A typology of sns users

28 Int. J. Web Based Communities, Vol. 7, No. 1, 2011

Copyright © 2011 Inderscience Enterprises Ltd.

A typology of social networking sites users

Petter Bae Brandtzæg* University of Oslo, and SINTEF ICT, Forskningsvn. 1, 0314 Oslo, Norway Fax: +47-22067350 E-mail: [email protected] *Corresponding author

Jan Heim SINTEF ICT, Forskningsvn. 1, 0314 Oslo, Norway Fax: +47-22067350 E-mail: [email protected]

Abstract: Social networking sites (SNSs) are said to be new important means of participating, communicating, and gaining social capital. Thus, increasingly fragmented user population and user behaviours in SNSs make it important to achieve more knowledge about SNS users and their participation inequality. This article proposes a user typology for SNSs, which identifies and describes the variety of ways in which people use SNSs. An analysis of the survey data from 5,233 respondents in four major Norwegian SNSs showed five distinct user types:

1 sporadics 2 lurkers 3 socialisers 4 debaters 5 actives.

Both quantitative and qualitative analysis validates the typology. The SNS user typology contributes to a more nuanced and precise measure of how future research should identify and predict SNS use and better understand participation inequality in SNSs. The identification of various user types indicates a 50-30-20 rule for participation in small and locally bounded online communities compared to the existing 90-9-1 rule. Finally, the results could help the design of SNSs in tailoring them to user type.

Keywords: social networking sites; SNSs; online communities; participation; users; user typology; personality types; Norway.

Reference to this paper should be made as follows: Brandtzæg, P.B. and Heim, J. (2011) ‘A typology of social networking sites users’, Int. J. Web Based Communities, Vol. 7, No. 1, pp.28–51.

Biographical notes: Petter Bae Brandtzæg is a Research Scientist at SINTEF ICT and a PhD candidate at the University of Oslo in the RECORD project. His research interests include social media, user behaviour, social capital and

A typology of social networking sites users 29

privacy issues. He received his MS in Psychology from the Norwegian University of Science and Technology, Trondheim, Norway in 2000. He is a member of ECREA and the ACM. He has recently been a Guest Editor for Computers in Human Behavior.

Jan Heim is a Chief Scientist at SINTEF. He has previously been an Associate Professor and Head of the Department of Psychology at University of Trondheim. He is at present working on human-computer interaction with a focus on user requirements and psychological aspects of mediated communication and has done so in various European research projects (Telecommunity, TASC, USER, INUSE, RESPECT, Vis-à-vis, Eye-2-Eye, CITIZEN MEDIA). He is the author or co-author of several international papers.

1 Introduction

The popularity of social networking sites (SNSs) such as Facebook, Orkut, LinkedIn, and MySpace has boosted social interactions and user engagement online among several hundred million users worldwide. In Norway, where this study takes place, 53% of the Norwegian online population between 15 and 74 years are users of SNSs (Brandtzæg and Lüders, 2008).

According to Bishop (2007), SNSs have changed the nature of online user participation into more democratising forums where people can communicate and add their user-generated content (UGC). SNSs are said to represent an important mechanism for knowledge exchange and sharing (Shen and Khalifa, 2009). The change from passive consumption to more active use indicates more varied use and forms of participation (Preece and Shneiderman, 2009) and a new digital divide between actives and non-actives (Brandtzæg, 2010). Participation inequality in SNSs might, therefore, affect people’s ability to gain social capital (Ellison et al., 2007; Brandtzæg et al., 2010), discuss, and become involved in new forms of information and civic engagement (Brandtzæg and Lüders, 2008).

