Connect, Talk, Engage
Citizen’s Engagement on Social Networking Sites
- in co-operation with Trafikverket (Swedish Transport Administration) -
Master’s thesis within BA
Author: Andreea Irinca & Rasa Strumskyte
Tutor: Tomas Müllern
Aknowledgement
We would have never had the chance to enjoy so much the thesis writing without the sup-port of Trafikverket, teachers, thesis group members and friends.
We would like to thank our supervisor, Tomas Müllern, who encouraged us to keep a high level of ambition and motivated us through a challenging research process. We would like to express the warmest gratitude to Adele Berndt for her constant willingness to help and the valuable consultancy regarding the statistical data analysis.
Many thanks to Sara Langgren and Alexandra Trollberg from Trafikverket for helping us collect the empirical data for the research, for the time spent discussing relevant questions and the most important – the opportunity to work for Trafikverket.
We also give a lot of thanks to our friends Moritz Justus Klein and Per Adolfsson who were always cheering us up and supporting us with valuable advices regarding the research paper.
Master’s Thesis in Business Administration
Title: Citizens’ Engagement on Social Networking Sites Author: Andreea Irinca & Rasa Strumskyte
Tutor: Tomas Müllern
Date: 2012-05-14
Subject terms: Social Media, Enagement, Trafikverket
Abstract
Background
Social Media is a communication tool employed nowadays not
only by multinational or middle-size and small companies, but also by non-profit organizations and governmental institutions. Today, more than 60% of the Swedish municipalities are participating on Social Media platforms with the purpose of engaging with the citizens. Online Engagement is not radically different from traditional Engagement, just as Citizen Engagement has close similarities with Customer Engagement. All of the Engagement facets are comparatively discussed in the current research paper, and four Dimensions, each including several factors that might differentiate Engaged and Unengaged users are described and used in the empirical study, in order to fulfill the purpose of the research.
Purpose
This research paper investigates the factors that differentiate
the users from different stages of engagment on public institution’s social media profiles.
Method
The factors differentiating users from different stages of
Engagement were identified by using relevant theories and previous research, and by conducting a quantitative study on Trafikverket, Swedish Transport Administration, Facebook profile followers.
Conclusion Five factors were tested throughout the statistical data
analysis. Three of them were found to be signifcantly different for users that are engaged within the profile and users that are not engaged. These factors are Usefulness, Privacy & Security and Communicaion & Relationship. Based on the findings and literature review, several suggestions for engaging with citizens, were proposed.
Table of Contents
1 Introduction ... 6
1.1 Problem definition ... 8 1.2 Purpose ... 8 1.3 Research Questions ... 8 1.4 Glossary of Terms ... 9The Paper at a Glance ... 10
2
Theoretical framework ... 11
2.1 Engagement definition ... 11
2.2 Citizen Engagement ... 12
2.3 Citizen Engagement Dimensions ... 13
2.4 Technology Acceptance Model ... 16
2.5 Social Networking Sites ... 18
2.5.1 Engagement in Terms of Online Behavior ... 18
2.5.2 Types of Social Media Behavior ... 19
2.6 Research model and Hypotheses ... 20
3
Method ... 23
3.1 Research Approach ... 23
3.2 Research Strategy ... 24
3.2.1 Sampling ... 24
3.2.2 Survey ... 25
Other methods considered ... 26
3.3 Pilot testing ... 26
3.4 Questionnaire design ... 27
3.5 Reliability and Validity ... 31
3.6 Generalization and Limitations ... 31
4
Data analysis ... 33
4.1 Overview of Respondents’ Characteristics ... 33
4.2 Identifying the Factors ... 35
4.3 Hypotheses Testing ... 37
4.3.1 Hypothesis 1: Privacy and Security ... 37
4.3.2 Hypothesis 2: Information Quality ... 38
4.3.3 Hypothesis 3: Relationship & Communication ... 38
4.3.4 Hypothesis 4: Community Capability & Inclusiveness ... 39
4.3.5 Summary of Hypotheses ... 40
4.4 Behavior on Facebook ... 40
5
Analysis in Accordance with the Theoretical Framework ... 42
5.1 Engagement Dimensions ... 42
5.2 Final Research Model ... 46
6
Discussion ... 48
6.1 Review of Purpose and Findings ... 48
6.2 Implications for Practice ... 49
6.3 Delimitations and Suggestions for Future Research ... 50
List of Figures
Figure 1-1 - Screenshot of Governmental Institutions Websites ... 6
Figure 2-1 - Citizen Engagement Stages (OECD, 2003) ... 12
Figure 2-2 -Technology Acceptance Model (Davis, 1985) ... 17
Figure 2-3 - Social Technographics (compiled by the authors) ... 19
Figure 2-4 – Proposed Research Model (compiled by the authors) ... 22
Figure 3-1 - Age Groups - Trafikverket Followers vs. Respondents (compiled by the authors) ... 32
Figure 4-1 – Respondents’ Demographics (compiled by the authors) ... 33
Figure 4-2 - Facebook usage vs. Trafikverket profile usage (compiled by the authors) .. 34
Figure 4-3 - Trafikverket Facebook profile - Participation and Reasons for Usage (compiled by the authors) ... 34
Figure 4-4 - Factor Analysis Results (compiled by the authors) ... 35
Figure 4-5 - Cronbach's Alpha Coefficient (compiled by the authors) ... 37
Figure 4-6 - Paired Sample T-test for Facebook Behavior (compiled by the authors) .... 41
Figure 5-1 - Final Research Model (compiled by the authors) ... 46
List of Tables
Table 2-1 – Engagement Dimensions (compiled by the authors) ... 15Table 3-1 – Questionnaire Design (compiled by the authors) ... 30
Table 4-1 - Independent Sample T-test - Privacy and Security (compiled by the authors) ... 37
Table 4-2 - Independent Sample T-test - Information Quality (compiled by the authors) ... 38
Table 4-3 - Independent Sample T-test - Relationship & Communication (compiled by the authors) ... 38
Table 4-4 - Independent Sample T-test - Ease of Use (compiled by the authors) ... 39
Table 4-5 - Independent Sample T-test – Usefulness (compiled by the authors) ... 40
Table 4-6 - Summary of Hypotheses Status (compiled by the authors) ... 40
Table 4-7 - Independent Sample T-test - Regular Behavior on Facebook (compiled by the authors) ... 41
Table 4-8 - Independent Sample T-test - Desired Behavior on Facebook (compiled by the authors) ... 41
List of Appendices
Appendix 1 - Trafikverket Social Media Policies ... 59Appendix 2 - Awareness Report, Trafikverket, 2011 ... 64
Appendix 3 - Demographic Distribution of Trafikverket followers - March 2012 ... 65
Appendix 4 - Questionnaire ... 66
Appendix 5 - Factor Analysis ... 70
Appendix 6 - Paired Sample T-Test ... 73
Appendix 7 - Mean Values for 6-point Likert items ... 74
1 Introduction
The background of the current situation and the research problem is presented. The social media landscape in Sweden and the theoretical background are introduced.
