Government Information Quarterly: Dongback Seo, Michel Bernsen

GOVINF-01154; No.
of pages: 13; 4C:

Government Information Quarterly xxx (2016) xxx–xxx
Contents lists available at ScienceDirect
Government Information Quarterly
journal homepage: www.elsevier.com/locate/govinf
Comparing attitudes toward e-government of non-users versus users in a

rural and urban municipality
DongBack Seo a,⁎, Michel Bernsen b
a
Dept. of Management Information Systems, Chungbuk National University, 1 Chungdae-ro, Seowon-Gu, Cheongju, Chungbuk 28644, Republic of Korea
b
Ernst & Young, Amsterdam, The Netherlands
a r t i c l e i n f o a b s t r a c t
Article history: Many e-government and Information Systems (IS) adoption studies have focused on people's attitudes during
Received 4 August 2014 the initial and post-adoption periods, but have not taken into account the fact that many people never use or ex-
Received in revised form 6 January 2016 perience e-government services. This paper investigates the attitudes of non-users versus users toward e-
Accepted 15 February 2016
government services in two locales: one urban and one rural municipality in the Netherlands. Although rural
Available online xxxx
and urban municipalities have distinct characteristics that may affect people's attitudes toward e-government,
Keywords:
the research thus far has not differentiated between them. We propose a model to investigate these differences
E-government using factors based on various IS acceptance and resistance theories, including enabling factors (e.g., perceived
E-municipality behavioral control), inhibiting factors (e.g., perceived risk) and other factors (e.g., trust and geographical close-
Perceived behavioral control ness). The model was tested via a survey of 337 non-users and users of e-government services from one rural
Rural versus urban and one urban municipality in the Netherlands. The findings reveal intriguing similarities and differences
Geographical closeness among the four groups, contributing a more nuanced perspective to the e-government and IS literature.
© 2016 Elsevier Inc. All rights reserved.
1. Introduction (municipality) government (e.g., applying and regularly renewing a