Despite the explosive interest in and use of SNSs worldwide, little is known about the various ways in which users engage in SNSs. According to a review by Boyd and Ellison (2007), scholars have a limited understanding of who is using these sites, why, and for what purposes, especially outside the USA. Therefore, this article explores the patterns of adoption and use of SNSs by developing a typology of the users of these sites. The development of a classification scheme of SNS users’ patterns identified through an empirically-based typology could reveal the mix of underlying activities and participation inequality of such sites (e.g., Bunn, 1993). Such a typology could also help researchers, designers, and managers understand what motivates technology-mediated social participation. This will enable them to improve the interface design and social support for such sites (e.g., Brandtzæg et al., 2010; Preece and Shneiderman, 2009; Shen and Khalifa, 2009).

More specifically, the main objective of this article is to outline an empirically-based SNS typology that can be used to describe and identify distinct user types characterised based on the ways in which SNS users behave in terms of participation level and participation objective. There were two sub-goals to achieve this:

30 P.B. Brandtzæg and J. Heim

• to empirically establish a user typology of SNS users based on questionnaire data on user behaviour from 5,233 SNS users in Norway

• to empirically validate the SNS user typology.

Validation of the cluster solution of the different user types is carried out by comparing the different user types’ responses to other survey questions reflecting a diverse user participation. In addition, different user types are validated by analysing typical qualitative responses from distinct user types.

It should be noted that the approach in this study is exploratory rather than one of testing theoretical hypotheses. This is because no firm body of empirically-based theoretical knowledge exists about users of SNS, especially not about typologies.

1.1 A user typology approach

The goal of a typology is to classify diversified behaviour into meaningful categories (Barnes et al., 2007). In contrast, media behaviour is usually understood in terms of frequency of use only (Heim et al., 2007) and lacks a framework to identify different media usage more precisely. The challenge of the research described in this paper is, therefore, to go beyond the traditional measures of time to include how people use SNSs, to better understand participation inequality in terms of a typology.

In this paper, we conceptualise usage in terms of users’ level of participation (intensity of use) and participation mode (objective and direction of participation) in the SNS, as shown in Figure 1. Both participation level and participation mode are treated as two equally important dimensions; at the same time, these dimensions cover a broad range of SNS activities such as time-killing, debating, and UGC production. This approach is similar to Shih and Venkatesh’s (2004) conceptualisation of user types in terms of variety and rate, where variety referred to the different ways in which the product could be used and usage rate referred to how often the product was used, regardless of its application.

Figure 1 A conceptualising of user types in terms of different modes of and level of participation in SNSs

Recreational modeInformational mode

High participation

Low


In this paper, user types in SNSs are conceptualised by

1 their participation objective

2 level of participation.

Users may be either high or low in terms of their participation, but some users may have both a recreational mode and an informational mode of communication.

2 Previous research on user types in SNSs

2.1 User typologies

To generate the research questions (RQs) and validate our approach for this article, this section provides a brief introduction to the study of user types that might relate to SNS usage. First, it should be noted that the categorisation of people based on a typology is not a new theoretical approach. In attempting to understand human nature, psychologists have been developing personality typologies for many years.

The most influential theory in research on technology use and adoption that applies to a typology is Rogers’ diffusion of innovations model (Rogers, 1962, 2003). This model explains the process by which innovations are adopted and offers the following categorisation of users, based on their rates of adoption of innovations over time:

1 innovators (around 2.5%)

2 early adopters (13.5%)

3 early majority (34%)

4 late majority (34%)

5 laggards (16%).

The theory’s weakness lies in the fact that it describes and classifies the innovation and adoption rate in general, rather than the differences in actual media behaviour. Still, the theory might prove useful in determining methods to describe, predict, and identify different user types in an SNS, since SNSs are a kind of innovation.

Some media studies have applied a typology approach to understand differences in media behaviour in general. Johnsson-Smaragdi (2001) identified four main user types among children in nine European countries:

1 low media users

2 traditional media users

3 specialists

4 screen entertainment fans.

A similar approach was suggested by Heim et al. (2007) in their study of children’s media usage in Norway, and by Ortega Egea et al. (2007) in their attempt to understand the diffusion and usage patterns of internet services in the European Union among citizens in general.


