying out a particular behaviour will cost them their control over the
situation.
Shine (2002) had argued that the elimination of written clinical notes by 2010 is a
reachable objective; but cautioned that health professionals would need to move from a
20th-century paradigm to a 21st-century one. This paradigm, noted Shine (2002),
constituted among other things a shift from physician autonomy to teamwork and
systems, solo practice to group practice, continuous learning to continuous improvement,
and infallibility to multi-disciplinary problem-solving and from knowledge to change.
This change of paradigm has far reaching consequences on the health professional’s
work environment. It means that their autonomy, power and workflow will be impacted.
The move from ‘knowledge’ to ‘change’ means that the professional’s world will never
be a calm, predictable and stable one. Consequently, it leads to a sense of loss of control
in power, autonomy and the flow of work. Mrayyan (2004) also noted that autonomy
An exploratory study of user resistance in healthcare IT 9
plays an important part in nurses’ job satisfaction and retention. He argues that nurses are
often dissatisfied with the lack of autonomy and constantly demand for greater autonomy
in decision-making (Mrayyan, 2004). Furthermore, Warren et al. (1998) commented
that: “[P]hysicians have lost control over who become their patients, the terms and
content of their work, the equipment and facilities needed for their work, and the amount
and rate of remuneration for their labour stemming from federal control and managed
healthcare”.
It is evident that some of the changes in the healthcare system are likely to generate
resistance due to the loss of control in autonomy and power of the health professional.
This loss of control is further exacerbated by the constraints placed on medical
professionals by governmental control and management of healthcare (Warren et al.,
1998). Since these changes do not originate from healthcare professionals, but rather
from policy makers, physicians and other professionals are likely to resist such changes.
Hence, it is hypothesised:
Hypothesis 1: Perceived loss of control due to EHR introduction will positively affect
user resistance to the system.
3.3 Perceived dissatisfaction
Perceived dissatisfaction is defined as an individual’s belief that carrying out a particular
behaviour will not be a gratifying thing. Festinger (1957) states: “[t]he resistance to
change would be a function of the satisfaction obtained from the present behaviour”
(p.26). Poon et al. (2006) observed that the use of non-interoperable HIT systems was
likely to negatively impact workflow and productivity.
Furthermore, Poon et al. (2006) also asserted that the income of healthcare providers
was directly tied to their productivity but not to their quality. Consequently,
dissatisfaction with productivity and workflows is likely to cause resistance to change
especially in an era of decreasing reimbursement (Poon et al., 2006). These productivity
and workflow challenges are then likely to contribute to the clinician’s dissatisfaction
with the system due to its threat on productivity and workflows.
Summarily, dissatisfaction due to the introduction of an information system in the
healthcare workplace can result in the alteration of reward systems, and impact
productivity and workflow. This means that health professionals whose productivity,
workflow and rewards are affected by electronic records introduction are likely to be
dissatisfied and hence, resist the technology. The more dissatisfied an individual is, vis-à-
vis a system, the more they are likely to resist it. It is therefore hypothesised:
Hypothesis 2: Perceived dissatisfaction with EHRs will positively affect user
resistance to the systems.
The perceived loss of control due to the introduction of an information system in the
workplace can also be a source of dissatisfaction in itself. For instance, Mrayyan (2004)
has stated that autonomy plays an important part in nurses’ job satisfaction and retention.
Hence, nurses are often dissatisfied with the lack of authority and demand better
working conditions and greater autonomy in decision-making. If this autonomy of
practice in the profession is not granted, dissatisfaction ensues. In this study’s context,
10 M.N. Ngafeeson and V. Midha
when IT is introduced into the healthcare workplace, the disruption of routines may
dictate and impose new ways of doing things, making professionals to feel unsafe and
insecure. This perceived loss of control is likely to cause dissatisfaction with the
introduced system. Research findings suggest that autonomy (lack of control) is the
strongest predictor of physician and nurses’ job satisfaction (Mrayyan, 2004; Warren
et al., 1998). In fact, in one particular study, nursing autonomy was positively correlated
with better perceptions of the quality of care delivered and higher levels of job
satisfaction (Rafferty et al., 2001). Evidently, the sense of loss of control in autonomy
and/or workflows due to systems introduction are also likely to increase the
professionals’ dissatisfaction with the given information system. It is therefore
hypothesised:
Hypothesis 3: Perceived loss of control due to EHR introduction will positively affect
perceived dissatisfaction with the system.