parking permit). Some services (e.g., changing a residential address)
Electronic government, commonly known as e-government, refers are used regularly, while others are more sporadic (e.g., renewing a
to the utilization of information and communications technologies, in- parking permit, paying parking fines, periodically borrowing sports fa-
cluding web-based technologies, to improve the efficiency and effec- cilities). Thus, “continuous use” refers to the usage of e-service when-
tiveness of service delivery and transactions in the public sector (Teo, ever a citizen needs to deal with this kind of government service. This
Srivastava, & Jiang, 2009). The European Action Program on e- means that citizens do not need to visit their city halls if they use the
government for 2011 to 2015 has the goal of 60% of European Union cit- services online. This phenomenon reveals issues. One is related to low
izens using e-government services by 2015. This kind of action program adoption and usage rates by citizens. The other is about different adop-
has led governments to put effort and financial investment into devel- tion and usage rates between municipalities.
oping and advocating for the use of e-government services This research investigates the attitudes of non-users versus users to-
(Overheidsmonitor, 2011). Governments have been able to convert ward e-government services by comparing two samples from two dis-
many paper-based services into e-services (Anthopoulos, Siozos, & tinct municipalities (rural versus urban area in the Netherlands). Each
Tsoukalas, 2007; Gouscos, Kalikakis, Legal, & Papadolpoulou, 2007). municipality has different characteristics such as geographical size, pop-
However, they have not been so successful in making their citizens ulation, population density and average age of population. Differences
adopt or continuously use those e-services, especially at a local (munic- can be more visible between a rural versus an urban area (Hofferth &
ipality) level. For example, the average rate of use of e-government ser- Iceland, 1998). As “rural communities have unique technological
vices provided by municipalities is 24% in the Netherlands (Van Dijk, needs,” people in urban and rural areas have different attitudes toward
Peters & Ebbers, 2008). Further, the usage rate of e-government services technologies (Gilbert, Karahalios, & Sandvig, 2010: p. 1367). For exam-
varies by municipality. ple, e-government services can be much more useful for residents
In this paper, e-government services refer to electronic services pro- who live far away from their city halls in a rural area. For this reason,
vided by a municipality website, whether the services are related to a we selected two distinct samples — one from an urban municipality
central government (e.g., applying for a passport) or to a local and the other from a rural municipality.
Although some studies focusing on users' experiences of e-
⁎ Corresponding author.
government systems have contributed to the development of e-
E-mail addresses: dseo@chungbuk.ac.kr (D. Seo), michel.bernsen@nl.ey.com government literature (e.g., Wang & Liao, 2008; Teo et al., 2009), there
(M. Bernsen). is little research that investigates the driving factors leading non-users
http://dx.doi.org/10.1016/j.giq.2016.02.002
0740-624X/© 2016 Elsevier Inc. All rights reserved.
Please cite this article as: Seo, D., & Bernsen, M., Comparing attitudes toward e-government of non-users versus users in a rural and urban
municipality, Government Information Quarterly (2016), http://dx.doi.org/10.1016/j.giq.2016.02.002
2 D. Seo, M. Bernsen / Government Information Quarterly xxx (2016) xxx–xxx
to initially adopt e-government services. This is important, considering continuously use e-government services are important. We can there-
that many citizens have not ever experienced e-government services of- fore distinguish what factors differ for non-users to adopt and for
fered by a local government in many European medium and small sized users to continuously use e-government services, because factors that
regions. For example, one municipality in The Netherlands reported that significantly affect initial adoption can possibly play no role in contin-
only four percent of its citizens had ever experienced and used its ued usage (Karahanna et al., 1999).
e-government services (Van Dijk et al., 2008). An important reason to To develop a model that identifies antecedents of intention to adopt
study non-users in the context of e-government services is the case and continuously use e-government services for non-users and users re-
when mandatory services are provided by a monopoly supplier. In the spectively, we address not only basic behavioral attitudes and subjective
context of commercial services, people usually have an option to select norm but also perceived behavioral control (Ajzen, 1988). Due to the
from various service providers or to not to use the service at all. In the fact that “successful performance of the intended behavior is contingent
context of government services, citizens must either go to their city on the person's control over the many factors that may prevent it”
halls or utilize e-government services, and they might not have a choice (Ajzen, 1988: p. 132), perceived behavioral control can play a significant
about whether or not to use the service: for instance, paying a parking role in a person's decision to adopt or continuously use e-government
fine is mandatory, while buying a book is not. This paper therefore in- services. Furthermore, we introduce perceived necessary knowledge
vestigates non-users' behavioral attitudes toward e-government, com- as a determinant of perceived behavioral control, because people who
paring to those of users. have experienced general e-commerce may believe that they have the
To compare the behavioral attitudes toward e-government services necessary knowledge to use e-government service, even though they
of non-users versus users in these two locales, we have developed a have not experienced any e-government service. Perceived ease of use
model that includes enabling factors (perceived behavioral control, per- (Riemenschneider, Harrison, & Mykytyn, 2003) and facilitating condi-
ceived usefulness, perceived ease of use and subjective norm); tions are also used as determinants of perceived behavioral control.
inhibiting factors (perceived risk, resistance to change); trust as an an- The relationships between these factors will be explained in detail in
tecedent of perceived risk; and an environmental factor (geographical the following section.
closeness between citizen's house and her/his city hall). In addition, as researchers (e.g., Bhattacherjee & Hikmet, 2007; Seo,
This paper contributes to extant literature in various ways. First, it Boonstra, & van Offenbeek, 2011) have urged, factors that inhibit people
finds a clear distinction between non-users' and users' perceptions. Sec- to accept and adopt IS are included in this model. Trust is a significant
ond, this research identifies an intriguing relationship between subjec- factor affecting other factors in IS adoption (Teo et al., 2009). Motivation
tive norm and resistance to change, by including inhibiting factors psychology assumes that motives only exert an influence on behavior if
(resistance to change and perceived risk). Third, this paper provides a they are roused by incentives (Schüler, 2010). Hence, motivation lies
model for studying people's attitudes toward e-government services. within the person and is shaped by incentives that reflect the environ-
Fourth, this paper compares findings across municipalities and presents ment. Incentives are defined as situational cues in the environment po-
new insights in the perceptions of citizens from two different munici- tentially associated with desired goal states. They therefore stimulate
palities. Therefore, practitioners can use the findings of this research goal-directed behavior (McClelland, 1985; Schneider & Schmalt, 2000;
to increase citizens' initial adoption and usage rates of e-government Beckmann & Heckhausen, 2008; Schüler, 2010). Motives can be stimu-
services. lated intrinsically and extrinsically. In the case of the latter, strengthen-
The paper is organized as follows: first, the theoretical background ing incentives for (non)users could possibly increase a person's
and hypotheses are presented. The research method is presented next, intention to adopt or use e-government services. Thus, geographical dis-
followed by a section on results. In the discussion section, the results tance between a citizen's house and her/his city hall is considered to
are discussed in depth. The conclusions are then presented. play a role as an incentive in developing her/his intention to initially
adopt or continuously use e-government services (Van Deursen, van
2. Theoretical background and hypotheses Dijk, & Ebbers, 2006). These factors will be explained in detail and
hypothesized in the section below.
One stream of e-government research has focused on identifying
motives that result in acceptance and intention to use e-government
services (e.g., Carter & Bélanger, 2005; Chang, Li, Hung, & Hwang, 2.1. Hypotheses
2005; Hung, Chang, & Yu, 2006; Yaghoubi, Kord, & Shakeri, 2010). How-
ever, there have been few empirical studies that distinguish between Perceived usefulness and perceived ease of use are important factors
people's initial adoption and post-adoption attitudes toward Informa- in accepting information systems (Davis, 1989). Perceived usefulness is
tion Systems (IS). Although the research of Karahanna, Straub, and defined as “the degree to which a person believes that using a particular
Chervany (1999) opens new avenues in this area, initial adoption and system would enhance his or her job performance” and perceived ease
the lagging acceptance behavior of non-users have not been explicitly of use is defined as “the degree to which using the technology will be
investigated much. free of effort” (Davis, 1989). The constructs of perceived usefulness
There are differences in the perceptions of non-users and users to- and perceived ease of use are directly correlated with intention to
ward IS (Karahanna et al., 1999; Ramayah, Ma'ruf, Jantan, & adopt or use (Gefen, Karahanna, & Straub, 2003; Straub, Keil, &
Mohamad, 2002). In this study, users are defined as citizens who have Brenner, 1997). In the e-government context, citizens who perceive
used e-government services and are continuously using them. Non- usefulness of e-services provided by their municipality tend to initially
users are people who have not adopted e-government services, or adopt or continuously use them, leading to the first hypothesis:
have only used them once or a few times but cannot recall their experi-
H1. A citizen's perceived usefulness of e-government services has a pos-
ences and do not plan to use at the moment. It is expected that there will
itive relationship with her/his intention to initially adopt or continu-
be fundamental differences between the perceptions and key motives of
ously use.
non-users and users, since non-users have to rely on expectations and
users can rely on past experiences. Expectations are not always formed Geographical distance might motivate citizens to use e-government
through deliberation, but may be developed based on imagination and services (Van Deursen et al., 2006). It seems logical that an individual
experience with alternative technology. Hence, non-users' and users' who has to travel a few kilometers to her/his city hall would increase
judgments and perceptions may differ greatly from each other. his/her perceived usefulness of e-government services. Incentive
Identifying potential factors from literature and applying them to is identified as a major situational cue to motivate one's behavioral in-
understand both non-users' and users' decisions to initially adopt or tent and is acknowledged in the theory of motivation psychology
D. Seo, M. Bernsen / Government Information Quarterly xxx (2016) xxx–xxx 3
(McClelland, 1985; Schneider & Schmalt, 2000; Beckmann & reason, the original definition of perceived facilitating conditions is
Heckhausen, 2008; Schüler, 2010). Therefore, geographical distance is redefined in this paper as “the degree to which an individual believes
hypothesized as having a relationship with both perceived usefulness that technical infrastructure exists to support use of the system.” Per-
and intention to initially adopt or continuously use. So the closer a citizen ceived facilitating conditions can be a determinant of perceived behav-
lives to her/his city hall, the less (s)he develops the intention to initially ioral control. For example, people are not able to use e-government
adopt or continuously use e-government services and the less usefulness services if they have no computer and/or Internet access. This results
(s)he perceives. The following hypotheses are thus formulated: in the following hypothesis:
H2. A citizen's geographical closeness from her/his residence to city hall H4a. A citizen's perceived facilitating conditions of e-government ser-
has a negative relationship with her/his intention to initially adopt or vices have a positive relationship with her/his perceived behavioral
continuously use e-government services. control.
H2a. A citizen's geographical closeness from her/his residence to city A term that is frequently used to explain one's technological knowl-
hall has a negative relationship with her/his perceived usefulness of e- edge of a system is self-efficacy. Self-efficacy is often hypothesized as
government services. being an antecedent of perceived behavioral control (Hsu & Chiu,
2004; Hung et al., 2006; Yaghoubi et al., 2010). Bandura (1982) defines
As mentioned, perceived ease of use is also an important factor in self-efficacy as “judgments of how well one can execute courses of ac-
accepting information systems (Davis, 1989). Some studies have tion required to deal with prospective situations” (p. 122). According
found little (e.g., Carter & Bélanger, 2005; Van Dijk et al., 2008) or no ev- to Davis (1989), self-efficacy is similar to perceived ease of use. If one
idence (e.g., Tseng & Lo, 2011) to support a relationship of perceived expects or perceives the use of a system as being free of effort, it does
ease of use with intention to use. Meanwhile, some other studies not prevent the person from using the system, which is a form of behav-
(e.g., Muthitcharoen, Palvia, & Grover, 2011; Tseng & Lo, 2011) have ioral control. Compeau and Higgins (1995) therefore coined the spin-off
found more evidence for the relationship of perceived ease of use with definition “computer self-efficacy,” which is in fact similar to perceived
perceived usefulness than for the relationship of perceived ease of use ease of use. Hence, perceived ease of use becomes one determinant of
with intention to use. To clarify these mixed research results in this par- perceived behavioral control.
ticular context, two hypotheses are formulated: Reconsidering the original definition of self-efficacy in the context of
this study, it can be hypothesized that people will prefer traditional gov-
H3. A citizen's perceived ease of use of e-government services has a
ernment services over a counter if they are unfamiliar with and insecure
positive relationship with her/his intention to initially adopt or contin-
about certain procedures, because they are guided through the proce-
uously use.
dures by a municipal employee at a city hall. Some people may experi-
H3a. A citizen's perceived ease of use of e-government services has a ence the procedure of filing taxes as being hard to understand and do
positive relationship with her/his perceived usefulness. not recognize terms used in the documents. In this sense, they need
basic knowledge to file taxes. This kind of knowledge can empower peo-
Over-emphasizing the significance of perceived usefulness and per- ple to perform relevant tasks. Thus, we create the factor of perceived
ceived ease of use in accepting information systems is based on the as- necessary knowledge and define it as “the knowledge one perceives to
sumption that a person is free to act without limitation when (s)he be required in understanding related terms and following a given proce-
forms an intention to act. In practice, constraints such as limited capabil- dure.” It means that a person would not limit her/his action if (s)he ex-
ities, time, environment and resources limit the freedom or ability to act. pects to have the knowledge to understand and use the system.
For this reason, the concept of perceived behavioral control has been in- Therefore, the next hypotheses are formulated:
troduced to bridge the intention to act and any limitations (Ajzen,
1991). Behavioral control is defined as “people's perception of the H4b. A citizen's perceived ease of use of e-government services has a
ease or difficulty of performing the behavior of interest” (p. 183). In positive relationship with her/his perceived behavioral control.
other words, one might want or intend to do something, but there
H4c. A citizen's perceived necessary knowledge of e-government ser-
might be obstacles keeping the person from performing the intended
vices has a positive relationship with her/his perceived behavioral
behavior (Ajzen, 1991). Therefore, perceived behavioral control plays
control.
an important role in determining one's behavioral intention
(Mathieson, 1991). Thus, the following hypothesis is formulated: Various researchers (e.g., Mayer, Davis, & Schoorman, 1995; De
Ruyter, Wetzels, & Kleijnen, 2001; Gefen, 2002; Wang & Emurian,
H4. A citizen's perceived behavioral control of e-government services
2005; Horst, Kuttschreuter, & Gutteling, 2007) have addressed the im-
has a positive relationship with her/his intention to initially adopt or
portance of trust as a factor that determines intention to use an elec-
continuously use.
tronic innovation. Trust is a complex and abstract construct that is
According to the Unified Theory of Acceptance and Use of Technol- difficult to define. The definition of Rotter (1967) is adopted for this
ogy (UTAUT) model (Venkatesh, Morris, Davis, & Davis, 2003), per- paper: “expectancy held by individuals or groups that the word, prom-
ceived facilitating conditions are similar to perceived behavioral ise, verbal or written statement of another can be relied on.” Translating
control. However, there is a fundamental difference. Perceived behav- this definition to the e-government context, it is about trust between an
ioral control is defined relatively globally, while perceived facilitating individual and the government and its IT facilities. One should have a
conditions are defined as being specific to the support of organizational certain level of trust in a service provider in order to adopt and use its
and IT infrastructure: “the degree to which an individual believes that e-services (Hoffman, Novak, & Peralta, 1999). Thus, lack of trust may
an organizational and technical infrastructure exists to support use of eventually have a negative impact on the adoption or use of e-
the system” (Venkatesh et al., 2003: p. 453). According to this definition, government services (Lee & Turban, 2001; McKnight & Chervany,
organizational support is not entirely applicable in the context of e- 2001; Bhattacherjee, 2002). In this paper, trust in a government is
government services, because citizens adopt and use the services in pri- based on perception of the integrity of its e-services, the government's
vate or public environments. A company can build every possible tech- ability to deal with the transaction, and the government's security in
nical infrastructure within its boundary for employees. However, a dealing with the transaction in the best interest of users. Therefore,
government is much more restricted in terms of utilizing resources trust is about relying on a government.
and managing a large territory. Citizens usually use technical infrastruc- Meanwhile, risk perception is formed by the presentation of an orga-
ture provided by companies to access e-government services. For this nization, the environment in which it acts, and the way it provides its
products or services. One's perception of a service as being risky may use of e-government services is entirely voluntary, so inaction can be
prevent the person from engaging in transactions. Pavlou (2003) mistaken for apathy resistance. Inaction can also be the product of not
pinpointed two forms of risk: i) risk of monetary loss associated with being aware of the existence of e-government services. Therefore, it is
vulnerability related to incomplete or distorted information provided important to establish the presence or absence of resistance to change.
by Web retailers and third parties (Lee, 1998), and ii) risk of loss of pri- User resistance has been recognized as one of the first-ranked chal-
vacy associated with providing personal information to online service lenges for the implementation of large-scale IS (Kim & Kankanhalli,
providers (Culnan & Armstrong, 1999). This perceived risk is applied 2009). Although inhibiting factors (e.g., resistance to change) is impor-
to government-to-citizen relationships and transactions in the e- tant, there has been little attention to this concept (Bhattacherjee &
government context. Hikmet, 2007). This results in the formulation of the following
When people trust others, they assume that those they trust will be- hypothesis:
have as expected, reducing the complexity of the interaction. Con-
sumers tend to assume that a trusted Web retailer will not engage in H7. A citizen's resistance to change has a negative relationship with
opportunistic behavior (Gefen, 2000), so trust reduces the perceived her/his intention to initially adopt or continuously use e-government
risk (Luhmann, 1979; Lewis & Weigert, 1985). There is only a longing services.
for trust in a risky environment (Mayer et al., 1995). However, theorists Fig. 1 visualizes the proposed thirteen hypotheses. The research
do not unanimously agree on the relationship between risk perception model was statistically tested in order to accept or reject the formulated
and trust. Also, they disagree about risk perception being the antecedent hypotheses. The applied method is explained in the next chapter.
of trust (e.g., Horst et al., 2007) and that trust is the antecedent of per-
ceived risk (e.g., Warkentin, Gefen, Pavlou, & Rose, 2002; Bélanger &
3. Research method
Carter, 2008). Since establishing this relationship is not the goal of this
research, the dissonance on this matter will be respected. Following
Heeks and Bailur (2007) call for an increase in the use of quantitative
the view that perceived risk is directly related to intention to use and
analysis with strict collection and analysis of data in e-government re-
that trust reduces perceived risk, two hypotheses are formulated:
search. To answer this call and in order to test the hypotheses, data
H5. A citizen's perceived risk of e-government services has a negative was collected and separated into four groups based on two criteria:
relationship with her/his intention to initially adopt or continuously i) users versus non-users, and ii) Groningen (urban municipality) ver-
use. sus Emmen (rural municipality).
H5a. A citizen's trust in e-government services has a negative relation-