When it comes to online communities or social media sites such as SNSs, only a few attempts have been made. One informal, but influential, study by Nielsen (2006) suggested a 90-9-1 rule, which refers to the unbalanced nature of participation in social media with UGC. An example of this rule is Wikipedia, where more than 99% of users do not contribute and only consume. Wikipedia has 68,000 active contributors, who make up 0.2% of the 32 million unique visitors to the site in the USA alone. A rough explanation of the 90-9-1 rule is as follows:

1 Lurkers (90%) read or observe, but do not contribute. Kollock and Smith (1996) described lurkers as free-riders, that is, non-contributing, resource-taking members. This is also referred to as the free-rider problem.

2 Intermittent contributors (9%) contribute from time to time, but other priorities dominate these contributors’ time.

3 Heavy contributors (1%) are active users who account for most of the contributions and system activity.

Verifying Nielsen’s (2006) categories is only somewhat possible because no detailed descriptions of these results or of the method exists, other than the defective information published on Nielsen’s Alterbox, a blog site. However, the 90-9-1 rule has been influential in the discussion on social media usage since the idea was launched, and it should be interesting to see if we find the same pattern in the study presented in this paper.

An online community user typology was suggested by Kozinets (1999). In a theoretical paper, he described various types of community members according to two factors: the degree of consumption activity and the intensity of relationships with other members of the virtual community. These two factors enable four distinct types of community members to be identified:

1 Tourists are users who simply drop by the community every now and then with only superficial interest and few social ties

2 Minglers are users who maintain strong social ties while being marginally interested in consumption activity

3 Devotees are users who maintain a strong interest in consumption but have few social attachments

4 Insiders are users who have strong social ties and a strong interest in consumption activity.

This typology is based not on empirical work but on theoretical assumptions. Kozinets also stressed the importance of categorisation in making the fragmentation of online community users more meaningful.

Recently, a report from OFCOM (2008) suggested a typology specific to SNS users, based on in-depth interviews with 39 users. Five types were identified:

1 alpha socialisers are regular users who visit SNSs often, but only for short bursts to flirt and meet new people

2 attention seekers are users who crave attention and comments from others, often by posting photographs of themselves and friends


3 followers are users who join sites to keep up with what their peers are doing

4 faithfuls are users who rekindle old friendships, often from school or university

5 functionals are users who log on for a particular purpose such as looking for music and bands.

The above user typologies related to SNSs are not based on quantitative information, but rather on sparse qualitative and theoretical judgments. This means that we know little about both user types and the distribution of user types in SNSs. Further, to the best of our knowledge, OFCOM’s typology is the only one that addresses SNSs specifically. Generally, the sparseness of empirical material related to SNSs and typologies of their users (i.e., a single research report) suggests that further research is needed.

However, we should note that, although much research into online communities (e.g., Preece, 2000) exists, SNSs are a new type of community that may play a role different from that described in the early literature on virtual communities (Ellison et al., 2007). This literature focused on how community participants who were geographically dispersed and motivated by ‘meeting’ new people online (e.g., Wellman and Gulia, 1999). New SNSs, on the other hand, are an ‘all in one place solution’ that merges several interactive services and application related to work, leisure, and information (Brandtzæg and Heim, 2008). For example, more than 350,000 applications are in use on the Facebook platform according to Facebook’s September 2009 statistics. It should, therefore, be highlighted that an empirically-based user typology that divides the various user behaviours into meaningful categories in SNSs is absent.

2.2 User types and social implications

A user typology could also benefit research into the social implications of various SNS usage, and hence, it is important to understand various forms of participation. Social capital can be used to investigate the social implications of SNSs. The concept has two complementary aspects:

1 social contact, which covers patterns of interpersonal communication, including the frequency of social contact and social events

2 civic engagement, which covers the degree to which people become involved in their community, both actively and passively, including political and general organisational activities (Quan-Haase and Wellman, 2004).