The new social and interactive media that has emerged in recent years creates new oppor-tunities not only for multinational firms or small and medium sized companies, but also for non-profit organizations and government sector. Today, most of the European govern-ment institutions use the Internet and Social Networks as one of the most innovative communication tools (Hellman, 2011). In order to facilitate the communication, public en-tities willing to engage in a two-way interaction with citizens, businesses or other govern-ments, need to develop comprehensive communication models, which incorporate the lat-est technological developments.
In March 2010, the Swedish Government published a proposal (Reinfeldt & Odell, 2010), acknowledging the fact that social media tools create new possibilities of communication with citizens and businesses, and can engage and empower communities. At the end of the same year, Sweden’s E-Government Delegation produced a set of guidelines (E– delegationen, 2010) that are intended as a support for all governmental bodies’ social media activities, describing the procedures that public institutions should follow and the resources that should be allocated. The Swedish E-Government delegation was established in March 2009 with the view to lead and coordinate the development of e-Government, which is de-scribed as ‘the use of information and communication technology in public administrations combined with
organizational change and new skills in order to improve public services and democratic processes and strengthen support to public policies [COM, 2003, section 3]’.
Sweden’s use of social media in public institutions communication is of specific and gen-eral interest in at least three different ways. First, Sweden has the highest annual ranking in the world in terms of usage of telecommunications technologies such as networks, cell-phones and computers (Waverman, 2011). Second, today more Swedes use social media on a typical day than traditional online media, with 35% vs. 28% in 2010 (Kullin, 2011). Third-ly, Sweden is the third most advanced country in the world in terms of e-government de-velopment and the first one in Europe (Waseda University Institute of e-Government, 2011). Other reasons might include the effectiveness of the administrative organization or the legislation on freedom of access to official records, which is in Sweden one of the most generous in the world (Klang, 2011).
According to SKL (Sveriges Kommuner och Landsting, 2011) around 60% of the Swedish governmental institutions are currently participating on social media. Municipalities, agen-cies and administrations are signing on Facebook and Twitter and popping up logos on their websites.
Citizens’ connection with public institutions via the Internet could parallel a variety of citi-zen engagement initiatives as face-to-face surveys or public meetings, which are a common means for involving citizens. CLG (2008), the UK Governmental department for Commu-nities and Local Government, describe the participatory process of citizen engagement as follows: ‘the process whereby public bodies facilitate citizen and community participation in order to
incor-porate their views and needs into decision-making processes. This includes reaching out to communities to create empowerment opportunities’ (CLG, 2008, p.7). Citizen engagement is a challenging topic in
a century of declining trust in public institutions, so the government needs to constantly remodel the engagement methods (Lukensmeyer & Torres, 2006).
Currently, companies and organizations are seeking for effective channels to engage with online customers and cannot ignore the social networking phenomenon, Facebook, with more then 845 million monthly active users in the world. The same applies for governmen-tal institutions online communication.
According to Kaplan and Haenlein (2010), social media can be defined as ‘a group of In-ternet-based applications that allow the creation and exchange of user-generated content’ (Kaplan & Haenlein, 2010, p.61). Given its two-way communication nature, the ability to give a ‘voice to all’ and its immediacy, social media offers a new opportunity for public in-stitutions to engage with the citizens. Social Networking Sites can be engaging and empow-ering for citizens as they provide them with a platform to speak, allowing everyone to par-ticipate, share and achieve common goals. (Bertot et.al, 2010)
The online engagement process is a long-term activity that can be constantly improved by investigating and understanding the influencing factors and taking action according to them. Several stages of Online Engagement are proposed by different authors(OECD, 2003; Eastin et al., 2011; Rowe & Frewer, 2005; etc.) and dimensions for engagement on Social Media platforms are identified throughout research area in the field (Ide-Smith, 2010; OECD, 2003; etc.). In order to investigate how the underlying dimensions of en-gagement can be integrated within the concept of enen-gagement, the current research will be conducted in co-operation with Trafikverket, Swedish Transport Administration.
Trafikverket is responsible for the long-term planning of the transport system for road traf-fic, rail traftraf-fic, maritime shipping, and air traffic. The administration also takes responsibil-ity for construction, operation and maintenance of the state road network and national railway network. According to Trafikverket, the use of social media in the communication plan is ‘a way to show commitment to our questions, to show that we are listening’ (see APPENDIX 1, p. 1). Building relationships with external audiences is also one of the ad-ministration’s purposes of using the new media channel. Through its presence on social media, Trafikverket wants to be perceived as a modern authority with valuable responsiveness and a holistic approach.
The agreement of working with Trafikverket was taken after discussing with decision mak-ers from the Transport Administration. It has been identified that at least three of the char-acteristics of citizen engagement described by Sheedy (2008) match Trafikverket purposes of going on social media. The criteria are as follows:
• Involve citizens in policy or program development;
• Generate innovative ideas and active participation in the form of dialogue with the customers;
The match between Trafikverket’s objectives and the current research purpose will allow this cooperation to result in a ‘real-life’ insight into the phenomena of citizen engagement throughout public institutions communication on social media.
1.1 Problem definition
The task for authorities of establishing a presence within social media is slightly different compared to corporations and private individuals. While corporations can be quite flexible when working on social media, governing bodies are influenced by regulated norms that define an appropriate relationship between citizens and those contacting the authorities (Klang, 2011). This can result in low levels of engagement on social media platforms and poor amounts of citizen initiated dialogue. These are also some of the barriers that Traf-ikverket, Swedish Transport Administration, faces in their social media communication. The Trafikverket case is of particular interest since the engagement barriers might be fur-ther intensified by the fact that the administration was founded two years ago, so its aware-ness and citizen’s knowledge about the organization’s reasons of existence are limited, as previous research shows (see APPENDIX 2).
1.2 Purpose
This research aims at identifying the differentiating factors in terms of users engagement stage on Trafikverket social media profile. The results of this research will be used for Traf-ikverket’s improvement of the communication on currently used social networking sites and it will serve as a starting point for the administration to initiate the communication on their Twitter account.
1.3 Research Questions
In order to be able to reach the purpose of the research, some sub-questions have been de- fined as follows:
1. Are the risks related to Privacy and Security perceived differently by users from different engagement stages?
2. Is the content, provided by Trafikverket perceived differently by users from dif-ferent engagement stages?
3. Is the online interaction between Trafikverket and the citizens perceived differ-ently based on the engagement stage?
4. Is the social networking platform of Trafikverket accepted differently by users in different engagement stages?
1.4 Glossary of Terms
E-Government Engagement Dimension Factor Facebook Profile Behavior on Facebook Like Facebook Pages News FeedThe usage of information technologies that have the ability to trans-form the relations with citizens, businesses, and other governmental institutions;
The research area within the concept of engagement;
The attribute within the Engagement Dimensions; (timeline) a complete picture of oneself on Facebook;
like, comment, follow, post, share;
The way to give positive feedback for a specific content or to con-nect to a Page on Facebook;
The space where businesses, organizations and brands can connect with people;
The ongoing list of updates on home page that shows what is new with the friends and pages one follows;
Regular
Face-book Behavior Desired Face-book Behavior
Regular level of activities, such as joining groups, following news, and sometimes interacting with the source;
High interaction activity level. Characteristic for creators and critics on social networking sites
The Paper at a Glance
Chapter Description Key Words
1. Introduction
The background of the current situation and the research problem are presented. The theoretical background is introduced.