ship with her/his perceived risk. 3.1. Data collection
Equity theory suggests that individuals are constantly concerned A survey on e-government services was conducted among users and
about their inputs, outcomes, and the fairness of the exchange in non-users from the municipalities of Groningen and Emmen in the
every exchange relationship. Individuals are also constantly comparing Netherlands. Groningen and Emmen were selected because they both
themselves with others in their reference group to assess whether the score relatively high on designing and providing e-government services
relative gains are the same (Adams, 1963; Adams, 1965; Walster, (Van Dijk et al., 2008). Although the distance between Groningen and
Walster, & Berschied, 1978). Based on this view, if they believe that peo- Emmen municipalities is relatively small, approximately 70 km, they
ple in their direct environment use e-government services and have rel- have fundamentally different characteristics (Table 1) as urban and
atively higher gains than themselves in using and receiving public rural municipalities, respectively.
services, they might be stimulated to adopt the e-government services. Urban and rural areas vary by countries and nations, but most coun-
Another theory that acknowledges the effect of social influence on tries classify an urban area as one with a large population with less
human behavior is Social Cognitive Theory (Bandura, 1986). Social Cog- space between structures. The average age of the population tends to
nitive Theory is based on the notion that environmental influences such be lower in an urban area than in a rural area. Analyzing the information
as social pressures, conditional characteristics, and cognitive and other
personal factors including personality and demographic characteristics
are reciprocally determined. Thus, individuals choose the environments
in which they exist and at the same time are influenced by those envi-
ronments. Applying this theory in the context of e-government, if
one's environment, consisting of friends, family, neighbors and/or co-
workers, is positive and supportive of the use of e-government services,
(s)he will be influenced by this environment in adopting or using e-
services.
Considering equity theory and social cognitive theory, the following
hypothesis is developed:
H6. Subjective norm has a positive relationship with one's intention to

initially adopt or continuously use.
There are different levels of resistance behavior from low levels of
use to dysfunctional and harmful use (Martinko, Henry, & Zmud,
1996). In practice, user resistance can take on a wide variety of behav-
ioral forms. Perhaps the most accepted theory today is the theory of
multi-level resistance by Lapointe and Rivard (2005). They distinguish
four levels of resistance: apathy, passive resistance, active resistance,
and aggressive resistance. Examples of these four levels are inaction, re-
fusal, voicing dissatisfaction, and rebellion, respectively. The behavior of
any of these levels is undesirable when diffusing an IS. This emphasizes
the importance of identifying whether there is resistance to change. The Fig. 1. Research model.
Table 1 3.4. Field survey

Demographic information on Groningen and Emmen.
Demographic information Groningen Emmen A field survey was conducted for five weeks. Participants were
Population 187,197a 109,493b
solicited through email and social media (e.g., LinkedIn, Facebook and
Gender distribution (M/F) 49.5%/50.5%a 49.4%/50.6%b Hyves). In addition, we asked people to post our survey invitation mes-
Average age 36.4c 41.7b sage on their social media. For this reason, it was not known that how
Geographical size (in m2) 83,690,000a 346,240,000b many people saw the survey invitation message. As a result, it was not
a
Gemeente Groningen (2010). possible to calculate the number of people who were contacted. How-
b
Gemeente Emmen (2010). ever, we were able to calculate the numbers of completed, partially
c
Groningen Toerisme (2010). completed, and refused or break-off surveys. In total, 628 individuals
started the online survey and 268 individuals completed all stages of
presented in Table 1, the population density and demographic informa- the online survey. 336 individuals initiated the online survey, but did
tion reveal Groningen as an urban area and Emmen as a rural area in The not finish less than 50% of questions, which were categorized as
Netherlands. break-off surveys. 24 individuals who filled out more than 50% of ques-
tions but did not complete all questions were categorized as partially
completed surveys. In a face to face survey, we did not calculate how
3.2. Measurement instrument many people we solicited. However, 110 hard copies of the survey
were distributed at the city halls and various neighborhoods. This re-
All constructs, except for geographical closeness, consisted of three sulted in 91 returned surveys. Among them, 69 were filled out
items. Geographical closeness in kilometers was determined by asking completely; 18 individuals were categorized as break-off surveys; and
for the distance from the respondent's home to her/his city hall. For all 4 individuals were categorized as partially completed surveys following
measurement items except geographical closeness and demographic the above rules. 19 not returned surveys were considered as refused
information, a seven-point Likert scale was adopted from (1) strongly surveys. Individuals who completed all questions were categorized as
agree to (7) strongly disagree (Lehmann & Hulbert, 1972). completed surveys. In total 337 surveys were completed. Although we
Measures were adopted and reformulated to fit the context. The sur- could not exactly calculate a response rate, we used a cooperation rate
vey items for users were formulated to address their experience with e- from American Association for Public Opinion Research (AAPOR)
government services, and the survey items for non-users were formu- standards. Especially, Cooperation Rate 3 (COOP3) considers those
lated to address their expectations of e-government services. Appendix unable to participate in a survey as incapable of cooperating, so
A shows the measurement items with sources. The items were random- they are excluded from the base. COOP3 only considers completed,
ized in the actual survey to avoid creating bias and to test consistency. partially completed, refused, and break-off surveys for its calcula-
As explained in the previous section, perceived necessary knowl- tion. COOP3 is the completed surveys divided by completed surveys
edge was added as an antecedent of perceived control behavior and plus partially completed surveys plus refused and break-off surveys.
the definition of perceived facilitating conditions was adjusted for the Therefore, COOP3 is calculated as (268 + 69) divided by
e-government context. Hence, new scale items for perceived necessary (268 + 69 + 336 + 24 + 19 + 18 + 4), which is 45.7%.
knowledge were developed and tested according the new scale develop- As discussed before, the 337 surveys were divided into the four
ment process. groups. Table 2 shows demographic information for these four groups.
The results are quite similar to the actual profile per municipality as il-
lustrated in Table 1. Groningen is a university city as well as the largest
3.3. Pretest city in the north of the Netherlands. For this reason, the respondents in
The survey was qualitatively and quantitatively tested on a small