3.4 Self-efficacy
The theory of cognitive dissonance also argues that the belief about ‘self’ is key to the
reducing dissonance and thus, changing behaviours. Compeau and Higgins (1995)
defined computer self-efficacy as “individuals’ beliefs about their abilities to competently
use computers”. Applied to technology, technological self-efficacy therefore refers to an
individual’s beliefs about their ability to competently use a technology. As has been
shown in previous research, and based on Bandura (1986), this belief is positively
associated with expectations of future use of technology (Compeau and Higgins 1995).
One reason for resistance of new technology lies in the unpredictable outcome on the use
of the technology, sometimes based on the lack of exposure to similar technology in the
past.
Simply put, individuals, who have used computers in the past, are likely to be more
favourable to use them in the future compared to non-users. Bandura’s (1977) theory of
self-efficacy hypothesises that “expectations of personal efficacy determine whether
coping behaviour will be initiated, how much effort will be expended, and how long it
will be sustained in the face of obstacles and aversive experiences”. Furthermore,
Bandura (1977) adds: “[p]ersistence in activities that are subjectively threatening but in
fact relatively safe produces, through experiences of mastery, further enhancement of
self-efficacy and corresponding reductions in defensive behaviour” (p.191). In other
words, past experiences of mastery of a particular behaviour are likely to reduce
‘defensive behaviour’ – or a sense of loss of control, in this case – even though the new
activity may be subjectively threatening. In the light of health IT, previous exposure to
similar technologies is likely to lessen the sense of loss of control due to the introduction
of a given technology. Hence, the greater the technological self-efficacy, the more less
likely an individual will feel threatened or have a sense of a loss of control over the new
technology. It is therefore hypothesised:
Hypothesis 4: Technology self-efficacy will negatively impact perceived loss of
control over the EHR system.
An exploratory study of user resistance in healthcare IT 11
3.5 Social enabling effect
Festinger (1957) also postulated that “when it is established by agreement with other
people, the resistance to change would be determined by the difficulty of finding persons
to support the new cognition” (p.27). In other words, when individuals do not find
support for ‘change ideas’ with significant or referent others around them, they tend to
feel a social nudge that encourages them to perform a particular contrary behaviour.
For example, if the prevailing belief within a healthcare facility or practice is not in
favour of a particular change, and individual within that social unit is likely to resist the
change behaviour in question. He or she does so, based on the perception that the
behaviour in question is somehow consistent with those of significant others around him
or her.
Social psychology research is conclusive on the potential influence of significant
others on an individual’s attitudes and behaviours. The subjective norm, described as
“a person’s perception that most people who are important to him think he should or
should not perform the behaviour in question” (Fishbein and Ajzen, 1975).
Social enabling effect is the perception by an individual that his or her behaviour is
consistent with significant others’ beliefs about the given behaviour. For example, if an
individual’s perception within a healthcare practice community is that significant others
are dissatisfied with a particular change behaviour, the individual is equally likely to be
dissatisfied. Hence, it is hypothesised:
Hypothesis 5: Social enabling effect will positively impact perceived dissatisfaction
with EHRs such that, if the perception about referent others towards the system is
dissatisfactory, then the behaviour of the subject in question will also be that of
dissatisfaction with the system.
4 Methodology
4.1 Sampling procedure
This exploratory study was conducted in the College of Health Sciences and Human
Services of a large southwestern university. This college houses among other academic
departments nursing, physician assistants, and rehabilitation departments. It also offers
degrees, both undergraduate and graduate professional degrees, in these disciplines. The
physician and nursing graduate programs specifically train professionals who use EHRs
in their routine work.
The sample was drawn from a class of final year students in the physician assistant
studies and a graduate nursing practitioner course. The subjects were handed the surveys
following a study recruitment notification and encouraged to participate voluntarily and
anonymously in the study. They were informed that their participation would help the
scientific community in understanding healthcare professionals’ behaviours towards HIT
usage. Of the 80 surveys that were distributed, 64 were found usable (80% response rate),
and the rest were incomplete with missing data. Given our sampling population, and
response rate, we did not find any non-response bias issues. Of the total surveyed,
43 were physician assistants and 21 were nurse practitioners. The general sample had an
average daily computer usage of 6 hrs. It must be noted that this survey excluded all
individuals who had no experience with both paper and electronic records.