Most studies on social effects (e.g., Kraut et al., 1998) and social capital (Ellison et al., 2007) have not focused on particular forms of use, but rather on the time spent online and the frequency of online visits. One study that we are aware of, by Shah et al. (2001), addressed four types of internet usage patterns and different facets of social capital and found that ‘social recreation’ usage (e.g., participation in chat rooms and game playing) was consistently negatively related to civic activities, trust in other people, and life contentment. In contrast, an ‘information exchange’ pattern had a positive impact on the same social variables. These results suggest a connection between different user types and various social outcomes.


3 Research questions

An exploratory approach is a natural point of departure in a relatively immature field of research.

Using the above review as a basis, we explore the following main RQ:

RQ1 Given that users in SNSs display different patterns of use in terms of the users’ participation and communication modes, how can users be classified into meaningful categories or user types?

We asked the following sub question:

RQ2 What are the associations between different types of SNS users and demographic factors such as age, gender, and education?

To further validate the typology, we introduced RQ3 and RQ4, below. The measures used to identify UGC participation, social contact, and political activity included distinct measures that were used to identify the typology, to avoid circular questions.

RQ3 How are diverse user types reflected in different levels of UGC participation?

RQ4 What are the relationships between distinct user types and social capital, in terms of number of friends in the SNS user’s profile, frequency of contact with friends, people, and political activity?

4 Method

4.1 Sample

The data were collected using an online questionnaire over a three-week period in March 2007 on four different SNSs. These SNSs were chosen because, at the time of the investigation, they were the most popular SNSs in Norway, which, consequently, might give us a good picture of what typical SNS users seek regarding relationships. The frequent usage and popularity of these sites were documented in a recent report for the Ministry of Government Administration and Reform in Norway (Brandtzæg and Lüders, 2008), which provides a detailed overview of the most popular SNSs in Norway.

A total of 5,233 persons responded to the survey, of which 3,078 were women and 2,155 were men, with a median age of 16 years. Demographic details given by the respondents on each SNS are as follows:

• Biip.no (n = 2,278) consists of adolescents with a median age of 15 years mostly in their final year of comprehensive school. The number of females (62%) greatly exceeds the number of males (38%). The users have no specific type of residence and live all over the country.

• HamarUngdom.no (n = 1,598) consists of adolescents with a median age of 16 years who are mostly in high school. There are somewhat more females (53%) than males (47%). The majority of users live in rural areas.

• Nettby.no (n = 512) consists mostly of young adults with a median age of 20 years. There are more females (58%) than males (42%). The sample members have no


special type of residence and live all over the country. About one third of the members are high school students or college/university students.

• Underskog.no (n = 335) consists of adults with a median age of 29 years. The members have a high level of education compared to the rest of the sample. There are somewhat more males (55%) than females (45%). Most live in the Oslo area (the largest city in Norway).

The four sites fit well into the definition of an SNS (e.g., Boyd and Ellison, 2007). In addition, they reflect some interesting differences (e.g., size and/or focus) that make it possible to investigate more closely how different user types are distributed in SNSs with different numbers of members and diversities in focus. Table 1 describes these SNSs in more detail. Table 1 Description of the four SNSs used in this study

SNS Origin Members Description

Biip.no 2005 June 280,000, March 2007

One of the most popular SNSs for teenagers in Norway. Mostly for socialising and sharing pictures, music, and videos.

HamarUngdom.no 2002 August

190,000, March 2007

Initially a local online community targeting youths in a small town in Norway called Hamar. The community has become very popular and grown outside its original borders. Mainly for socialising and discussions.

Nettby.no 2006 September

320,000 March 2007

Norway’s most popular SNS, connected to Norway’s largest online newspaper. It attracts a wider sample of the population. Popular for discussion and interests groups, as well as socialising.

Underskog.no 2005 November

10,000, March 2007

Originally a user-generated cultural calendar for Oslo. Very academic and cultural oriented, but also for socialising and discussions. This is an invitation-only community.

Note: These numbers were collected by e-mail from the different community owners at the same time as this study was executed, March 2007.