Table 2
scale. Citizens from Groningen and Emmen were asked to take part in Demographic information of the four groups.
the pretest. Respondents identified themselves as users or non-users
by answering the questions: 1) “Do you take care of your own transac- Demographic Category NE (N NG (N UE (N UG (N
variable = = = =
tions with a municipality? If not, you do not need to complete this sur-
111) 73) 70) 83)
vey”; 2) “Are you aware of and know municipal e-services?”; and
3) “Have you used and remembered your experience on municipal e- Age 18–29 25.2% 45.2% 31.4% 50.6%
30–49 30.6% 31.5% 38.6% 25.3%
services in the past?” If a respondent answered ‘yes’ for the three ques- 50–64 33.3% 15.1% 22.9% 20.5%
tions, then (s)he was considered a user. If a respondent answered ‘yes’ 65+ 6.3% 8.2% 5.7% 3.6%
for the first two questions but ‘no’ for the third question, then (s)he Unwilling to disclose 4.5% − 1.4% −
was considered a non-user. Others were excluded. The test was con- Gender Male 50.5% 50.7% 44.3% 54.2%
Female 49.5% 49.3% 55.7% 45.8%
ducted face-to-face. The goal was to test the clarity of the survey and re-
Unwilling to disclose − − − −
liability of the measurement items, not to create preliminary results. Education Academic university 4.5% 27.4% 17.1% 45.8%
The pretest had 23 respondents. Thirteen of them were non-users and Polytechnic college 18% 40.1% 35.7% 37.3%
ten were users. It took them an average of 4 min and 32 s to complete Secondary vocational 43.2% 21.9% 25.7% 14.5%
the test. school
Primary vocational 21.6% 5.5% 15.7% 2.4%
The questionnaire was accurately translated into Dutch based on the
school
measurement items shown in Appendix A. An independent individual Secondary school 12.6% 2.7% 5.7% −
fluent in both Dutch and English translated the survey back into English. Unwilling to disclose − 1.4% − −
The retranslated survey was compared to the original measurement Geographical 0–5 km 12.6% 45.2% 11.4% 56.6%
closeness 6–10 km 23.4% 28.8% 20% 26.5%
items. Based on this comparison, a few minor changes were made.
11–15 km 36.9% 13.7% 30% 8.4%
Then, the survey was sent to the participating municipalities. They 15–20 km 26.1% 9.6% 34.3% 8.4%
recommended changing the order of the survey to test the measure- 21+ 0.9% − 4.3% −
ment items and gain demographic information. Examples of e- Unwilling to disclose − 2.7% − −
government services were included in the introduction of the survey NE: non-users in Emmen; NG: non-users in Groningen; UE: users in Emmen; and UG:
to provide a clear idea about e-government services. users in Groningen.
Groningen are younger, more educated and live closer to their city hall small letter of ‘dropped’ means that the same items in other groups
than those in Emmen. The gender breakdown of participants is also were dropped, even though they exceed 0.70, so that we can compare
representative. the hypotheses across the four groups.
Second, Cronbach's alpha (α) for each construct was calculated for
4. Results construct reliability (Cronbach, 1984). Crohnbach's alpha must exceed
0.70 to be considered reliable. All constructs met this requirement
Two different methods were applied for statistical testing: Dunnett's (see Table 3).
2-sided t-test and Partial Least Squares. The results of these tests are Third, after establishing convergent and discriminant validities at a
presented next. factor level, the validity of a construct was tested by calculating Average
Variance Extracted (AVE). All constructs met the requirement of more
4.1. Comparing constructs than 0.50 (Fornell & Larcker, 1981) (see Table 3).
Fourth, the inter-construct correlation matrices (Table 4 and
The constructs of the research model were compared across the four Table 5) show the comparison by inter-correlations among latent vari-
groups for two reasons. First, there was no need to distinguish the ables and corresponding square roots of AVEs. Each square root of AVE
groups from each other and they could be merged if there were no dif- is greater than its correlation with any of the other constructs. These
ference between the groups. Second, if there was a difference, it was matrices also show that convergent and discriminant validities hold
necessary to provide insight on the size of the difference in order to an- for each latent construct.
swer the research questions.
For these reasons, Dunnett's 2-sided t-test was selected. Dunnett's t- 4.2.2. Hypotheses testing
test treats one group as a control group and compares other groups After confirming that the measurement model was considered to be
against it. The means per construct are compared. This test was con- reliable, the research model as a whole and individual structural paths
ducted using IBM SPSS Statistics 19. The results of the ten constructs were statistically tested. The result of each group is presented.
and the factor of distance are represented in Appendix B.
To compare the groups to each other, six possible comparisons could 4.2.2.1. Non-users in the rural area (Emmen). The model of non-users in
be made. However, comparing non-users in Emmen to users in Gro- the rural area (Fig. 2) explains 69% of the variance for intention to
ningen and non-users in Groningen to users in Emmen seemed irratio- adopt. Chin (1998) prescribes labeling R2 values of 0.15, 0.35, and 0.67
nal, because user and non-users should be compared within the same as weak, moderate, and substantial, respectively; therefore this model
environment. Thus, we compared between user groups and between is considered to have substantial predictive power. The model explains
non-user groups in the two areas. 52% of the variance for perceived behavioral control, and eleven out of
Comparing user groups in Groningen and Emmen, two constructs – the thirteen hypotheses are supported.
trust (−0.48) and perceived facilitating conditions (−0.38) – were sig- Perceived usefulness (β = 0.120, ρ b 0.05) is positively and signifi-
nificantly different (ρ b 0.05). The results also indicated that geograph- cantly correlated with intention to adopt, providing support for hypoth-
ical closeness (−1.31) was different. esis H1. Geographical closeness has a negative and significant
Comparing non-user groups in Groningen and Emmen, two con- relationship with intention to adopt (β = − 0.293, ρ b 0.001) and
structs – resistance to change (0.66) and intention to use (− 0.57) – with perceived usefulness (β = −0.197, ρ b 0.001), supporting both hy-
were significantly different (ρ b 0.05). Similar to the comparison of potheses H2 and H2a, respectively. Perceived ease of use has no signifi-
the two user groups, geographical closeness (1.00) was different. cant relationship with intention to adopt (β = 0.036, ρ N 0.05).
The comparison between users and non-users in Emmen presented Therefore, hypothesis H3 is rejected. Perceived ease of use shows posi-
clear differences. Eight out of eleven constructs were significantly differ- tive and significant correlations to perceived usefulness (β = 0.380,
ent. No evidence was found to support a difference for trust, subjective ρ b 0.001) and perceived behavioral control (β = 0.351, ρ b 0.001),
norm, and geographical closeness. Almost equal results were produced supporting hypotheses H3a and H4b, respectively. Perceived behavioral
in the comparison between users and non-users in Groningen except control exerts a positive and significant effect on intention to adopt
trust, which was also significantly different. These results make sense (β = 0.352, ρ b 0.001), providing support for hypothesis H4. Perceived
considering that citizens from the same area tend to share subjective facilitating conditions (β = 0.266, ρ b 0.001) and perceived necessary
norm and geographical closeness. knowledge (β = 0.334, ρ b 0.001) exercise a positive and significant re-
Although the two non-user groups and two user groups across mu- lationship with perceived behavioral control, supporting hypotheses
nicipalities had similar perceptions on the various constructs, for empir- H4a and H4c, respectively. Perceived risk (β = −0.185, ρ b 0.001) exerts
ical research, the few differences were enough to consider the groups as a negative and significant effect on intention to adopt, supporting hy-
distinct. However, non-users and users of the same municipality had pothesis H5. Perceived risk is negatively and significantly influenced
different behavioral perceptions of e-government services. Hence, the by trust (β = − 0.407, ρ b 0.001), providing support for hypothesis
four compared groups were too different from each other to be merged
together.
Table 3
Measurement properties. (Cronbach's alpha and average variance extracted).
4.2. Model testing
Non-users in Non-users in Users in Users in
Emmen Groningen Emmen Groningen
4.2.1. Scale validation
To answer the research questions by comparing the groups, the par- α AVE α AVE α AVE α AVE
tial least square (PLS) procedure was adopted to validate the proposed FC 0.80 0.83 0.95 0.95 0.84 0.86 0.81 0.84
model using SmartPLS 2.0. INT 0.94 0.89 0.96 0.93 0.96 0.92 0.91 0.85
The measurement model was tested in three ways. First, loadings of PBC 0.88 0.81 0.91 0.85 0.91 0.85 0.96 0.92
PEOU 0.79 0.71 0.96 0.93 0.94 0.90 0.94 0.89
all items were calculated for convergent and discriminant validities. If
PR 0.80 0.84 0.88 0.89 0.85 0.87 0.88 0.89
the loading of an individual indicator exceeded 0.70, it was considered PNK 0.94 0.90 0.92 0.86 0.88 0.81 0.94 0.89
valid (Fornell & Larcker, 1981). In total, the common sets of three PU 0.75 0.67 0.79 0.70 0.88 0.81 0.85 0.77
items that did not meet this requirement were dropped. All factor load- RC 0.76 0.68 0.93 0.88 0.74 0.66 0.93 0.87
ings are listed in Appendix C. The capital letter of ‘DROPPED’ means that SN 0.84 0.76 0.83 0.75 0.81 0.71 0.82 0.73
T 0.86 0.88 0.72 0.77 0.72 0.77 0.93 0.93
a corresponding item within a given group is less than 0.70, while the
Table 4
Inter-construct correlation matrices.
Non-users in Groningen
FC INT PBC PEOU PR PNK PU RC SN T

0.97a 0.35 0.74 0.71 −0.35 0.70 0.54 −0.38 0.66 0.72 FC
0.96a 0.68 0.71 −0.82 0.51 0.78 −0.57 0.56 0.53 INT
0.92a 0.92 −0.59 0.79 0.75 −0.50 0.74 0.69 PBC
FC 0.91a 0.96a −0.65 0.68 0.75 −0.57 0.77 0.66 PEOU
INT 0.23 0.94a 0.95a −0.36 −0.70 0.47 −0.54 −0.50 PR
PBC 0.45 0.68 0.90a 0.93a 0.52 −0.46 0.64 0.56 PNK
PEOU 0.37 0.52 0.62 0.84a 0.84a −0.52 0.64 0.63 PU
PR −0.27 −0.51 −0.28 −0.26 0.91a 0.94a −0.48 −0.45 RC
PNK 0.17 0.43 0.56 0.51 −0.34 0.95a 0.87a 0.74 SN
PU 0.45 0.51 0.42 0.44 −0.38 0.21 0.82a 0.88a T
RC −0.33 −0.44 −0.33 −0.27 0.41 −0.13 −0.16 0.82a
SN 0.49 0.48 0.60 0.49 −0.21 0.31 0.49 −0.23 0.87a
T 0.54 0.35 0.40 0.38 −0.41 0.19 0.51 −0.13 0.63 0.94a
Non-users in Emmen
a
Square-root of average variance extracted; Non-users in Emmen and Groningen.
H5a. No evidence is found to support a relationship of subjective norm (β = 0.165, ρ b 0.01), supporting hypothesis H4. Hypothesis H4a is not
(β = 0.077, ρ N 0.05) with intention to adopt. Hypothesis H6 is therefore supported, since there is no evidence found to support a positive rela-
rejected. Resistance to change (β = −0.172, ρ b 0.001) exerts a nega- tionship of perceived facilitating conditions (β = 0.050, ρ N 0.05) with
tive and significant effect on intention to adopt, supporting hypothesis perceived behavioral control. Perceived necessary knowledge (β =
H7. 0.297, ρ b 0.001) exercises a positive and significant relationship with
perceived behavioral control, supporting hypothesis H4c. Perceived
4.2.2.2. Non-users in the urban area (Groningen). The model of non-users risk (β = −0.488, ρ b 0.001) exerts a negative and significant relation
in the urban area (Fig. 3) explains 78% of the variance for intention to with intention to adopt, supporting hypothesis H5. Perceived risk is neg-
adopt. Therefore, this model is considered to have substantial predictive atively and significantly influenced by trust (β = −0.499, ρ b 0.001),
power. It is notable that the model explains 90% of the variance for per- supporting hypothesis H5a. No evidence is found to support a relation-
ceived behavioral control that has the antecedents of perceived ease of ship of subjective norm (β = − 0.085, ρ N 0.05) with intention to
use, perceived facilitating conditions, and perceived necessary knowl- adopt, rejecting hypothesis H6. Resistance to change (β = − 0.144,
edge. Among them, perceived ease of use and perceived necessary ρ b 0.001) exerts a negative and significant effect on intention to
knowledge show significant relationships with perceived behavioral adopt, supporting hypothesis H7.
control. In addition, nine of the thirteen hypotheses are supported.
Perceived usefulness (β = 0.290, ρ b 0.001) is positively and signif- 4.2.2.3. Users in the rural area (Emmen). The model of users in the rural
icantly correlated to intention to adopt. Hence, hypothesis H1 is sup- area (Fig. 4) explains 86% of the variance for intention to continuously
ported. There is no evidence to support a relationship between use, which is considered to have substantial predictive power. This
geographical closeness (β = −0.039, ρ N 0.05) and intention to adopt. model explains 82% of the variance for perceived behavioral control
Therefore, hypothesis H2 is rejected. Geographical closeness has a nega- and eleven out of the thirteen hypotheses are supported.
tive and significant relationship with perceived usefulness Perceived usefulness (β = 0.359, ρ b 0.001) is positively and signif-
(β = − 0.103, ρ b 0.01), supporting hypothesis H2a. Perceived ease of icantly correlated to intention to use. Hence, hypothesis H1 is supported.
use (β = 0.002, ρ N 0.05) is not significantly correlated to intention to Geographical closeness has a negative and significant relationship with
adopt, so hypothesis H3 is rejected. Perceived ease of use exerts a posi- both intention to use (β = −0.064, ρ b 0.05) and perceived usefulness
tive and significant relationship on perceived usefulness (β = 0.748, (β = − 0.165, ρ b 0.001). Therefore, hypotheses H2 and H2a are sup-
ρ b 0.001) and perceived behavioral control (β = 0.680, ρ b 0.001), ported, respectively. Perceived ease of use exerts a positive and signifi-
supporting hypotheses H3a and H4b, respectively. Perceived behavioral cant relationship with intention to use (β = 0.176, ρ b 0.01), perceived
control exercises a positive and significant effect on intention to adopt usefulness (β = 0.761, ρ b 0.001) and perceived behavioral control
Table 5
Inter-construct correlation matrices.
Users in Groningen