12 M.N. Ngafeeson and V. Midha
A summary of the descriptive statistics of the sample is presented in Table 2a and
Table 2b. As can be seen from Tables 1a and 1b, 40 of the total number of respondents
were females, while the remaining 24 were males. Particularly over nearly 60% of all
respondents had an EHR experience of less than a year, while a third had an experience
exceeding two years. Additionally, a majority of respondents (nearly 60%) had a practice
experience of less than two years while about a quarter (about 27%) had practiced for
more than five years.
Table 2a Sample EHR experience
EHR experience Count Percentage
Below one year 38 59.38
Male 16 25.00
Female 22 34.38
1–2 years 6 9.38
Male 4 6.25
Female 2 3.13
More than two years 20 31.25
Male 4 6.25
Female 16 25.00
Table 2b Sample years of practice
Years of practice Count Percentage
Less than two years 38 59.38
Male 18 28.13
Female 20 31.25
2–5 years 9 14.06
Male 3 4.69
Female 6 9.38
More than five years 17 26.56
Male 3 4.69
Female 14 21.88
4.2 Measures
The five constructs used in the study were either adapted from previous studies.
Each of these constructs was measured on a five-point Likert scale depicting
the respondent’s level of agreement with a particular item. The range was from
strongly disagree to strongly agree. Items on User resistance were adapted from
Bhattacherjee and Hikmet (2007) who also investigated physician resistance of HIT.
Items of perceived loss of control, perceived dissatisfaction and social enabling effect
were adapted from Ngafeeson (2013). Lastly, items on technology self-efficacy were
adapted from Compeau and Higgins (1995). A summary of the original instrument:
An exploratory study of user resistance in healthcare IT 13
including construct definitions and sample items used are included in the summarisation
on Table 3.
Table 3 Construct definition and derivation
Construct Definition Sample Item Source
User resistance
(UR)
A covert or overt intention
that opposes change towards
the use of an information
system
I don’t want the EHR system
to change the way I order
patient tests
Bhatacherjee
and Hikmet
(2007)
Perceived loss of
control (PLC)
An individual’s belief that
carrying out a particular
behaviour will cost them
their control over the
situation
The EHR system makes me
lose my sense of autonomy as
a professional
Ngafeeson
(2013)
Perceived
dissatisfaction (PD)
An individual’s belief that
carrying out a particular
behaviour will not be a
gratifying thing
I am not satisfied with the
way the EHR system
interferes with my
professional autonomy
Ngafeeson
(2013)
Technology self-
efficacy (TSE)
An individual’s belief that
they are able to competently
use technology
I feel confident that I could
complete my job using the
EHR with productivity as a
professional if I had never
used a system like that before
Compeau and
Higgins (1995)
Social enabling
effect (SEE)
An individual’s belief that
his/her beliefs are consistent
with those of referent others
around them
My peers would agree with
me that the EHR system has
flaws that prevent usage
Ngafeeson
(2013)
5 Data analysis and results
5.1 Data analysis
To test the research hypotheses, partial least squares (PLS) was used. PLS is a
component-based algorithm for structural equation modelling (SEM), that is quite similar
to the covariance-based SEM technique but with some unique differences. Unlike the
latter, PLS does not make assumptions that observations must follow a specific
distributional pattern and that each observation is independently distributed (Chin, 2010,
p.659). This characteristic of PLS makes it suitable for exploratory studies, which have a
limited sample sizes and where claims of multivariate normality of distribution may not
be made. Nevertheless, Fornell and Bookstein (1982) have shown that PLS can generate
similar loadings and structural path values comparable to other SEM techniques without
requiring these distributional assumptions.
The minimum sample size consideration for this study was determined using two
criteria suggested by Hair et al. (2014). First, the general rule of thumb is to use a sample
size that is ten times the largest number of structural paths directed at a particular
construct in the structural model. Since the largest number of arrowheads pointing to a
latent variable in the proposed model was 2, the 10 times arrowhead rule required a
sample size of at least 20. However, like Hair et al. (2011) have noted, PLS-SEM like
14 M.N. Ngafeeson and V. Midha
every other statistical technique must also consider the background of model and data
characteristics. Specifically, power analyses have been highly recommended. Given the
characteristics of the proposed model (i.e., with a maximum of 2 arrowheads to a latent
variable); it will require a least sample size of 52 to yield a statistical power of 80% at
95% confidence level for a minimum R2 of 0.25 (see Hair et al., 2014, p.21). The sample
size of 64 satisfied both the rule-of-thumb and the more stringent power analysis
calculations; and hence, proving adequacy for use in this study.