4.2 Procedure

We used an online survey questionnaire that contained open and fixed-response questions. We urged the site owners to distribute the survey to all of their members inside the SNS, either by a tag in the members’ user profiles or with a message. This was done on Biip.no and HamarUngdom.no, whereas on Nettby.no and Underskog.no, the survey was merely introduced by a banner ad on the front page and by blogging contributions that urged users to participate. This type of survey allowed us to easily access a very large number of users of these SNSs, while the users were actually using the sites. To motivate as many users as possible (i.e., including less active to very active users) to participate in the survey and complete the questionnaire, all the participants from the four SNSs were entered into a raffle to win a travel gift coupon worth US $1,750. A technical solution made it impossible for any user to participate multiple times in the study.


4.3 Measures

All the measures and questions were identical in all four online surveys.

4.3.1 Socio-demographics

We used the socio-demographic measures of residence, gender, age, and education.

4.3.2 User typology measure (fixed question)

To identify distinct user types in the SNS (RQ1), we asked the question, “What are your reasons for visiting this SNS today?” This is a very concrete question about the users’ actual activities when they are on the site and is not subject to interpretation on the part of the respondents. This question was followed by 18 yes/no alternatives that reflect different modes of communication (informational vs. recreational) and levels of participation (high vs. low). These alternatives are shown in Table 2. Some examples for the informational mode are ‘look for new information’, ‘discuss’, and ‘make appointments for meetings with other people’, while examples for the recreational mode include ‘profile surfing’, ‘kill some time’, and ‘look for a new friend’. These alternatives aimed to cover most of the possible and common user activities or behaviours within the SNS. The fact that very few participants used the alternative ‘other’ also indicates that we considered most of the possible user activities in the present questionnaire. The 18 activities were defined based on a discussion with the developers of Nettby.no and a pilot investigation conducted with ten target users between the ages of 14 and 38 years who had experience with SNS usage.

4.3.3 User typology measure (open-ended questions)

After the fixed question (RQ1), we included the following three open-ended questions:

1 What is your most important reason for using this SNS?

2 Why have you stopped using, or are less active in, other SNSs? (This question was answered by only 48% of the sample who reported being less active in an SNS.)

3 Under what conditions would you consider contributing more video footage to the SNS?

These three questions were designed to encourage full, meaningful answers using the subject’s own SNS experiences, so that we could achieve a more in-depth understanding of the user typology yielded. In addition, the questions were selected with the aim of conducting a qualitative analysis that could verify the validity of the different user types we identified in the statistical cluster.

4.3.4 UGC participation

To answer RQ3 concerning how user types relate to levels of UGC production, we measured UGC contribution in terms of how often users reported text production, image contribution, and video contribution on a six-point scale (i.e., ranging from ‘never’ to ‘several times a week/daily’).


4.3.5 Social capital – social contact and civic engagement

The two aspects of social capital (RQ4), frequency of contact and level of civic engagement, are commonly used measures in surveys to reveal people’s level of involvement in a community (Quan-Haase and Wellman, 2004). Social contact was measured in terms of frequency of contact with different forms of relationships:

a friends

b family

c other people, using the same six-point scale as above (i.e., ranging from ‘never’ to ‘several times a week/daily’).

In addition, the number of friends in the user’s own profile was reported. Civic engagement was measured in terms of the user’s attitude toward his or her political engagement with the question, “How important is it for you to express yourself politically inside your SNS?” using a six-point Likert scale (i.e., ranging from ‘not important’ to ‘very important’).

4.4 Analysis

A k-means cluster analysis (SPSS, v.14) was employed to identify typical user types in the SNS. Cluster analysis is an empirical technique designed to partition a heterogeneous sample into homogeneous subgroups to discover classifications within complex datasets (Blashfield and Aldenderfer, 1978). The method implies the classification of similar objects into different groups or, more precisely, the partitioning of a dataset into subsets (clusters) so that the data in each subset (ideally) share some common trait – often proximity, according to some defined distance measure. It is a method of categorising objects into naturally occurring groups and is used in medicine (disease groups), biology (animal and plant groups), and marketing (groups of people with similar buying habits) (Gore, 2000). The method could, therefore, be employed to determine similarities of usage patterns into occurring user types in SNSs as well.