0.92a 0.53 0.58 0.57 −0.56 0.64 0.61 −0.28 0.25 0.48 FC
0.92a 0.79 0.78 −0.67 0.77 0.80 −0.64 0.42 0.45 INT
0.96a 0.92 −0.80 0.87 0.84 −0.32 0.27 0.61 PBC
FC 0.93a 0.94a −0.85 0.91 0.86 −0.25 0.31 0.63 PEOU
INT 0.61 0.96a 0.94a −0.80 −0.79 0.17 −0.18 −0.65 PR
PBC 0.65 0.87 0.92a 0.94a 0.86 −0.30 0.28 0.56 PNK
PEOU 0.75 0.83 0.84 0.95a 0.87a −0.39 0.44 0.62 PU
PR −0.36 −0.60 −0.62 −0.54 0.93a 0.93a −0.42 −0.21 RC
PNK 0.57 0.61 0.82 0.68 −0.58 0.90a 0.85a 0.13 SN
PU 0.69 0.90 0.87 0.83 −0.55 0.63 0.90a 0.97a T
RC −0.48 −0.60 −0.63 −0.64 0.58 −0.59 −0.63 0.81a
SN 0.27 0.32 0.15 0.23 0.16 −0.03 0.36 −0.10 0.85a
T 0.37 0.23 0.30 0.30 −0.24 0.31 0.30 −0.15 0.10 0.88a
Users in Emmen
a
Square-root of average variance extracted; Users in Emmen and Groningen.
Fig. 2. PLS analysis of the research model for non-users in the rural area (Emmen). Path Fig. 4. PLS analysis of the research model for users in the rural area (Emmen). Path
significance: *ρ b 0.05, **ρ b 0.01, ***ρ b 0.001. significance: *ρ b 0.05, **ρ b 0.01, ***ρ b 0.001.
(β = 0.548, ρ b 0.001). Thus, hypotheses H3, H3a and H4b are supported. Perceived risk is negatively and significantly influenced by trust
Hypothesis H4a is not supported, since there is no evidence found to (β = −0.238, ρ b 0.001), providing support for hypothesis H5a. Subjec-
support a positive relationship between perceived facilitating condi- tive norm (β = 0.115, ρ b 0.05) exerts a positive and significant effect on
tions (β = −0.03, ρ N 0.05) and perceived behavioral control. Perceived intention to use, supporting hypothesis H6. No evidence is found to sup-
necessary knowledge (β = 0.465, ρ b 0.001) exercises a positive and port a relationship of resistance to change (β = 0.040, ρ N 0.05) with in-
significant relationship with perceived behavioral control, which in tention to use. Hypothesis H7 is therefore rejected.
turn exercises a positive and significant relationship with intention to
use (β = 0.297, ρ b 0.001), supporting hypotheses H4c and H4, respec- 4.2.2.4. Users in the urban area (Groningen). The model of users in the
tively. Perceived risk (β = −0.151, ρ b 0.01) exerts a negative and sig- urban area (Fig. 5) explains 83% of the variance for intention to contin-
nificant relationship with intention to use, supporting hypothesis H5. uously use, which is considered to have substantial predictive power.
Fig. 3. PLS analysis of the research model for non-users in the urban area (Groningen). Fig. 5. PLS analysis of the research model for users in the urban area (Groningen). Path
Path significance: *ρ b 0.05, **ρ b 0.01, ***ρ b 0.001. significance: *ρ b 0.05, **ρ b 0.01, ***ρ b 0.001.
This model explains 86% of the variance for perceived behavioral control influences rural users in developing their intentions to continuously
and nine out of the thirteen hypotheses are supported. use, but not others. Hypothesis H7 is rejected by rural users, but sup-
Perceived usefulness (β = 0.169, ρ b 0.01) practices a positive and ported by the three other groups. Thus, resistance to change influences
significant relationship with intention to use, providing support for hy- non-users in both municipalities in developing their intentions to adopt
pothesis H1. Geographical closeness has a negative and significant rela- and urban users in their intentions to continuously use.
tionship with perceived usefulness (β = − 0.056, ρ b 0.05), but no
significant relationship with intention to use (β = 0.018, ρ N 0.05),
supporting hypothesis H2a, but rejecting hypothesis H2. Perceived ease 5. Discussion and conclusions
of use has a positive and significant relationship with intention to use
(β = 0.322, ρ b 0.001), perceived usefulness (β = 0.840, ρ b 0.001), In this section, we begin with a discussion of the research results.
and perceived behavioral control (β = 0.749, ρ b 0.001), supporting hy- First, comparing the two non-user groups between the rural (Emmen)
potheses H3, H3a and H4b, respectively. Perceived behavioral control ex- and urban (Groningen) municipalities, there are only two differences:
erts a positive and significant effect on intention to use (β = 0.174, i) geographical closeness influences only rural non-users in developing
ρ b 0.01), supporting hypothesis H4. Perceived facilitating conditions their intentions to initially adopt e-government services and ii) per-
(β = 0.044, ρ N 0.05) have no significant relationship with perceived be- ceived facilitating conditions influence only rural non-users' percep-
havioral control. Hypothesis H4a is therefore rejected. Perceived neces- tions of behavioral control.
sary knowledge (β = 0.165, ρ b 0.01) exercises a positive and During the pre-adoption period, non-users across the urban and
significant relationship with perceived behavioral control, supporting rural municipalities show similar behavioral perceptions except per-
hypothesis H4c. Hypothesis H5 is not supported, because there is no ev- ceived facilitating conditions. Urban non-users do not care about per-
idence found to support a negative relationship of perceived risk ceived facilitating conditions in developing their perceptions of
(β = − 0.054, ρ N 0.05) with intention to use. Perceived risk is nega- behavioral control. Geographical closeness as an actual fact influences
tively and significantly influenced by trust (β = − 0.649, ρ b 0.001), people who live in the rural area.
providing support for hypothesis H5a. No evidence is found to support Second, comparing the two user groups between the rural (Emmen)
a relationship of subjective norm (β = 0.017, ρ N 0.05) with intention and urban (Groningen) municipalities, there are four differences. To de-
to use. Hypothesis H6 is therefore rejected. Resistance to change velop intention to continuously use e-government services,
(β = −0.427, ρ b 0.001) exerts a negative and significant effect on in- i) geographical closeness influences the rural users, but not the urban
tention to use, supporting hypothesis H7. users. This means that the factor of geographical closeness is relevant
to people regardless of whether they are users or non-users in the
4.3. Comparisons of the hypotheses across the four groups rural municipality. Further, ii) social norm and iii) perceived risk affect
the rural users, but not the urban users; and iv) resistance to change
Table 6 shows the results and comparisons of hypotheses. Hypothe- does not influence the rural users, but does influence the urban users.
ses H1, H2a, H3a, H4, H4b, H4c and H5a are supported across all four groups. During the post-adoption period, the two user groups (urban and
Hypothesis H2 is supported by the two groups in the rural area rural) show more differences. Both user groups are not influenced by
(Emmen), but not by the groups in the urban area (Groningen). This in- perceived facilitating conditions in developing perceptions of behav-
dicates that geographical closeness plays a role in determining citizens' ioral control. For rural users, the environmental factors (distance and
intentions to initially adopt or continuously use in the rural municipal- social norm) influence their intentions to continuously use e-
ity, but not in those in the urban municipality. Hypothesis H3 is sup- government services, but these same factors do not affect urban users'
ported across the two user groups, but rejected across the two non- intentions. However, the inhibiting factors of perceived risk and resis-
user groups. This indicates that perceived ease of use matters to users tance to change influence the rural and urban users, respectively, so
in both municipalities in determining their intentions to continuously rural users' continued use depends on their risk perception, while
use, but not to non-users in both municipalities. Hypothesis H4a is sup- urban users' continued use depends on their degree of resistance to
ported by the group of rural non-users, but rejected across the three change. These findings are confirmed by literature that shows that IS
other groups. This stresses that rural non-users' perceptions of behav- adoption in rural areas is different from that in urban areas
ioral control are influenced by perceived facilitating conditions, which (e.g., Gilbert et al., 2010; Hollifield & Donnermeyer, 2003; LaRose,
is not the case for other groups. Hypothesis H5 is rejected by urban Gregg, Strover, Straubhaar, & Carpenter, 2007; Peronard & Just, 2011).
users, but accepted by the three other groups. This indicates that risk Third, comparing non-users to users in the rural municipality
perception plays a role in determining intention to initially adopt or (Emmen), there are four differences: i) perceived facilitating conditions
continuously use for both rural non-users and users as well as for influence only non-users' perceptions of behavioral control; ii) resis-
urban non-users. Hypothesis H6 is supported by the rural users, but tance to change as an inhibitor influences only non-users in their inten-
rejected across the three other groups. Therefore, subjective norm tions to initially adopt; iii) perceived ease of use influences only users in
Table 6
Hypotheses overview.
Hypo-thesis Relationship Rural non-users (E.) Urban non-users (G.) Rural users (Emmen) Urban users (Groningen)
1 PU → INT Supported Supported Supported Supported