PLS allows for a combined principal component factor analysis as well as regression
analysis. Hence, PLS is clearly superior to the traditional regression analysis as it
assesses the measurement model is assessed within structural model context (Thompson
et al., 1991). Consequently, PLS is a clearly useful tool for exploratory research (Chin,
2010, p.660) capable of handling complex models using smaller samples. SmartPLS
version 2.0 was particularly used in the analysis of this data.
After coding of sample research data, seven missing values were found for different
observations. The missing values were treated in a two-step process. First, missing data
cells were replaced with a sentinel value in each cell (in this case –99) and resaved.
Second, the data were then imported to SmartPLS and the missing values settings
corrected to reflect the sentinel value before proceeding with validation of data. In order
to estimate the model, a case-wise replacement algorithm was chosen. This process forces
PLS to not discard useful information for the non-missing values cells. The model was
then estimated using both the PLS algorithm and bootstrapping techniques. The PLS
algorithm helps us the determine path coefficients while bootstrapping enables us to
determine the significance of these paths and to finally test the proposed hypotheses.
Table 4 presents the descriptive statistics of all the constructs – their mean and standard
deviation values, as well as the number of items that were used of represent the construct.
Table 4 Descriptive statistics of constructs
Construct Number of items Mean Standard deviation
User Resistance (UR) 4 3.336 1.005
Perceived Loss of Control (PLC) 5 2.391 0.833
Perceived Dissatisfaction (PD) 3 2.406 0.909
Social Enabling Effect (SEE) 3 3.448 0.871
Self-Efficacy (SEF) 3 1.823 0.576
5.2 Model evaluation: measurement model results
Generally, first part of model evaluation is to present the measurement model results.
This portion focuses on ascertaining how accurate or reliable the measures are as well as
assessing the convergent and discriminant validities of the proposed model. The
measurement model was assessed for internal consistency by computing both the
Cronbach’s alphas and composite reliability values. Composite reliability measures
the internal consistency of a construct, but unlike Cronbach’s alpha, it does not assume
equal indicator loadings (Hair et al., 2014). Composite reliability measure is therefore a
suitable measure for use in lieu of Cronbach’s alpha in this study. Hair et al. (2014)
suggest a threshold of 0.70 in exploratory research or a range of 0.06–0.07 to be
considered acceptable (p.115). The Cronbach’s alpha values for the measurement were
An exploratory study of user resistance in healthcare IT 15
all adequate: ranging from 0.71 (for technology self-efficacy) to 0.92 (for social enabling
effect). The composite reliabilities measures also confirmed reliability given that
these measures are all greater than the recommended 0.7 threshold level (see Table 5).
The inter-construct correlations and reliabilities are also included in Table 5. It should be
noted here that these measurements represent the final values after inter-item cross-
loadings were identified and some items dropped from the analysis. According to these
results, the reliability measures were considered to be adequate given that all were greater
than the recommended 0.70 level (Nunnally, 1978). An additional check on reliability
was also done by measuring the average variance extracted (AVE) and the composite
reliability measures. The AVE serves to further support the reliability of these measures
as recommended by Fornell and Larcker (1981). AVEs are expected to be greater than
the squared inter-construct correlations to establish discriminant validity. Results reveal
that the AVEs are all above the 0.5 threshold level, meaning that more than 50% of the
variance in the indicators is accounted for; and that all AVEs were greater than the
squared inter-construct correlations, establishing discriminant validity.
Table 5 Inter-construct and reliability measures
Squared correlations among constructs
Construct Composite reliability AVE PD PLC SEE EF UR
PD 0.9225 0.7990 1
PLC 0.8602 0.6070 0.5849 1
SEE 0.9501 0.8640 0.1614 0.0754 1
SEF 0.8342 0.6277 0.0004 0.0001 0.0005 1
UR 0.8926 0.6754 0.1335 0.3207 0.0766 0.0133 1
Convergent validity was also assessed. Convergent validity refers to the extent to which
blocks of items strongly agree or ‘converge’ in their representation of the underlying
construct they were created to measure Chin (2010). It answers the question as to how
high each of the loadings is and whether they are more similar or dissimilar. Though
there is no generally accepted rule of thumb, Chin (2010) recommends that loadings be
high enough and to have about a difference in range of about 0.02. Except for the self-
efficacy item – SEF4 – which loaded highly but had a wider range of 0.58–0.86 all of the
rest of the items both loaded highly and within an acceptable narrow range. We therefore
see evidence of convergent validity from the data. Table 6 reveals the squared factor
cross-loadings for a more intuitive assessment of the convergent validity.