Governmental Institu-tion, e-Govenrment, So-cial Media Landscape in Sweden;
2. Theoretical Framework
The theoretical background that was used in this research paper in order to fulfill the purpose is presented. Engagement, Citizen Engagement, Technology Ac-ceptance Model, Social Networking Sites and Social Media user types are defined and discussed from different perspec-tives. The theories are used in the process of constructing the research model that serves as a theoretical foundation for the empirical study.
Engagement, Citizen Engagement, Engage-ment Stages, Technology Acceptance Model (TAM), Social Media, Social Networking Sites, Social Technographics, Dimensions of Engage-ment;
3. Methodology
This section of the paper presents the choice of the research design, approach and the method used in the research. The process of data collection and the way in which the survey was constructed will be presented. The reliability, validity and generalization of the current research are also discussed.
Deductive/Inductive, Descripto-Explanatory, Quantitative, Judgmental Sampling, Online
Sur-vey,
Struc-tured/Unstructured questions;
4. Data analysis
This section summarizes the results of the data analysis. An overview of the re-spondents is presented, and the factors that will be used for testing the hypothe-ses are identified. The section ends with the testing of the hypotheses and pre-senting the research findings in the area of behavior on social media.
Factor Analysis, Inde-pendent Sample T-Test, Cronbach’s Alpha;
5. Analysis in Accordance with
the Theoretical Framework
The findings of the research are dis-cussed from a theoretical perspective in terms of the relationships found to be significant and the possible explanations for the non-significant results.
Privacy & Security, Ease of Use, Usefulness, In-formation Quality, Rela-tionship & Communica-tion, Regular Facebook Behavior, Desired Face-book Behavior;
6. Discussion A summary of the main findings is
pro-vided. The limitations and suggestions for future research are presented. The chapter is closed with the implications for practice and a final conclusion.
Theoretical Limitations, Methodology Limita-tions, Future Research Suggestions, Conclusion.
2 Theoretical framework
This section presents the theoretical background used in this research paper in order to fulfill the research purpose. Engagement, Citizen Engagement, Technology Acceptance Model and Social Media user types will be discussed as theoretical foundations for the empirical study.
2.1 Engagement definition
The new technologies and possibility to join the Internet everywhere is not just a usable and convenient tool that helps content creators to be reached by online users, it is also a tool to engage users (O’Brien & Toms, 2008).
According to O’Brien & Toms (2008), despite the obvious need for users engagement, there is no agreement on a clear definition of the construct of engagement. Currently, many different engagement definitions and models can be found, but they are incomplete or need some changes in line with the constantly renewing online environment (O’ Brien, 2009).
A general definition of engagement is suggested by Attfield, Kazai, Lalmas & Piwowarski (2011, p. 2): ‘the emotional, cognitive and behavioral connection that exists, at any point in time and possibly over time, between a user and a resource.’ A more simplified definition is provided by O’Brien (2009, p. 2), who states that ‘Engagement has been defined as both the act of emotionally involving someone or the personal state of being in gear’. Looking at engagement from an even more narrow perspective, it has been described based on a set of factors: attention and motivation (Jacques, 1996), perceived control (Jacques, 1996) or the action that can be observed between the user and the system (Hutchins & Holland & Norman, 1986).
Most of the Engagement theories and methods can be easily adapted to the online envi-ronment, but according to Sheedy (2008), it requires careful planning, support and even creativity.
Maclean & O’Brien (2009) state that Online Engagement mirrors a quality of user experi-ences with a particular online platform. Quesenbery (2003) suggests that online engage-ment is dependent on the user’s first impression of the website/profile and satisfaction that comes from using it.
In academic literature, several theories related to the concept of Online Engagement can be found. The Flow Theory (Csikzentmihayi, 1988) is focused on the users’ state of concen-trating entirely on a particular online activity and losing awareness of the outside world. The Flow Theory is closely related to engagement, in terms of the factors that are found within the theory. These factors include: immediate feedback, potential control, personal skills, concentration or loss of self-consciousness (Csikzentmihayi, 1988; Choi & Kim & Kim, 2007). The last factor differentiates Flow Theory from Engagement, in a way that makes the Engagement a more passive state of Flow. The research areas of highly interac-tive online platforms, such as video games, are fields where Flow Theory becomes more appropriate than Online Engagement theory, since the level of concentration on these plat-forms is considered to be very high.
Another theory closely related to the Online Engagement, is The Information Interaction Theory (Ingwersen, 1992). The Information Interaction Theory focuses on the experience that users have based on the content provided, while engagement is an expression of that interactivity (O’Brien, 2008). The Information Interaction Theory leaves aside the focus on feedback quality or getting attention factor, thus differing from the Online Engagement in a way that the Information Interaction Theory is less complex. The theory thus becomes more appropriate for research in the area of regular websites or blogs, platforms that have as a main feature the content provided.
The concept of Online Engagement, instead of Flow Theory and Information Interaction Theory, has been chosen for the purpose of this study, due to the fact that it was consid-ered as being more appropriate for Social Media platforms, and due to its adaptation for different research areas (O’Brien, 2009), such as e-learning (Quinn, 2005), e-commerce (Kappelman, 1995), multimedia and games online (Said, 2004) and e-government engage-ment with citizens (OECD, 2003).
2.2 Citizen Engagement
The term citizen engagement is sometimes confused with other terms such as community or public involvement, so Sheedy (2008) explains that engagement is just one of the several approaches that fit within the concept of involvement. The author states that engagement emphasizes the sharing of information, providing opinion and showing respect between citizens and public institution.
The OECD (2003) developed a model (Figure 2-1 – Citizen Engagement Stages) for en-gaging citizens with governmental institutions through information communication tech-nologies. Citizens in this context take the form of individuals, communities and business. The three levels of citizen engagement proposed by OECD (2003) are described below and further explored in this paper through a comparison with other similar models suggested by different authors.
Figure 2-1 - Citizen Engagement Stages (OECD, 2003)
Stage 1 is called information stage because it is a simple one-way relationship. Generally,
online users are getting the information provided by governmental institutions.
Stage 2 is a two-way relationship and a longer communication process in which citizens
provide feedback on the information posted.
opinions and inputs they can shape governmental institutions’ decisions. Despite of that, government retains all the responsibility for the decision-making.