2 Distance → INT Supported Not supported Supported Not supported
2a Distance → PU Supported Supported Supported Supported
3 PEOU → INT Not supported Not supported Supported Supported
3a PEOU → PU Supported Supported Supported Supported
4 PBC → INT Supported Supported Supported Supported
4a FC → PBC Supported Not supported Not supported Not supported
4b PEOU → PBC Supported Supported Supported Supported
4c PNK → PBC Supported Supported Supported Supported
5 PR → INT Supported Supported Supported Not supported
5a T → PR Supported Supported Supported Supported
6 SN → INT Not supported Not supported Supported Not supported
7 RC →INT Supported Supported Not supported Supported
their intentions to continuously use; and iv) social norm influences only closeness as a factor in developing their intentions to initially adopt or
users in their intentions to continuously use. continuously use e-government services.
During the pre-adoption period in the rural municipality (Emmen), Third, this paper contributes to identifying the roles of inhibiting fac-
non-users care about facilitating conditions in developing perceptions tors in the e-government context. Both inhibiting factors (perceived risk
of behavioral control, and also care about resistance to change in ini- and resistance to change) play important roles for all non-users in de-
tially adopting e-government services. However, perceived facilitating veloping their intentions to initially adopt e-government services. This
conditions and resistance to change become not important once people indicates that non-users tend to have fear about adopting e-
adopt e-government services, and then their concerns move to per- government services. Once people adopt them, however, perceived
ceived ease of use and social norms in developing their intentions to risk decreases for rural users and disappears for urban users. However,
continuously use e-government services. resistance to change becomes the most influential determinant only
Fourth, comparing non-users to users in the urban municipality for urban users to continuously use e-government services. According
(Groningen), there are only two differences: i) perceived risk influences to the status quo bias perspective (Samuelson & Zeckhauser, 1988;
only non-users in developing their intentions to initially adopt e- Kim & Kankanhalli, 2009), it can be explained that users develop inten-
government services; and ii) perceived ease of use influences only tions not to continuously use e-government services if they have to
users for their intentions to continuously use. change the way they interact with their municipality as a result.
During the pre-adoption period in the urban area, non-users worry In conclusion, this paper proposes a model to assess citizens' inten-
about the risk involved in engaging e-government services, but this tions to adopt or use e-government services based on the literature on
worry seems to disappear once they adopt the services. Then, ease of IS acceptance, resistance, and e-government. The model is applied to
use appears for users as a new factor in developing their intentions to four different groups divided by usage and spatial criteria (at the pre-
continuously use the e-government services. adoption versus post-adoption period and in an urban versus rural
To summarize the comparisons of non-users (pre-adoption) versus area) and the results are compared. This paper confirms that enabling
users (post-adoption) and urban versus rural municipality, rural resi- factors such as perceived usefulness and perceived behavioral control con-
dents tend to change their behavioral attitudes more than urban resi- sistently play important roles across the four groups in influencing people's
dents between pre-adoption and post-adoption periods, confirming intentions to adopt or use e-government services. However, geographical
the finding of LaRose et al. (2007) that the experience of rural users in- closeness, subjective norm and inhibiting factors such as perceived risk
fluences the continuous usage of IS. In addition, these changed attitudes and resistance to change do not show consistent roles across the four
of rural residents do not evolve toward, but remain different from those groups. This implies that these factors need to be researched further.
of urban residents. Additional research could also be conducted in a longitudinal study
This research also has some contributions. First, this research con- to investigate behavioral attitudes over time, which this research as a
tributes to an extension of e-government literature by focusing on citi- cross-sectional study does not do. In this way, researchers can under-
zens' perceptions of the e-services that their municipalities provide. In stand how some non-users become initial adopters, how some adopters
the e-government context relevant to the relationship between a gov- become permanent users, and why others never become users.
ernment and citizens, this paper provides a model with strong predic- Another limitation of this research is that it is based on two munici-
tive power. palities in The Netherlands. We also acknowledge that there is potential
Second, this study compares not only non-users with users, but also bias in the data collection method. For example, the distribution of hard
those in a rural municipality with those in an urban municipality. As copy surveys at city halls and various neighborhoods could create bias
mentioned, a government has monopolistic power in dealing with citi- against people who do not go to city halls and particular areas. The on-
zens. Therefore, it is in the government's interest to understand their line distribution of the survey could also create bias against people who
citizens and persuade them to initially adopt and continuously use its do not use email or social media. Disproportional responses between
e-government services. This paper reveals that we need to consider offline and online surveys could create bias, even though we did not
characteristics of each municipality and its citizens in researching e- find any difference between them. We think that the findings are appli-
government related subjects. For example, one factor that distinguishes cable for e-government services provided by many European local gov-
people in a rural municipality from those in an urban municipality is ernments, but may not be applicable for regions governed by a strong
geographical closeness. Only people in rural areas consider geographical central government.
Appendix A. Measurement items and sources

Municipality eServices are defined as e-government services provided by respondent's municipality at the beginning of the survey. In addition, ex-
amples of municipality eServices are provided to avoid the vagueness of the term. The series 1 to 3 are indicators for users and the series 4 to 6 are
indicators for non-users. Indicator FC3 (FC6) is a reversed item.
Item Description
Users (non-users): intention to use (Venkatesh et al., 2003)

INT1 (INT4) I intend to use (adopt) the municipality eServices in the next 12 months
INT2 (INT5) I predict I will use (adopt) the municipality eServices in the next 12 months.
INT3 (INT6) Next time I have business with my municipality, I intend to use (adopt) municipality eServices.
Users (non-users): perceived usefulness (Davis, 1989)

PU1 (PU4) Using municipality eServices enabled (would enable) me to accomplish tasks more quickly.
PU2 (PU5) The eServices offered online by my municipality were (would be) useful to me.
PU3 (PU6) I found (would find) using municipality eServices useful.
Users (non-users): perceived ease of use (Venkatesh et al., 2003)

PEOU1 (PEOU4) (I expect that) The use of municipality eServices was (would be) clear and understandable.
PEOU2 (PEOU5) I found (expect to find) municipality eServices easy to use.
PEOU3 (PEOU6) Learning to operate municipality eServices was (would be) easy for me.
Users (non-users): perceived behavioral control (Taylor & Todd, 1995)

PBC1 (PBC4) I was (would be) able to use municipality eServices.
(continued)
Item Description
PBC2 (PBC5) Using municipality eServices was (would be) entirely within my control.
PBC3 (PBC6) I had (would have) the facilities, the knowledge and the ability to use municipality eServices.
Users (non-users): perceived necessary knowledge (Newly developed items)

PNK1 (PNK4) I had (expect to have) the knowledge necessary to follow the procedures of municipality eServices.
PNK2 (PNK5) I had (expect to have) the knowledge to interact through municipality eServices.
PNK3 (PNK6) I had (expect to have) the knowledge necessary to understand the underlying procedures and mentioned terminology in the municipality eServices.
Users (non-users): perceived facilitating conditions (Venkatesh et al., 2003)

FC1 (PFC4) I had (expect to have) the facilities necessary to use municipality eServices.
FC2 (PFC5) It was (would be) easy for me to get access to facilities to use municipality eServices.
FC3 (PFC6) I had (expect to have) facility difficulties in attempting to use municipality eServices. (reversed)
Users (non-users): trust (Bhattacherjee, 2002)

T1 (T4) My municipality kept (would keep) my best interest in mind during transactions and interactions through municipality eServices.
T2 (T5) I trusted (trust) the municipality to secure my data shared through municipality eServices.
T3 (T6) Overall, I had (have) trust in the integrity of municipality eServices.
Users (non-users): perceived risk (Malhotra et al., 2004)

PR1 (PR4) There was (would be) too much risk using municipality eServices.
PR2 (PR5) Using municipality eServices was (would be) risky.
PR3 (PR6) There was (would be) too much uncertainty using municipality eServices.
Users (non-users): resistance to change (Bhattacherjee & Hikmet, 2007)

RC1 (RC4) I did not (do not) want municipality eServices to change the way I do business with my municipality.
RC2 (RC5) I did not (do not) want municipality eServices to change the way I interact with my municipality.
RC3 (RC6) Overall, I did not (do not) want municipality eServices to change the way I currently deal with my municipality.
Users (non-users): subjective norm (Thompson, Higgins, & Howell, 1994; Venkatesh et al., 2003)
SN1 (SN4) People who were (are) important to me thought (think) that I had to (should) use municipality eServices.
SN2 (SN5) I used (would use) municipality eServices because (if) friends or family recommended (recommend) the use to me.
SN3 (SN6) My direct environment is (would be) positive about the use of municipality eServices.
Appendix B. Multiple comparisons
Dependent variable Variable group (I) Control group Mean difference Std. error Sig. 95% confidence interval
(J) (I–J)
Lower bound Upper bound
Perceived facilitating conditions Emmen non-users Groningen non-users −0.1520319 0.150724151 0.608 −0.50648424 0.20242052
Groningen users Groningen non-users −0.9596005a 0.160493954 0.000 −1.33702818 −0.58217285
Emmen non-users Emmen users 0.42413751a 0.152660132 0.016 0.06558875 0.78268627
Groningen users Emmen users −0.3834311a 0.162313448 0.048 −0.76465239 −0.00220989
Intention to use Emmen non-users Groningen non-users −0.57a 0.190 0.009 −1.01 −0.12
Groningen users Emmen users 0.10 0.205 0.916 −0.38 0.59
Perceived behavioral control Emmen non-users Groningen non-users −0.37658759 0.178211583 0.089 −0.79568114 0.04250597
Groningen users Emmen users −0.11895083 0.191914410 0.861 −0.56969506 0.33179341
Perceived ease of use Emmen non-users Groningen non-users −0.0209663 0.18249387 0.999 −0.4501304 0.4081977
Perceived risk Emmen non-users Groningen non-users 0.38920923 0.194009216 0.112 −0.06703503 0.84545349
Groningen users Groningen non-users 1.99439924a 0.206584718 0.000 1.50858164 2.48021684
Emmen non-users Emmen users −1.1686748a 0.196501172 0.000 −1.63019185 −0.70715777
Groningen users Emmen users 0.43651520 0.208926735 0.093 −0.05418542 0.92721583
Perceived necessary knowledge Emmen non-users Groningen non-users −0.19148501 0.198944516 0.641 −0.65933543 0.27636542
Perceived usefulness Emmen non-users Groningen non-users −0.10541232 0.153728258 0.826 −0.46692935 0.25610471
Resistance to change Emmen non-users Groningen non-users 0.66a 0.215 0.007 0.15 1.16
Groningen users Groningen non-users 1.75a 0.229 0.000 1.21 2.29
Emmen non-users Emmen users −1.36a 0.218 0.000 −1.87 −0.84
Subjective norm Emmen non-users Groningen non-users −0.10239989 0.172442725 0.877 −0.50792704 0.30312725
Groningen users Groningen non-users 0.03508964 0.183620306 0.995 −0.39672340 0.46690269
Emmen non-users Emmen users −0.17853480 0.174657670 0.595 −0.58874859 0.23167899
(continued on next page)
(continued)
Dependent variable Variable group (I) Control group Mean difference Std. error Sig. 95% confidence interval
(J) (I–J)
Lower bound Upper bound