Table 6 Squared factor cross-loadings between constructs
PD PLC SEE SEF UR
PD1 0.8838 0.5483 0.1451 0.0009 0.1350
PD2 0.8017 0.5108 0.0715 0.0015 0.1356
PD4 0.7117 0.3398 0.1972 0.0028 0.0527
PLC1 0.3966 0.6942 0.0581 0.0001 0.2366
PLC2 0.3846 0.6631 0.0214 0.0051 0.1590
PLC4 0.3026 0.5109 0.0277 0.0055 0.1897
16 M.N. Ngafeeson and V. Midha
Table 6 Squared factor cross-loadings between constructs (continued)
PD PLC SEE SEF UR
PLC5 0.3361 0.5600 0.0895 0.0103 0.1947
SEE1 0.2270 0.1080 0.8844 0.0030 0.0620
SEE2 0.0978 0.0483 0.8586 0.0102 0.0761
SEE3 0.0686 0.0276 0.8490 0.0044 0.0637
SEF3 0.0045 0.0001 0.0107 0.5909 0.0150
SEF4 0.0003 0.0001 0.0013 0.7526 0.0137
SEF6 0.0000 0.0000 0.0130 0.5395 0.0003
UR1 0.0212 0.1915 0.0050 0.0144 0.6392
UR2 0.0795 0.1974 0.0419 0.0043 0.6876
UR3 0.1708 0.2317 0.1478 0.0583 0.6187
UR4 0.1319 0.2470 0.0654 0.0065 0.7560
5.3 Model evaluation: structural model results
The results of the structural model are summarised in Figure 2. These results show
the path coefficients, R-square values as well as the significance levels. The t-statistics
for significance levels were obtained from the bootstrapping procedure of PLS.
Bootstrapping is a non-parametric technique that does not require the normality
assumptions associated with regression models. In order to obtain reliable structural path
results and their t-values a bootstrapping procedure of 5000 samples and 64 cases was
run. Results for the structural model show that the hypotheses were only partially
supported for the proposed cognitive dissonance model of resistance. Specifically, three
hypotheses were supported while two were not. Examining the path coefficients as well
as the directionality, we can see that Hypotheses 1, 3 and 5 were supported while
Hypotheses 2 and 5 were not (P < 0.05).
Figure 2 Structural model showing path coefficients
n.s.: not significant: *significant at 0.05 level; **significant at 0.01 level.
An exploratory study of user resistance in healthcare IT 17
User resistance was predicted positively by perceived loss of control (β = 0.69;
P < 0.01), but not by perceived dissatisfaction as originally hypothesised. Perceived
dissatisfaction, on the other hand, was positively predicted by perceived loss of control
(β = 0.71; P < 0.01) and social enabling effect (β = 0.21; P < 0.05). Technology
self-efficacy’s influence on perceived loss of control turned out to be non-significant.
The total variance in user resistance that was explained by perceived loss of control and
perceived dissatisfaction was 33.2%, while the total variance in perceived loss of control
and perceived dissatisfaction were 0% and 62.5% respectively.
6 Findings, implications and limitations
6.1 Findings
The purpose of this study was to understand why user resistance to IT happens. The goal
was to develop a theory-based model that could be empirically tested and to make sense
of these exploratory findings. The study utilised the theory of cognitive dissonance and
built on the Lapointe and Rivard (2005) generic framework to propose an empirically
testable model.
This study found that user resistance originates from perceived threats that may come
from two sources, namely: perceived loss of control and perceived dissatisfaction.
Earlier research (e.g., Bhattacherjee and Hikmet, 2007; Lapointe and Rivard, 2005) had
looked at perceived threat as a singular construct. This research suggests that there are
two possible types of threats that can generate user resistance (UR), namely perceived
loss of control (PLC) and perceived dissatisfaction (PD). However, while the PLC-UR
relationship was clearly strong and positive, the PD-UR was non-significant. It is possible
that the limited sample size for the study was a contributing factor to this non-significant
relationship, given that previous studies show that dissatisfaction with technology
outcomes may cause users to resist its use (see Mrayyan, 2004).
The relationship between perceived loss of control and perceived dissatisfaction was
also strong and positive. This means that threats of loss of control can also trigger
dissatisfaction with outcomes. This finding is very important because it shows that
perceived loss of control and perceived dissatisfaction should not be considered under the
umbrella term of perceived threats because not only are there are at least two distinct
types of threats, but that one of them (perceived loss of control), could actually lead to the
other (perceived dissatisfaction).