The models used for comparing the stages described by OECD (2003) are the Hierarchy of Engagement with Customers in the Online Environment (Eastin, Daugherty & Burns, 2011) and Engagement with Public Participation (Rowe & Frewer, 2005).
Eastin et al. (2011) look from the strategic perspective and propose that there are 4 levels of engagement as follows: one-way communication, two-way communication, service co-creation and community building. The 4th level, community building, is the deepest level of
engagement that shows active interaction between the customer and a company. This strong two-way communication forms an online community where people actively share experiences or opinions, discuss and give inputs for the company. According to Eastin et al. (2011) these four stages, applied to the ‘real world’, can be successfully adapted to the online environment, but the 3rd stage, service co-creation, is oriented towards e-commerce
customers and involving them in co-creating services together with the firm. This interac-tion is not in line with citizens and public instituinterac-tions communicainterac-tion online, so this stage is not relevant for the purpose of this research. Rowe & Frewer (2005) analyze several stag-es of engagement within the public participation and prstag-esent the three typstag-es of public en-gagement: public communication, public consultation and public participation.
In general, the models for engagement stages, describe the same communication types in a particular engagement stage as was suggested by OECD (2003). The models differ just to a small extent in terms of structure and are adapted for different purposes of being online, such as online purchase or online consultation.
2.3 Citizen Engagement Dimensions
According to Sheedy (2008), a challenging question for public institutions on Social Media platforms is not only how to reach citizens online, attract them, involve them into the Online communication or decision-making process, but also the consideration of how to adapt communication to different citizen types and avoid barriers/threats that can occur. In this case, the dimensions of engagement and the factors within the dimensions should be analyzed, and the importance of them evaluated.
Ide-Smith (2010) conducted a research in UK with the purpose to provide insights into the attitudes and provide understanding towards the usage of Social Media as a tool to engage with citizen communities. A qualitative study, considering the opinions of both councils and citizens, gave valuable findings that became a basis when formulating the engagement dimensions in terms of attitudes and perceptions. Although the sample size is rather small, the respondents represented four of the six main socio-demographic types of the area, thus increasing the reliability of the sampling procedure.
Ide-Smith (2010) identified six dimensions within engagement that might play an important role in the process of citizen engagement on social media platforms. The six dimensions are: Privacy, Trust & Confidence, Information Gathering, Managing Participation, Digital Exclusion and Usability. In order to conceptualize the dimensions within engagement, the factors identified by Ide-Smith (2010) were also compared with findings in other relevant literature.
User-generated Content Websites’, in order to determine how different factors, such as transpar-ency, social accessibility, risk and technical features, influence user’s experience and further on their engagement behavior. Through a qualitative and quantitative analysis conducted on undergraduate and graduate students, Di Gangi (2010) identifies that transparency of communication has a direct impact on the Engagement Behavior. The Privacy and Security dimension is extensively discussed throughout the paper, but the influence of it on En-gagement Behavior is not supported throughout the empirical study. The result is quite surprising, as concerns for risk related to information security and privacy are previously identified by the same author as being one of the largest concerns in terms of using and creating User-Generated Content.
In his research paper, Cavaye (2004) outlines the insights of Community Engagement based on relevant literature review (Anon, 2001; Bush, 2001; Putnam, 1993; National Eco-nomics, 1999) and observations from the Australian Government. Cavaye (2004) suggests new approaches and describes the actions that would better support citizen participation and engagement. The dimensions identified by Cavaye (2004) as being an important sup-port for citizen engagement mostly refer to relationship building, continuity of dialog or maintaining the quality of communication.
Throughout a literature review paper, Quesenbery (2003) discusses online engagement in terms of ease of use. The author suggests that clear language and appropriate terminology or an appropriate and helpful tone, are criteria for creating an engaging online environ-ment.
Looking from the perspective of usefulness, Chen & Dimitrova (2006), examine the civic engagement via social media platforms and the results confirm the significant impact of perceived benefits, or usefulness, in developing online engagement. In terms of Online Cit-izens Engagement, perceived usefulness refers to the belief that interacting with govern-mental institution online benefits participating citizens, as they can get easy access to the information, quick response, etc. If citizens believe that they can benefit significantly from the governmental institution presence on social media platform, they are more likely to use the system and to be engaged.
Hochheiser & Lazar (2007) admit the important and constructive role that Human-Computer Interaction currently plays in responding to concerns and questions raised by governmental institutions, citizens and other stakeholders. The researchers discuss the range of factors that might stimulate the involvement of HCI. In terms of online interac-tion, Hocheiser & Lazar (2007) found security to be an important concern that can de-crease the intention to use and engage.
In an article published in April 2010, Bertot (2010) explores the potential impacts of in-formation and inin-formation & communication technologies, in terms of e-goverment on social media. Bertot (2010) incorporates the technology literacy, usability, accessibility and functionality as criteria to understand and use technologies and engage with them.
As it shown in Table 2-1 – Engagement Dimension, different citizen engagement dimen-sions proposed by Ide-Smith (2010) were compared and matched with the different factors identified in the Online Engagement literature, described above. After a careful compari-son, the authors of this paper grouped the identified factors into four significant dimen-sions, as follows: Privacy & Security, Information Quality, Responsibility & Relationship and Community Capability & Inclusiveness. The decision was taken after comparing and
summarizing the concept of each dimension, predicting its’ importance with respect to en-gagement on social media and after a conversation with decision makers from Trafikverket Web Communication department.
Table 2-1 – Engagement Dimensions (compiled by the authors) Engagement dimensions by M. Ide-Smith (2010) Concept of the dimen-sion
Other authors Final
En-gagement Dimension Author Factors Theory used
Privacy - Personal identity; - Anonymity; - Conse-quences of revealing per-sonal infor-mation; Di Gangi (2010) Privacy con-cerns; Engagement behaviors in user generated
content; Privacy & Security; Hochheiser & Lazar (2007) Security is-sues; Human com-puter Interac-tion & Society Engagement Trust &
Confidence - Transparen-cy of infor-mation; - Manage-ment of ex-pectations; Di Gangi (2010) Bertot (2010) - Information Transparency; - Information comprehen-siveness; - Information completeness; Community engagement; Information Communica-tion Technolo-gies; Information Quality Information Gathering - Collecting opinions for the purpose of decision making; - Gathering evidence; - Pro-active monitoring; Cavaye (2004) Based on: Anon (2001); Bush (2001); Putnam (1993); National Economics (1999) - Accounta-bility; - Relationship & Trust; Community engagement Relationship & Communi-cation Di Gangi
(2010) Level of re-sponsibility: users expecta-tions; Co-created val-ue theory: so-cial interaction; Managing
Participation - Immediacy; - Quality of dialogue; - Control; - Feedback; - Responsibil-ity and roles;
Cavaye (2004) Based on: Anon (2001); Bush (2001); Putnam (1993); National Economics (1999) - Sustainabil-ity: ongoing feedback; Community engagement Digital
Exclusion - Skills, apti-tude; - Access to information; Cavaye (2004) Based on: Anon (2001); Bush (2001); Putnam (1993); National Economics (1999) - Skills: quali-ty & diversiquali-ty; - Inclusive-ness: equity, opportunity to participate; Community engagement Community Capability & Inclusiveness
Usability - Ease of use and utility;
Quesenbery (2003) Chen & Dimi-trova (2006) Bertot (2010) - Ease of use; - Usefulness; Dimensions of usability; Online civic engagement
The four dimensions identified in Table 2-1 – Engagement Dimensions, will be used as a foundation for the Hypotheses that will help fulfill the purpose of this research. The de-scription for each of the dimensions is as follows.