Trust Emmen non-users Groningen non-users 0.10925508 0.143169396 0.777 −0.22743107 0.44594123
Emmen non-users Emmen users 0.15667971 0.145008339 0.554 −0.18389750 0.49725691
Geographical closeness Emmen non-users Groningen non-users 1.00a 0.155 0.000 0.63 1.36
Groningen users Groningen non-users −0.11 0.165 0.845 −0.50 0.28
Emmen non-users Emmen users −0.21 0.157 0.397 −0.58 0.16
Based on observed means. The error term is Mean Square (Error) = 1.057.
a
The mean difference is significant at the .05 level.
Appendix C. Factor loadings
Item UE UG NE NG
Calculated FC3/FC6 DROPPED DROPPED 0.74356 0.845185

FC1/FC4 0.946485 0.94187 0.89583 0.945187
FC2/FC5 0.905456 0.890369 0.830749 0.98013
INT1/INT4 0.949042 0.951508 0.94894 0.967794
INT2/INT5 0.976809 0.919511 0.968508 0.985383
INT3/INT6 0.954354 0.898592 0.904962 0.933483
PBC1/PBC4 0.886901 0.973872 0.894136 0.944049
PBC2/PBC5 0.872328 0.953255 0.896632 0.918451
PBC3/PBC6 0.931111 0.95182 0.909733 0.906591
PEOU1/PEOU4 0.855226 0.924707 0.777317 0.946556
PEOU2/PEOU5 0.910628 0.96075 0.910366 0.968305
PEOU3/PEOU6 0.899261 0.939578 0.832066 0.976871
PR1/PR4 0.895245 0.914786 0.85817 0.959889
PR2/PR5 0.915658 0.942799 0.896521 0.931555
PR3/PR6 0.825232 0.941301 0.817626 DROPPED
PNK1/PNK4 0.880491 0.956665 0.959029 0.937468
PNK2/PNK5 0.902789 0.946246 0.958885 0.926062
PNK3/PNK6 0.919988 0.923052 0.929135 0.911707
PU1/PU4 0.902656 0.912088 DROPPED 0.851173
PU2/PU5 0.925093 0.83577 0.921277 0.802383
PU3/PU6 0.87439 0.874858 0.883099 0.859031
RC1/RC4 0.804225 0.911098 0.753447 0.908101
RC2/RC5 0.837717 0.957627 0.899759 0.956016
RC3/RC6 0.788492 0.930632 0.814548 0.94774
SN1/SN4 0.784827 0.923998 0.848019 0.92876
SN2/SN5 0.920348 0.745645 0.869275 0.940744
SN3/SN6 0.825715 0.884726 0.891665 0.714491
T1/T4 DROPPED 0.744358 0.829282 0.744007
T2/T5 0.777484 0.923569 0.917887 0.830958
T3/T6 0.955758 0.928544 0.933535 0.894121
References Bhattacherjee, A. (2002). Individual trust in online firms: Scale development and initial
test. Journal of Management Information Systems, 19(3), 211–241.
Adams, J. S. (1963). Towards an understanding of inequity. Journal of Abnormal and Bhattacherjee, A., & Hikmet, N. (2007). Physicians' resistance toward healthcare informa-
Normal Social Psychology, 67(1), 422–436. tion technologies: A theoretical model and empirical test. European Journal of Infor-
Adams, J. S. (1965). Inequity in social exchange. In L. Berkowitz (Ed.), Advances in exper- mation Systems, 16(6), 725–737.
imental psychology (pp. 267–299). New York: Academic Press. Carter, L., & Bélanger, F. (2005). The utilization of e-government services: Citizen trust, in-
Ajzen, I. (1988). Attitudes, personality, & behavior. Chicago, IL: The Dorsey Press. novation and acceptance factors. Information Systems Journal, 15(1), 5–25.
Ajzen, I. (1991). The theory of planned behavior. Organizational Behavior and Human Chang, I., Li, Y., Hung, W., & Hwang, H. (2005). An empirical study on the impact of quality
Decision Processes, 50(2), 179–211. antecedents on tax payers' acceptance of internet tax-filing systems. Government
Anthopoulos, L. G., Siozos, P., & Tsoukalas, I. A. (2007). Applying participatory Information Quarterly, 22(3), 389–410.
design and collaboration in digital public services for discovering and re- Chin, W. W. (1998). The partial least squares approach to structural equation modeling. In
designing e-Government services. Government Information Quarterly, 24(2), G. A. Marcoulides (Ed.), Modern methods for business research (pp. 295–336).
353–376. Mahwah: Lawrence Erlbaum Associates, Inc.
Bandura, A. (1982). Self-efficacy mechanism in human agency. The American Psychologist, Compeau, D. R., & Higgins, C. A. (1995). Application of social cognitive theory to training
37(2), 122–147. for computer skills. Information Systems Research, 6(2), 118–143.
Bandura, A. (1986). Social foundations of thought and action: A social cognitive theory. En- Cronbach, L. J. (1984). Essentials of psychological testing. New York: Harper and
glewood Cliffs, NJ: Prentice-Hall. Row.
Beckmann, J., & Heckhausen, H. (2008). Motivation as a function of expectancy and Culnan, M. J., & Armstrong, P. K. (1999). Information privacy concerns, procedural fairness
incentive. In J. Heckhausen, & H. Heckhausen (Eds.), Motivation and action and impersonal trust: An empirical investigation. Organization Science, 10(1),
(pp. 275–299). Cambridge: Cambridge University Press. 104–115.
Bélanger, F., & Carter, L. (2008). Trust and risk in e-government adoption. The Journal of Davis, F. D. (1989). Perceived usefulness, perceived ease of use and user acceptance of in-
Strategic Information Systems, 17(2), 165–176. formation technology. MIS Quarterly, 13(3), 319–340.
De Ruyter, K., Wetzels, M., & Kleijnen, M. (2001). Customer adoption of e-service: An ex- Pavlou, P. A. (2003). Consumer acceptance of electronic commerce — Integrating trust
perimental study. International Journal of Service Industry Management, 12(2), and risk with the technology acceptance model. International Journal of Electronic
184–207. Commerce, 7(3), 101–134.
Gemeente Emmen. (2010). Emmen in cijfers, The Netherlands. Retrieved July 23, 2014, Peronard, J., & Just, F. (2011). User motivation for broadband: A rural Danish study.
from http://www.emmen.nl/fileadmin/files/www.emmen.nl/Tekstpaginas/Over_de_ Telecommunications Policy, 35(8), 691–701.
gemeente/Emmen_in_cijfers/Emmen_in_cijfers_uitgave_2010/Emmen_ Ramayah, T., Ma'ruf, J. J., Jantan, M., & Mohamad, O. (2002). Technology acceptance
jaaroverzicht_2010.pdf. model: Is it applicable to users and non-users of internet banking? Proceedings of
Fornell, C., & Larcker, D. F. (1981). Evaluating structural equations with unobservable var- The International Seminar. Banda Aceh, Indonesia.
iables and measurement error. Journal of Marketing Research, 18(1), 39–50. Riemenschneider, C. K., Harrison, D. A., & Mykytyn, P. P. (2003). Understanding it adop-
Gefen, D. (2000). E-commerce: The role of familiarity and trust. Omega: The International tion decisions in small business: Integrating current theories. Information
Journal of Management Science, 28(6), 725–737. Management, 40(4), 269–285.
Gefen, D. (2002). Reflections on the dimensions of trust and trustworthiness among on- Rotter, J. B. (1967). A new scale for the measurement of interpersonal trust. Journal of
line consumers. ACM SIGMIS Database, 33(3), 38–53. Personality, 35(4), 651–665.
Gefen, D., Karahanna, E., & Straub, D. (2003). Trust and TAM in online shopping: An inte- Samuelson, W., & Zeckhauser, R. J. (1988). Status quo bias in decision making. Journal of
grated model. MIS Quarterly, 27(1), 51–90. Risk and Uncertainty, 1(1), 7–59.
Gilbert, E., Karahalios, K., & Sandvig, C. (2010). The network in the garden: Designing so- Schneider, K., & Schmalt, H. D. (2000). Motivation. Stuttgart: Kohlhammer.
cial media for rural life. The American Behavioral Scientist, 53(9), 1367–1388. Schüler, J. (2010). Achievement incentives determine the effects of achievement-motive
Gouscos, D., Kalikakis, M., Legal, M., & Papadolpoulou, S. (2007). A general model of per- incongruence on flow experience. Motivation and Emotion, 34(1), 2–14.
formance and quality for one-stop e-Government service offerings. Government Seo, D., Boonstra, A., & Van Offenbeek, M. (2011). Managing IS adoption among ambiva-
Information Quarterly, 24(4), 860–885. lent groups. Communications of the ACM, 54(11), 68–73.
Gemeente Groningen. (2010). Statistisch Jaarboek 2010, The Netherlands. Retrieved July Straub, D., Keil, M., & Brenner, W. (1997). Testing the technology acceptance model across
23, 2014, from http://os-groningen.nl/images/stories/rapport/Statistisch_Jaarboek_ cultures: A three country study. Information Management, 33(1), 1–11.
2010.pdf. Taylor, S., & Todd, P. A. (1995). Assessing IT usage: The role of prior experience. MIS
Heeks, R. B., & Bailur, S. (2007). Analyzing e-Government research: Perspectives, philoso- Quarterly, 19(2), 561–570.
phies, theories, methods and practice. Government Information Quarterly, 24(2), Teo, T. S. H., Srivastava, S. C., & Jiang, L. (2009). Trust and electronic government success:
243–265. An empirical study. Journal of Management Information Systems, 25(3), 99–131.
Hofferth, S., & Iceland, J. (1998). Social capital in rural and urban communities. Rural Thompson, R. L., Higgins, C. A., & Howell, J. M. (1994). Influence of experience on personal
Sociology, 63(4), 574–598. computer utilization: Testing a conceptual model. Journal of Management Information
Hoffman, D. L., Novak, T. P., & Peralta, M. A. (1999). Building consumer trust online. Systems, 11(1), 167–187.
Communications of the ACM, 42(4), 80–85. Groningen Toerisme. (2010). Facts & Figures Groningen, The Netherlands. Retrieved July