Perceived dissatisfaction, on the other hand, was found to be influenced by
social enabling effect. As discussed earlier, when an individual belief about
the introduction of a system in the workplace is that this technology will cause significant
others to be dissatisfied with it, they are likely to be equally dissatisfied with the
technology. This finding is consistent with normative behavioural theories which suggest
that individuals’ important others can directly or indirectly influence their behaviours
(see Ajzen, 1991). In fact this model showed that up to 62.5% of perceived dissatisfaction
was jointly predicted by perceived loss of control and social enabling effect.
No support was found for the relationship between technology self-efficacy and
perceived loss of control. It is possible that the small sample size would have impacted
this relationship by providing very small variability in the sample. Additionally, it is
possible that if items were adapted to be more specific in capturing technological
18 M.N. Ngafeeson and V. Midha
self-efficacy in general, but healthcare technology self-efficacy, better results could be
yielded.
6.2 Implications
The study offers both theoretical and practical implications. Theoretically, this research
extends the body of knowledge in IT user resistance by leveraging a social psychological
theory (cognitive dissonance) to explain the concept of resistance. As Piderit (2000) has
cautioned, user resistance is a complex phenomenon that requires a multidimensional
approach examining it. This research therefore introduces a relevant body of literature
from which resistance studies can be viewed from. Second, this study introduces the
concept of two distinct but identical types of perceived threats which heretofore has only
regarded as one and the same thing. Because each threat is different, the strategy to
combat each threat will be different and would improve our knowledge perspectives on
the subject.
Practically, change managers will find this research helpful for two major reasons.
First, it attempts the answer as to why people resist technology. From the standpoint of
the two major ways that perceived threats are manifested, managers may design strategies
for combating or at least mitigating user resistance. By proactively dealing with issues of
perceived loss of control due to power imbalances or autonomy concerns resulting
from the introduction of a system, managers can organise programs or campaigns
to deal with the threats. Additionally, the items used in this study could be utilised
pre-implementation to identify potential threat areas. For instance, if a manager finds out
that employees are afraid to lose the power vested in their positions, campaigns to assure
threatened employees may be in order. By the same token, if the managers notice a
general perception of dissatisfaction with the outcomes of the systems, they may equally
work out strategies designed to target this concern. Knowles and Linn (2004) have noted
that theoretical understanding of resistance, can lead to the right application of persuasion
– its antithesis. Lastly, Vendors of EHR software will find this research useful in that,
elements of the system that conflict flagrantly with workflows and autonomy leading to
threats could be minimised so as to mitigate user resistance due to the system.
6.3 Limitations and future research
These findings and implications must, however, be interpreted within the confines of the
limitations of this study. First, the resistance model proposed was based on the theory of
cognitive dissonance. Different perspectives are also needed to study a complex concept
as user resistance. Second, the sample data was small and could present issues of
generalisability of results. Lastly, the lack of pre-validated scales for resistance and the
perceived threats constructs means that further testing would be required in the future.
Nevertheless, this research has its merits, as it can serve as a departure point for future
empirical research in user resistance: by leveraging the cognitive dissonance theory with
IT user resistance and testing it empirically. Additionally, though a small sample size is
used, it still met the minimum requirements for SEM using the PLS technique.
Future research could consider the use of more theoretical paradigms that lend an
understanding of the concept of IT user resistance for greater insights. It might be that
more theories lead to the discovery of new types of threats as Bhattacherjee and Hikmet
(2007) have noted. Furthermore, future studies might test the validity of the study’s new
An exploratory study of user resistance in healthcare IT 19
scales here-developed and by using new paradigms arrive at a testable comprehensive
model.
7 Conclusion
Understanding user resistance is crucial in the current US healthcare transition.
Why people resist the use of health technology must be sufficiently answered if the
promise of quality outcomes would be realised. This preliminary research shows that the
theory of cognitive dissonance can be a useful lens through which an understanding the
role of user resistance to IT can be gained. It also revealed that perceived loss of control,
social enabling effect and perceived dissatisfaction are important antecedents of IT user
resistance.
If change managers would be successful in managing the current healthcare
transition, they must convince healthcare professionals that the benefits of the system far
outweigh the inconveniences of change. While change is never static, the challenge of
managers of change must seek for a way to ‘normalise’ the inevitable change.