The first dimension, ‘Privacy & Security’ represents the overall risk perceived in terms of Privacy & Security. Privacy & Security as a perceived risk incudes concerns about the in-formation disclosure, affect of the online identity perception or consequences of revealing personal information. In the case of a governmental institution, the risks associated can be both technical and personal, meaning that they are not only connected to Social Network features, but also to the expression of personal believes and opinions that might influence the perception of the personal identity. Consequently, organizations develop privacy & se-curity policies that minimize the perceptions of the risks associated with engaging in online activities.
The second dimension, ‘Information Quality’, represents the overall perceptions towards the information presented. The elements included into this dimension are information transparency, trustworthiness and relevance. The transparency of information is the main aspect of this dimension, since together with the transparency of information, trust and confidence in the governmental institutions is built.
‘Relationship & Communication’ is the biggest dimension that represents the perception of the overall communication and relationship between the governmental institution and citi-zens online. The dimension is constructed based on the features of communication, such as quality of dialogue, immediacy of feedback and the way in which the communication be-tween the governmental institution and citizens is handled as a basis for relationship. The communication building from the part of the organization is defined as the caring, individ-ualized attention that the users feel they are given by an organization (Kettinger et al., 1994).
The ‘Community Capability & Inclusiveness’ dimension refers to the users’ capabilities and relations with new information technologies. The ‘Community Capability & Inclusiveness’ dimension is closely related to the human-computer interaction literature (HCI), since it describes the relationship with the technology. Out of the HCI literature, the Technology Acceptance Model (TAM), was chosen to be employed in this research paper in order to analyze the ‘Community Capability & Inclusiveness’ dimension. The TAM model (Davis, 1985) is presented in the following section together with the reasons for choosing this model.
2.4 Technology Acceptance Model
Since most of the previous studies done on engagement focus on offline/face-to-face en-gagement, not many authors incorporated humcomputer interaction literature when an-alyzing online processes such as citizens’ engagement with governmental institutions on social media platforms.
In the process of conducting the theoretical part there have been found several theoretical models that represent user acceptance of information technologies. Different models in-clude perceived ease of use as a determinant of acceptance (Venkatesh, 2000), but the
technology acceptance model (TAM) was chosen because of the widest application in terms of user acceptance and usage.
Figure 2-2 -Technology Acceptance Model (Davis, 1985) The Technology Acceptance Model (TAM) is an information system theory that models how users accept and use a technology. Davis (1985) developed this model based on the theory of reasoned action that was adapted in order to understand the causal link between the external variables & the usage intention and the actual use of new technologies. TAM is based on two pillars: perceived usefulness and perceived ease-of-use, and researches on the usage of TAM suggest that perceived usefulness and perceived ease of use both have a sig-nificant effect on technology acceptance (Schepers & Wetzels, 2007). Generally, TAM states that perceived usefulness will be affected by perceived ease of use because the easier a technology is to use, the more useful it can be. According to Bertot et al. (2010), the ease of use covers different areas such as functionality or literacy of technology and it generally describes the design of the technology that includes the characteristics desired by users. The perceived usefulness is defined as ‘the extent to which a person believes that using the system will enhance his or her job performance.’ (Venkatesh & Davis, 2000, pp 187). The close link between TAM and its quite predictive power, makes it easy to use it in dif-ferent situations (Venkatesh, 2000), but at the same time it can become a limitation of the model. TAM is very appropriate when predicting acceptance, but Mathieson (1991) states that even though it is predictive, TAM does not give a clear and sufficient understanding for system designers to develop user acceptance of the platform/system. For example, based only on TAM, public institutions cannot fully predict the acceptance of a new system that is going to be created. Since this study is conducted on Facebook, an already devel-oped social media platform, there is no need for predicting a new technology’s acceptance. Mathieson’s (1991) critical approach on TAM is considered not significant for the current research purpose.
TAM is used in this paper in order to measure the ‘Community Capability & Inclusiveness’ engagement dimension. As presented in Table 2-1 – Engagement Dimensions, ‘Communi-ty Capabili‘Communi-ty & Inclusiveness’ is defined in terms of access to information and ease of use and utility. Today no published studies have investigated the use of TAM to measure the intention to engage with government using social media platforms. Close studies to the field, such as citizen interaction with government via websites, have however put TAM in
close relationship with online governmental engagement. Chen & Dimitrova (2006) show that the higher the level of utilization of information channels, the higher will be the level of engagement with the government online. Perceived usefulness, as a part of TAM, is a concept that can be applied to citizen engagement with government in general (Chen & Dimitrova, 2006). Since social media platforms are another channel of interaction with government, the applicability of TAM is expected.
2.5 Social Networking Sites
Just like customers, citizens can be engaged digitally/online and non-digitally/offline. Digi-tal engagement can be achieved with the help of Web 2.0, which in terms of communica-tion differs from the Web 1.0, in the way that it allows organizacommunica-tions and businesses not only to broadcast their information to the users, but also to receive information from them (Singh, Kumar & Singh, 2010). In Web 2.0 people can interact, collaborate and share con-tent online. The Web 2.0 platform facilitates processes such as streaming media, messaging, news feeds and Social Networking (O’ Reily, 2005).
Although all terms such as Social Networks, Web 2.0 or users’ generated content are usual-ly associated with the term ‘Social Media’, Kaplan & Haenlein (2010) state that Social Net-working Sites are only one of the elements of social media.
The nowadays Internet phenomenon, Social Networking Sites (SNS), are defined as web-based services that give the possibility for users to construct a profile within the web-page, establish connections with other people and communicate (Ellison, B. N, 2008). SNS are actively encouraging business professionals to change communication habits and also in-clude the analysis of the interaction that customers have among themselves, instead of fo-cusing only on the two-way communication between the company and customer. (Brand & Peteohoff, 2010)
In terms of engagement, it is discussed a lot about traditional Web sites and blogs, but very little investigations has been published on social networking sites (O’Brien & Mclean, 2009). To understand better how the engagement process works on social networking sites, it is necessary to analyze users’ behavior and differences that can occur based on that.
2.5.1 Engagement in Terms of Online Behavior
In their article ‘Towards a Science of User Engagement’, Attfield et al. (2011) identify two broad types of online user engagement evaluation metrics: subjective and objective. While the subjective measure is defined as the perceptions that a user reports, the objective meas-ure only reports metrics related to the online behavior. The objective method is considered as being an appropriate measurement system, since the subjective perceptions of the users have consequences that can be objectively measured (Attfield et al., 2011).