Hollifield, C. A., & Donnermeyer, J. F. (2003). Creating demand: Influencing information 23, 2014, from http://toerisme.groningen.nl/over-groningen/stad-groningen/facts-
technology diffusion in rural communities. Government Information Quarterly, 20(2), figures.
135–150. Tseng, F. M., & Lo, H. Y. (2011). Antecedents of consumers' intentions to upgrade their
Horst, M., Kuttschreuter, M., & Gutteling, J. M. (2007). Perceived usefulness, personal ex- mobile phones. Telecommunications Policy, 35(1), 74–86.
periences, risk perception and trust as determinants of adoption of e-government Van Deursen, A., Van Dijk, J., & Ebbers, W. (2006). Why e-Government usage lags behind:
services in The Netherlands. Computers in Human Behavior, 23(4), 1838–1852. Explaining the gap between potential and actual usage of electronic public services in
Hsu, M. H., & Chiu, C. M. (2004). Predicting electronic service continuance with a The Netherlands. In M. A. Wimmer, H. J. Scholl, Å. Grönlund, & K. V. Andersen (Eds.),
decomposed theory of planned behaviour. Behaviour & Information Technology, 23 EGOV 2006 LNCS, 4084. (pp. 269–280). Heidelberg: Springer.
(5), 359–373. Van Dijk, J. A. G. M., Peters, O., & Ebbers, W. (2008). Explaining the acceptance and use of
Hung, S. Y., Chang, C. M., & Yu, T. J. (2006). Determinants of user acceptance of the e- government internet services: A multivariate analysis of 2006 survey data in the
Government services: The case of online tax filing and payment system. Netherlands. Government Information Quarterly, 25(3), 379–399.
Government Information Quarterly, 23(1), 97–122. Venkatesh, V., Morris, M. G., Davis, G. B., & Davis, F. D. (2003). User acceptance of informa-
Karahanna, E., Straub, D. W., & Chervany, N. L. (1999). Information technology adoption tion technology: Toward a unified view. MIS Quarterly, 27(3), 425–478.
across time: A cross-sectional comparison of pre-adoption and post-adoption beliefs. Walster, R., Walster, G., & Berschied, E. (1978). Equity: Theory and Research. Needham
MIS Quarterly, 23(2), 183–213. Heights: Allyn and Bacon, Inc.
Kim, H. W., & Kankanhalli, A. (2009). Investigating user resistance to information systems Wang, Y. D., & Emurian, H. H. (2005). An overview of online trust: Concepts, elements and
implementation: A status quo bias perspective. MIS Quarterly, 33(3), 567–582. implications. Computers in Human Behavior, 21(1), 105–125.
Lapointe, L., & Rivard, S. (2005). A multilevel model of resistance to information technol- Wang, Y. S., & Liao, Y. W. (2008). Assessing e-government systems success: A validation of
ogy implementation. MIS Quarterly, 29(3), 461–491. the DeLone and McLean model of information system success. Government
LaRose, R., Gregg, J. L., Strover, S., Straubhaar, J., & Carpenter, S. (2007). Closing the rural Information Quarterly, 25(4), 717–733.
broadband gap: Promoting adoption of the internet in rural America. Warkentin, M., Gefen, D., Pavlou, P. A., & Rose, G. M. (2002). Encouraging citizen adoption
Telecommunications Policy, 31(6–7), 359–373. of e-government by building trust. Electronic Markets, 12(3), 157–162.
Lee, H. G. (1998). Do electronic marketplaces lower the price of goods? Communications of Yaghoubi, N. H., Kord, B., & Shakeri, R. (2010). E-Government services and user accep-
the ACM, 41(1), 73–80. tance: The unified models' perspective. European Journal of Economics, Finance and
Lee, M. K. O., & Turban, E. (2001). A trust model for consumer internet shopping. Administrative Sciences, 24, 36–49.
International Journal of Electronic Commerce, 6(1), 75–92.
Lehmann, D. R., & Hulbert, J. (1972). Are three-point scales always good enough? Journal
DONGBACK SEO earned her Doctor of Philosophy and Masters of Science in Management
of Marketing Research, 9(4), 444–464.
Information Systems from the University of Illinois at Chicago. Prior to pursuing the Ph.D.
Lewis, J. D., & Weigert, A. (1985). Trust as a social reality. Social Forces, 63(4), 967–985.
program, she worked as a software engineer in a wireless communications firm and as a
Luhmann, N. (1979). Trust and power. London, England: John Wiley.
small business owner. Her publications include three books, recently published Evolution
Malhotra, N. K., Kim, S. S., & Agarwal, J. (2004). Internet users' information privacy con-
and Standardization of Mobile Communications Technology, as well as a class manual and
cerns (IUIPC): The construct, the scale, and a causal model. Information Systems
several chapters. Her papers have been published in many journals (e.g., Communications
Research, 15(4), 336–355.
of the ACM, European Journal of Information Systems, Telecommunications Policy, etc.)
Martinko, M. J., Henry, J. W., & Zmud, R. W. (1996). An attributional explanation of indi-
and conference proceedings (e.g., International Conference on Information Systems,
vidual resistance to the introduction of information technologies in the workplace.
European Conference on Information Systems, etc.). Her research interests include the
Behaviour & Information Technology, 15(5), 313–330.
areas of adoption of IT/IS-enabled services from the perspectives of individuals and
Mathieson, K. (1991). Predicting user intentions: Comparing the technology acceptance
organizations, organizational standards strategy, business convergence, mobile com-
model with the theory of planned behavior. Information Systems Research, 2(3),
merce, and analysis of competitive dynamics in rapidly changing industries. She is a
173–191.
currently associate professor at the department of Management Information Systems,
Mayer, R. C., Davis, J. H., & Schoorman, F. D. (1995). An integrative model of organizational
Chungbuk National University.
trust. The Academy of Management Review, 20(3), 709–734.
McClelland, D. C. (1985). Human motivation. Glenview: Scott Foresman.
MICHEL BERNSEN was a graduate student in the Department of Business & ICT at the
McKnight, D. H., & Chervany, N. L. (2001). What trust means in e-commerce customer re-
Faculty of Economics and Business, University of Groningen, the Netherlands. He received
lationships: An interdisciplinary conceptual typology. International Journal of
a Master of Science in Business and ICT from the University of Groningen, the Netherlands
Electronic Commerce, 6(2), 35–39.
in 2011. His research interests include e-government, adoption and diffusion of IT, IT
Muthitcharoen, A., Palvia, P. C., & Grover, V. (2011). Building a model of technology pref-
strategy, and human-computer interaction. He is currently manager of the IT Risk and
erence: The case of channel choices. Decision Sciences Journal, 42(1), 205–237.
Assurance practice of Ernst & Young in Amsterdam, the Netherlands.
Overheidsmonitor (2011). Overheid.nl Monitor|Ranglijsten, The Netherlands. Retrieved
April 13, 2011, from http://monitor.overheid.nl/ranglijsten?organisatietype=
gemeente.

Government Information Quarterly: Dongback Seo, Michel Bernsen

Uploaded by

Copyright:

Available Formats

Government Information Quarterly: Dongback Seo, Michel Bernsen

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Government Information Quarterly: Dongback Seo, Michel Bernsen

Uploaded by

Copyright:

Available Formats

GOVINF-01154; No.

of pages: 13; 4C:

Contents lists available at ScienceDirect

Government Information Quarterly

journal homepage: www.elsevier.com/locate/govinf

Comparing attitudes toward e-government of non-users versus users in a

1. Introduction (municipality) government (e.g., applying and regularly renewing a

H5a. A citizen's trust in e-government services has a negative relation-

H6. Subjective norm has a positive relationship with one's intention to

Table 1 3.4. Field survey

The survey was qualitatively and quantitatively tested on a small

FC INT PBC PEOU PR PNK PU RC SN T

FC INT PBC PEOU PR PNK PU RC SN T

1 PU → INT Supported Supported Supported Supported

Appendix A. Measurement items and sources

Users (non-users): intention to use (Venkatesh et al., 2003)

Users (non-users): perceived usefulness (Davis, 1989)

Users (non-users): perceived ease of use (Venkatesh et al., 2003)

Users (non-users): perceived behavioral control (Taylor & Todd, 1995)

Users (non-users): perceived necessary knowledge (Newly developed items)

Users (non-users): perceived facilitating conditions (Venkatesh et al., 2003)

Users (non-users): trust (Bhattacherjee, 2002)

Users (non-users): perceived risk (Malhotra et al., 2004)

Users (non-users): resistance to change (Bhattacherjee & Hikmet, 2007)

Appendix B. Multiple comparisons

(continued on next page)

Groningen users Emmen users −0.04104526 0.185701980 0.992 −0.47719853 0.39510800

Appendix C. Factor loadings

Calculated FC3/FC6 DROPPED DROPPED 0.74356 0.845185

You might also like