References
Ajzen, I. (1991) ‘The theory of planned behavior’, Organizational Behavior and Human Decision
Processes, Vol. 50, No. 2, pp.179–211.
Aronson, E. (1969) ‘The theory of cognitive dissonance: a current perspective’, Advances in
Experimental Social Psychology, Academic Press, New York, Vol. 4, pp.1–34.
Bandura, A. (1977) ‘Self-efficacy: toward a unifying theory of behavioral change’, Psychological
Review, Vol. 84, pp.191–215.
Bandura, A. (1986) ‘The explanatory and predictive scope of self-efficacy theory’, Journal of
Social and Clinical Psychology, Vol. 4, pp.359–373.
Bhattacherjee, A. and Hikmet, N. (2007) ‘Physicians’ resistance toward healthcare information
technology: a theoretical model and empirical test’, European Journal of Information Systems,
Vol. 16, No. 6, pp.725–737.
Blumenthal, D. and Tavenner, M. (2010) ‘The ‘meaningful use’ regulation for electronic health
records, [Perspective]’, The New England Journal of Medicine, Vol. 363, No. 6, pp.501–504.
Buntin, M.B., Burke, M.F., Hoaglin, M.C. and Blumenthal, D. (2011) ‘The benefits of health
information technology: A review of recent literature shows predominantly positive results’,
Health Affairs, Vol. 30, No. 3, pp.464–471.
Chin, W.W. (2010) ‘How to write and report PLS analyses’, in Vinzi, V.E., Chin, W.W.,
Henseler, J. and Wang, H. (Eds.): Handbook of Partial Least Squares: Concepts, Methods and
Applications, Springer-Verlag, New York, pp.655–687.
Coetsee, L. (1999) ‘From resistance to commitment’, Public Administration Quarterly, Vol. 23,
No. 2, pp.204–222.
Compeau, D.R. and Higgins, C.A. (1995) ‘Computer self-efficacy: development of a measure and
initial test’, MIS Quarterly, Vol. 19, No. 2, pp.189–211.
Cooper, J. and Fazio, R.H. (1984) ‘A new look at dissonance theory’, in Berkowitz, L. (Ed.):
Advances in Experimental Psychology, Vol. 17, pp.229–264.
Festinger, L. (1957) A Theory of Cognitive Dissonance, Stanford University Press, Stanford, CA.
Fishbein, M. and Ajzen, I. (1975) Belief, Attitude, Intention, and Behavior: An Introduction to
Theory and Research, Addison-Wesley, Reading, MA.
20 M.N. Ngafeeson and V. Midha
Fornell, C. and Bookstein, F.L. (1982) ‘Two structural equation models: LISREL and PLS applied
to consumer exit-voice theory’, Journal of Marketing Research, Vol. 19, pp.440–452.
Fornell, C. and Larcker, D.F. (1981) ‘Structural equation models with unobservable variables and
measurement error: algebra and statistics’, Journal of Marketing Research, pp.382–388.
Hair, J.F., Ringle, C.M. and Sarstedt, M. (2011) ‘PLS-SEM: Indeed a silver bullet’, Journal of
Marketing Theory and Practice, Vol. 19, pp.139–151.
Hair, J.F., Hult, G.T., Ringle, C.M. and Sarstedt, M. (2014) A Primer on Partial Least Squares
Structural Equation Modeling (PLS-SEM), Sage, London.
Harmon-Jones, E. and Harmon-Jones, C. (2002) ‘Testing the action-based model of cognitive
dissonance: the effect of action-orientation on post-decisional attitudes’, Personality and
Social Psychology Bulletin, Vol. 28, pp.711–723.
Harmon-Jones, E., Amodio, D.M. and Hamon-Jones, C.H. (2010) ‘Action-based model
of dissonance on cognitive conflict and attitude change’, in Forgas, J.P., Joel, C. and
Crano, W.D. (Eds.): The Psychology of Attitude & Change, Psychology Press, NY,
pp.163–181.
Hirschheim, R. and Newman, M. (1988) ‘Information systems and user resistance: theory and
practice’, The Computer Journal, Vol. 31, pp.1–11.
Jha, A.K., DesRoches, C.M., Campbell, E.G., Donelan, K., Rao, S.R., Ferris, T.G., Shields, A.,
Rosenbaum, S. and Blumenthal, M.D. (2010) ‘Use of electronic health records in U.S.
hospitals’, The New England Journal of Medicine, Vol. 360, No. 16, pp.1628–1638.