As previously described in Section 2.2, the citizen engagement stages are mainly measured in terms of participation. In the same time, focusing on online behavior and participation while measuring the online engagement, is a common practice and web-analytics are devel-oped and adapted for different online platforms. Online behavior as an engagement metric is mainly related to the number of interactions that one user engages with during a certain period of time. The measurement of engagement in terms of online behavior and interac-tion is also suggested in Peterson’s (2006) definiinterac-tion of engagement as an approximainterac-tion of the degree and depth of the visitors’ interaction on a clearly defined set of objectives. Pe-terson (2006) proposes metrics such as reporting the visitors that consume content slowly,
the visitors that return directly to a website or that subscribe to the website’s news feeds. Social Networking sites have a considerably different user’s participation process that can be classified into several activities. Activities as measurement tools for engagement through Social Networking sites can be found in Li, Bernoff, Fiorentino & Glass (2007) Social Technographics research, in which several types of social media behavior are identified.
2.5.2 Types of Social Media Behavior
Based on online behaviors on social media platforms, Li et al. (2007) from the Forrester Research Group, identified six different social media user categories, as follows: inactives, joiners, critics, collectors, spectators and creators (Russo et. al, 2009). The categories are defined after having surveyed almost 10 000 US people in order to learn about their social computing technology use. 4,556 US adults were survey in December 2006 and 4,556 youth in October 2006. The 6 identified groups, called Social Technographics, are de-scribed below.
Figure 2-3 - Social Technographics (compiled by the authors) According to Li et al. (2007), joiners are the category of users who usually tend to join dif-ferent groups and other social activities but not necessarily follow them. Some joiners can be engaged with social computing activities such as reading on particular topics in different discussions. Generally, joiners represent approximately 19 % of the adult online popula-tion.
Spectators are the online users category, which is a very important audience for social
con-tent created by others. Belonging to spectators group takes little effort because these users usually consume what the rest produces. They follow content posted, read information, watch videos and represent 33% of the online adults population. (Li et al., 2007)
Around 52% of the online adult population are inactives and usually do not participate in social computing activities but can get affected when content generated by other active us-ers gets covered in the news media.
Collectors tend to tag web pages (Russo et al., 2009), gather interesting information and
sometimes share it among friends or groups. The action of collecting and aggregating in-formation has an important role in organizing the vast amount of content being produced by two other groups, Creators and Critics (Li et al., 2007).
Inactives Joiners Spectators Collectors Critics Creators
In general, critics post comments on blogs and online forums, post ratings or reviews or edit wikis, but on Facebook they mostly are active in interest groups discussions and keep the “two ways communications”. They like to share their expertise opinion, provide de-tailed feedback and sometimes suggestions, so 4 out of 10 critics also can became creators. (Li et al., 2007)
Creators are the most desired users category, which represents just 13% of the online
adults population. These users are active participants in generating content, creating groups and discussion, sharing information and opinions on various topics.
Since in order to fulfill the purpose of this paper, engaged and unengaged users from Traf-ikverket Facebook profile need to be identified, the activities identified by Li et al. (2007) system will be employed as an engagement measurement tool.
2.6 Research model and Hypotheses
Throughout the literature review and relevant research in the area, four dimensions of en-gagement were identified. The literature suggests that the factors within each of the dimen-sions can differentiate engaged and unengaged users. Four hypotheses were thus created for the purpose of this research, and they will be tested throughout the empirical study. Privacy & Security is an important element in terms of communication on Social Media platforms because, while some content providers as Trafikverket are willing to engage with the users through social media channels, users may not feel free and confident in providing their data, publishing opinions and showing their personality (Di Gangi, 2010). In terms of social media platforms such as Facebook, privacy settings are mostly not expected to be highly developed and usually they are only briefly defined, so more and more users get into a higher level of concerns for privacy and security that affect the usage of social networking sites (Dwyer, Hiltz & Passerini, 2007). Consequently, privacy and security concerns can be-come an important issue when engaging users (Di Gangi, 2010) with governmental institu-tions online. The main concern areas in the communication process with governmental in-stitution can be personal data, other peoples’ opinion and fear to express own opinion in public, especially in terms of critical comments.
The following hypothesis was developed in order to answer the research question number one:
Social Networking Sites provide a new avenue for transparency and openness by giving ac-cess to content provided on social media platforms (Gross & Acquisti, 2005). Consistent standards, such as transparency or openness of the information, need to be considered in order to build an online relationship. In order to deal with possible users fears and con-cerns about transparency of information, governmental institutions should not only create high expectations but also fulfill them through the high frequency of reasonable, convinc-ing and trustworthy information (Di Gangi, 2010). If users perceive the information pro-vided as open and accessible for everyone interested in it, it should have a positive influ-ence on their experiinflu-ence (Mishra, Heide & Cort, 1998), which also influinflu-ences the level of engagement (Di Gangi, 2010). Gattiker & Goodhue (2005) state that the quality, accuracy and relevancy of information encourage positive users attitudes towards information
nology. All these aspects thus become an important factor to consider when talking about engagement through social media platforms.
The following hypothesis was developed in order to answer research question number two:
Bertot (2010) states that social media and presence on social media provides a faster and a cost-efficient way for communication. Engagement with citizens through social media is perceived to be more immediate than other existing methods, which makes users to be more actively involved in the communication with public institutions. When building strong relationships as a basis for engagement, the quality of the dialogue and interactions between public institutions and citizens drive positive attitudes towards the organization’s presence online (Di Gangi, 2010). Then, besides feeling a meaningful person, users get more involved into the communication process. The two-way communication, that de-scribes constant interaction between active users and the governmental institutions (OECD, 2003), helps to build engaging relationships.
The following hypothesis was raised in order to answer research question number three:
The authors of this research paper consider the dimension ‘Community Capability & Inclu-siveness’ to be an important one in terms of Engagement on Social Media, since it refers to the users’ skills and the relations that the users build with new information technologies or platforms of interaction. Bertot et al. (2010) states that the main challenge in the engage-ment process is not the acceptance of the rapidly developing technologies, but access to it. Since in terms of Social Media, Sweden is an advanced country (Kullin, 2011), where more than 85% of the population uses the Internet (Findahl, 2010), eventual ease of use and use-fulness concerns seems to be more probable then barriers for limited access. The authors Schepers & Wetzels (2007) state that perceived usefulness and perceived ease of use, both have a significant effect on technology acceptance and interaction. The two factors, useful-ness and ease of use, which are part of Technology Acceptance Model (see Section 2.3.1), are used in this research paper in order to define the ‘Community Capability & Inclusive-ness’ dimension. When analyzing the users’ relationship with information technologies, ‘Community Capability & Inclusiveness’ might be important factors to differentiate en-gaged and unenen-gaged users since they can explain how and why the relationship with the new technology is developed. While the how question can be explained by the ease of use, the why question mostly refers to the perceived usefulness and benefits associated with the content within a specific technology. Consequently, in order to be engaged in a higher lev-el, users should be able to access and understand information at the desired level of speci-ficity (Di Gangi, 2010).
The following hypothesis was raised in order to answer the research question number four:
H3: Relationship & Communication factors differentiate users in terms of engagement stage.
H2: Information Quality factors differentiate users in terms of engagement stage.
The hypotheses described above are summarized in The Research Model (Figure 2-4) pro-posed by the authors in order to test the differential dimensions within the social media cit-izen engagement process.
Figure 2-4 – Proposed Research Model (compiled by the authors)
Each dimension believed to differentiate users in terms of engagement level is represented in the model. The factors pertaining to each dimension will be provided after the analysis of the empirical data. In the same time, the difference in how the factors are perceived by engaged and unengaged users will be shown after the analysis.
3 Method
This section of the paper presents the choice of the research design and approach and the method used in the research. The process of data collection and the way in which the survey was constructed will be presented. The reliability, validity and generalization of the current research are also discussed.
As a pre-step before designing the current research, a discussion with the decision makers from Trafikverket took place. According to Malhotra & Birks (2006), the discussion with decision makers before defining the problem and the research questions is usually vital. The discussion helped in areas such as understanding the research problem, designing the research questions and objectives and also gaining deeper understanding on aspects such as the ‘Dimensions of Engagement’. In the same time, in-depth data was collected from the organization in the form of documents showing statistics on the level of activity on Traf-ikverket Facebook profile. Also, the decision-makers should know the capabilities of the researchers and possible limitations of the study (Malhotra & Birks, 2006) so significant re-search questions were discussed.
Because of the need for decision makers to be directly contacted by researchers on relevant questions (Malhotra & Birks, 2006), the responsible persons from Trafikverket were in-formed about the process, constantly updated and the company’s resources were also used in order to gain access to the respondents for the survey.
3.1 Research Approach
When designing the research approach, there are two main methods to be considered: the deductive approach and the inductive approach (Saunders, Lewis & Thornhill, 2009; Wil-liamson, 2002; Arbnor & Bjerke, 1997; Schindler & Cooper, 2006).
The deductive approach is mainly related to testing a specific theory and it is characterized by a search to explain causal relationships between variables (Saunders et. al, 2009). As op-posed to the deductive approach, the inductive approach starts with the field work and ob-servations and hypothesis are generated from the analysis of the data collected (Williamson, 2022; Saunders et al., 2009).
This research employs both deductive and inductive method in order to reach its purpose. As identified by Robson (2002), the deductive research progresses through: (1) deducing hypotheses from the theory; (2) indicating how different concepts or variables will be measured; (3) testing the hypothesis; (4) examining the outcome; (5) if necessary, modifying the theory in the light of the finding.
The deductive component of the current research uses theories in order to identify the ‘Dimensions of Engagement’. Hypotheses are constructed in the beginning of the research based on the ‘Dimensions of Engagement’ and ‘Engagement Stages’ derived from the the-ory. Hypotheses are particularly appropriate when the study has a quantitative approach and they together with the research questions, help to clarify the purpose of the research (Williamson, 2002).
However, since the research that has been done in the area of engagement on social media is rather narrow, an inductive approach is also needed in order to fulfill the purpose of the current research. In order to decide how the different ‘Dimensions of Engagement’ will be measured, inductive methods are also employed. Different factors will be identified in the
data analysis in order to further clarify the hypothesis designed in the beginning of the re-search. The effect of the factors based on the overall engagement level is analyzed in this research.
The hypotheses will be tested based on the empirical findings and the theory will be in the end discussed in the light of the findings.
After the research approach has been decided, the research design will help the researchers detail the procedures necessary in order to implement the approach (Malhotra & Birks, 2006). The overall research design is defined as exploratory, descriptive or causal (Malhotra & Birks, 2006; Schindler & Copper, 2006; Aaker, Kumar & Day, 2003), or exploratory, de-scriptive and explanatory (Saunders et. Al, 2009).
This research uses both descriptive and explanatory techniques in order to reach its pur-pose.
The descriptive technique is a type of pre-planned and structured research strategy, as op-posed to the exploratory technique, and it has the purpose of describing a phenomenon (Malhotra & Birks, 2006). In this research, both based on theory and empirical findings, the ‘Dimensions of Engagement’ are portrayed in terms of factors based on empirical findings and theory. This stage constitutes the descriptive part of the study and it is the forerun of defining and testing the hypothesis constructed for the purpose of this research.
According to Schindler et al. (2006) the hypotheses can be either descriptive or relational. While the descriptive hypotheses state the existence, size, form, or distribution of some variables, the relational hypotheses describe a relationship between two variables with re-spect to a specific case. The relational hypothesis can further become correlational or ex-planatory. Explanatory, also called causal hypothesis are used in this research, as there is an implication that the existence of or a change in one variable – ‘Engagement Levels’ - causes or leads to a change in the other variables – ‘Dimensions of Engagement’.
It can be concluded that based on the research questions and objective and the hypothesis employed, this research fits the type of study known as ‘descripto – explanatory’. (Saunders et al., 2006).
3.2 Research Strategy
As previously mentioned in this section, the current research is partly deductive since it us-es hypothus-esis in order to tus-est existing theory. An important characteristic of the deductive approach is the concept that the research needs to be operationalized in a way that enables facts to be measured quantitatively (Saunders et al., 2006). A quantitative strategy was thus considered appropriate in the form of an online survey.
3.2.1 Sampling
A sampling technique is useful if the population size is large and both the cost and the time that is associated with obtaining information from the required population is high (Aaker et al., 2003). Sampling techniques provide a range of methods that enable reducing the data by considering data from only a subgroup rather than from all the possible cases (Saunders et al., 2006).
The main activities that are associated with the sampling process are: (1) identifying the tar-get population; (2) determining the sampling frame; (3) selecting the sampling procedure;
(4) determining the relevant sample size; (5) obtaining information from respondents; (6) validate the sample (Malhotra & Birks, 2006).
The current research analyzes the engagement dimensions in terms of social media com-munication of Trafikverket Swedish Transport Administration Facebook profile. In order to decide upon the sample, a first look is given to the target of Trafikverket Facebook pro-file.
Three groups are identified by Trafikverket as a target group for this channel: road travel-ers, local residents and prospective employees (see APPENDIX 1).
Road travelers: Travelers who are traveling on any transport means, for example train
travel-ers, bus travelers and other public transport users. The road travelers also include all citi-zens moving in the traffic environment, such as motorists, pedestrians, cyclists, motorcy-clists, scooter riders, commuters, tourists, etc. Mobility can be both privately or for profes-sional purposes. This target segment includes also interest groups, industry associations and partners, like for example, NTF, train travelers or snowmobile associations. It is the team that scores the most posts in the log and creates the most activity.