Joshi, K. (1991) ‘A model of users’ perspective on change: the case of information systems
technology implementation’, MIS Quarterly, Vol. 15, No. 2, pp.229–242.
Kim, H-W. and Kankanhalli, A. (2009) ‘Investigating user resistance to information systems
implementation: a status quo bias perspective’, MIS Quarterly, Vol. 33, No. 3, pp.567–582.
Kling, R. (1980) ‘Social analyses of computing: theoretical perspectives in recent empirical
research’, Computing Surveys, Vol. 12, pp.60–110.
Knowles, E.S. and Linn, J.A. (Eds.) (2004) Resistance and Persuasion, Academic Press,
New York.
Lapointe, L. and Rivard, S. (2005) ‘A multilevel model of resistance to information technology
implementation’, MIS Quarterly, Vol. 29, No. 3, pp.461–491.
Marakas, G.M. and Hornik, S. (1996) ‘Passive resistance misuse: overt support and covert
recalcitrance in IS implementation’, European Journal of Information Systems, Vol. 5, No. 3,
pp.208–220.
Markus, M.L. (1983) ‘Power, politics, and MIS implementation’, Communications of the ACM,
Vol. 26, No. 6, pp.430–444.
Martinko, M.J., Henry, J.W. and Zmud, R.W. (1996) ‘An attributional explanation of individual
resistance to the introduction of information technologies in the workplace’, Behavior and
Information Technology, Vol. 15, No. 5, pp.313–330.
Mrayyan, M.T. (2004) ‘‘Nurses’ autonomy: influence of nurse managers’ actions’, Journal of
Advanced Nursing, Vol. 45, No. 3, pp.326–336.
Ngafeeson, M.N. (2013) Understanding User Resistance to Information Technology: Toward A
Comprehensive Model in Health Information Technology, Doctoral Dissertation,
Proquest/UMI Manuscript No. 10530.
Nunnally, J.C. (1978) Psychometric Theory, 2nd ed., McGraw-Hill, New York.
Piderit, S.K. (2000) ‘Rethinking resistance and recognizing ambivalence: a multidimensional
view of attitudes toward an organizational change’, Academy of Management Review, Vol. 25,
No. 4, pp.783–794.
An exploratory study of user resistance in healthcare IT 21
Poon, E.G., Blumenfeld, B., Hamann, C., Turchin, A., Graydon-Baker, E., McCarthy, P.C., and
Broverman, C.A. (2006) ‘Design and implementation of an application and associated services
to support interdisciplinary medication reconciliation efforts at an integrated healthcare
delivery network’, Journal of the American Medical Informatics Association, Vol. 13, No. 6,
pp.581–592.
Rafferty, A.M., Ball, J. and Aiken, L.H. (2001) ‘Are teamwork and professional autonomy
compatible, and do they result in improved hospital care?’, Quality in Health Care, Vol. 10,
pp.32–37.
Sherman, D.K. and Cohen, G.L. (2002) ‘Accepting threatening information; self-affirmation and
the reduction of defensive biases’, Current Directions in Psychological Science, Vol. 11,
pp.119–123.
Shine, K.I. (2002) ‘Health care quality and how to achieve it’, Academic Medicine, Vol. 77,
pp.91–99.
Steele, C.M. (1988) ‘The psychology of self-affirmation: sustaining the integrity of the self’,
Advances in Experimental Social Psychology, Academic Press, San Diego, CA, Vol. 21,
pp.261–302.
Stone, J. and Cooper, J. (2001) ‘A self-standards model of cognitive dissonance’, Journal of
Experimental Social Psychology, Vol. 37, pp.228–243.
Stone, J. and Cooper, J. (2003) ‘The effect of self-attribute relevance on how self-esteem moderates
attitude change in dissonance processes’, Journal of Experimental Social Psychology, Vol. 72,
pp.54–65.
Thompson, R.L., Higgins, C.A. and Howell, J.M. (1991) ‘Personal computing: toward a conceptual
model of utilization’, MIS Quarterly, pp.125–143.
Waddell, D. and Sohal, A.S. (1998) ‘Resistance: a constructive tool for change management’,
Management Decision, Vol. 36, No. 8, pp.543–548.
Warren, M., Guptill, R., Weitz, R. and Stephen, K. (1998) ‘Physician satisfaction in a changing
health care environment: The impact of challenges to professional autonomy, authority, and
dominance’, Journal of Health and Social Behavior, Vol. 39, pp.356–367.