Neighbors: People who are concerned by the Agency's construction and maintenance
pro-jects, or persons affected by the nearby road and rail.
Prospective employees: Individuals seeking employment and students seeking the opportunity
to do thesis work. Persons looking for work experience or trainee position and summer jobs.
After a careful evaluation of the target population, Trafikverket Facebook followers are chosen as a sample for this research, through a process that is called judgmental sampling. The judgmental sampling is a non-probability sampling technique in which the population elements are selected based on the judgment of the researcher (Malhotra & Birks, 2006). The approach of selecting Trafikverket followers for the current research is highly relevant for four main reasons. Firstly, most individuals in the group are likely to have knowledge of governmental institution communication on Facebook, hence their perceptions towards Trafikverket communication on Facebook are more reliable. Secondly, the individuals in this group play an important role in Trafikverket Facebook profile, since they follow the news, they create discussions and through their engagement new users can be attracted to the Facebook profile. Thirdly, as the definition of Trafikverket target audience shows, the main target of the Facebook profile, the road travelers, are also the users who create the most posts and activity on the Facebook page, so their opinion is highly significant. Lastly, by gathering responses from Trafikverket followers, insights can be gathered from both cit-izens that engage with public institutions on Facebook profile, and citcit-izens who do not en-gage or participate in this kind of activities and analyze the differential factors comparative-ly. The sample consists of 1.823 followers, as per 30th May 2012 (see APPENDIX 3).
3.2.2 Survey
It was taken into consideration that some authors argue that observations are more appro-priate to capture accurate data and qualitative studies are a better option in order to investi-gate behavioral sciences (Andrews, Nonnecke & Preece, 2003), but, the authors of this re-search paper have chosen the web survey because of the possibility to quickly access the
target audience and collect opinions on particular questions. Also, web based surveys, con-ducted using tools such as Qualtrics, provide the possibility to verify and automatically store answers using database technology (Andrews at al., 2003). In the same time, Qualtrics as a survey tool is one of the most valuable software in the industry of online research and free to use for students, which made it rewarding to work with.
Surveys are popular since they allow the collection of large amounts of data. The survey strategy allows the researcher to collect data that can be analyzed quantitatively and it is usually employed when there is a purpose of showing relationships between variables and to produce models of these relationships. As this research has the purpose to analyze the Dimensions of Engagement and show relationships between the Dimensions and the Stag-es of Engagement, the survey strategy was chosen.
Other methods considered
For the purpose of this research, two other research methods were considered before de-ciding to use the Survey method.
Netnography was considered as a method choice for conducting the current research. Netnography is described as ’a written account resulting from fieldwork studying the cul-tures and communities that emerge from on-line, computer mediated, or Internet-based communications’ (Kozinets, 1998, p. 366). Netnographic studies are similar to ethnograph-ic studies, whethnograph-ich are mostly used in inductive research (Saunders at al., 2009). The current research has a strong deductive component, for which questionnaires are considered to be more appropriate. The netnographic method was also considered not to be appropriate be-cause of the lack of user-generated content on Trafikverket Facebook profile and the low number of active participants in the discussions. In the same time, netnography would not have permitted the investigation of personal opinions and feelings, since the conversations are based on general problems, for example road works or information about traffic jams. Although the engagement stage could have been established based on Netnography, a con-nection between the Engagement Stage and the personal opinion about the communication would not have been possible.
In-depth interviews were also considered as a method to serve the purpose of this research, but this method was excluded due to sampling drawbacks identified for this particular re-search. Another reason for choosing the online survey of a larger sample rather than in-depth interviews of a smaller sample is the fact that gathering a large number of respond-ents produced balanced data to work on the analysis.
3.3 Pilot testing
Before gathering the empirical data used in the study, a pilot test survey was sent to a small sample of both followers and non-followers of Trafikverket Facebook profile. The results of the pilot study lead to rephrasing and removing of several attributes that the respond-ents considered as being hard to answer. The original question related to ‘Information Quality’ consisted of ten items, but the respondents reported confusion and incapacity of distinguishing between several items. In order to give an example, the Believable-Unbelievable item was excluded from the questionnaire, since it overlaps with the Trust-worthy – UntrustTrust-worthy item and the participants reported difficulties in distinguishing be-tween the two based on Trafikverket Facebook communication.
3.4 Questionnaire design
The method chosen for collecting primary data within this research is the survey research method, in the form of an electronic questionnaire. The web-based survey was launched and advertised during the period 1April and 16 of April 2012 using Trafikverket Facebook profile. The link for the questionnaire was posted 4 times during the mentioned period, in order to remind users and keep the link in the Facebook news feed.
The questionnaire consists of 21 questions and incorporates the following types of ques-tions:
• filtering questions that measure ’familiarity, product use and past experi-ence’ (Malhotra & Birks, 2006); this type of questions were asked before the questions about the topic itself in order to filter out the respondents who are not adequately informed about the topic of the research; two questions of this type were used;
• unstructured questions – ’open-ended questions that respondents answer in their own words (Malhotra & Birks, 2006); these questions carried no ’force response’ feature, given the fact that they constitute no key questions for testing the hypothesis and are used as additional information; three ques-tions of this type were used;
• structured questions that ’specify the set of response alternatives and the re-sponse format’ (Malhotra & Birks, 2006); the possible answers of the struc-tured questions were in the form of multiple-choice answers, dichotomous answers and scale answers.
Engagement is measured in the paper in terms of social media behavior and participation (see Section 2.5) and a scale was developed based on Li et al. (2007) research. Four dimen-sions were incorporated in the survey design to measure the factors found to be relevant for the engagement levels: ‘Privacy & Security’, ‘Information Quality’, ‘Relationship & Communication’, and ‘Community Capability & Inclusiveness’.
In order to measure the ‘Privacy & Security’ factor, a 3 item, 5 – point Likert scale was adapted from Dinev & Hart (2004).
The ‘Information Quality’ factor was measured on a 5 item, 6-point Likert scale adapted from Beltramini (1982) Believability scale. Out of the ten attributes used by Beltramini (1982), five were used for this research since the pilot testing has shown that the content on Trafikverket social media platform cannot be distinguished in terms of all of the ten at-tributes. Furthermore, in order to measure the same dimension, the ‘Easy to understand’ – ‘Hard to Understand’ item was added to Beltramini (1982) Believability scale.
The ‘Relationship & Communication’ dimension was measured using two different prede-fined factors: ‘Communication Style’ and ‘Interactivity Satisfaction’. For the ‘Communica-tion style’ a 4 attribute, 6-point scale was developed for the purpose of this research. For the ‘Interactivity satisfaction’, a 5 attribute, 5-point Likert scale was adapted from Liu (2003) Two-Way Communication scale.
In order to measure the ‘Community Capability and Inclusiveness’ dimension, two factors were considered: usefulness and ease of use. The perceived usefulness and perceived ease of use are both measured on a 3 item, 5-point Likert scale adapted from Davis (1989). The