K-means cluster analysis (see Tan et al., 2006) is recommended when the number of entities (persons) in the analysis is high (more than 1,000). All variables in the analysis were dichotomous: zero or 1. We used 18 variables and 5,233 respondents, and there were no missing values.

However, it is always a challenge to define how many cluster solutions should be chosen. Several different cluster solutions were evaluated for their effectiveness in producing cohesive groups that were distinct from one another, large enough in size to be analytically useful, and substantively meaningful (Horrigan, 2007). To validate the final cluster solution, we repeated the same analysis ten times with a random sample of 40% of all cases in the material. In six out of the ten analyses, all the clusters were reproduced. In one of the ten analyses, four of the five clusters were reproduced. In three out of the ten analyses, three of the five clusters were reproduced. This indicates that the final solution was robust.

A five-cluster solution seems to be an adequate size when we compare this solution with the cluster outcome in similar studies identifying user types (e.g., Ortega Egea et al., 2007). According to a review by Brandtzæg et al. (2009), cluster solutions vary between four and six.


To further validate the cluster solution and to illustrate the different user types more in depth, one or two typical representatives for each pattern were extracted from the dataset and presented. These typical respondents were chosen from those with the least deviation from their cluster mean. A further requirement was that they gave a reasonable number of comments in the open-ended questions for statistical data analysis.

5 Results

5.1 The user typology

The analysis suggested five distinct clusters, interpreted as user types that reflect different participation levels and communication modes or deeper motivations for using an SNS. Table 2 shows the proportion of cluster members indicating their reason “for visiting the SNS today” in terms of user activities reflecting the users’ typical behaviour inside the community. The clusters were given user-type labels, the formulation of which was based on an interpretation of the values in Table 2, indicating the users’ level of activity in different types of behaviour. Table 2 Five user types based on results from k-means cluster analysis of the 18 variables

(N = 5,233)

SNS usage Cluster 1 Cluster 2 Cluster 3 Cluster 4 Cluster 5 1 Write a contribution 0.04 0.03 0.08 0.78 0.53 2 Find an announcement for

an event 0.04 0.01 0.03 0.00 0.17

3 Publish or share pictures 0.01 0.11 0.52 0.22 0.81 4 Publish or share

audio/music 0.01 0.00 0.02 0.03 0.09

5 Publish or share film/video

0.00 0.00 0.02 0.04 0.02

6 See if somebody has tried to contact me

0.28 0.94 0.94 0.90 0.98

7 Look for a new friend 0.10 0.13 0.67 0.22 0.77 8 Read new contributions 0.14 0.41 0.37 0.94 0.86 9 See other people’s

videos/pictures 0.06 0.42 0.55 0.38 0.92

10 Make appointments 0.03 0.06 0.17 0.12 0.43 11 Look for new information 0.05 0.21 0.11 0.30 0.52 12 Write letters or messages 0.09 0.34 0.83 0.54 0.92 13 Discuss 0.03 0.04 0.10 0.70 0.64 14 Run community groups 0.00 0.00 0.02 0.09 0.13 15 Profile surfing 0.06 0.44 0.39 0.39 0.90 16 Contact others 0.17 0.54 0.92 0.65 0.96 17 Kill some time 0.13 0.89 0.28 0.67 0.71 18 Other reasons 0.11 0.11 0.13 0.10 0.30 Final cluster labels: Sporadics Lurkers Socializers Debaters Actives

Note: To make this table more readable; values below 0.10 are in italic, while values above 0.70 are in bold.


As seen in Table 2, actives seem to be even more social than socialisers. Similarly, actives seem to be at least as involved in ‘debating’ as debaters, but on closer inspection, debaters seem to mainly debate and almost completely ignore other activities. Socialisers mostly socialise and engage in other activities to a lesser extent. Actives are so-labelled because they are active in almost every type of participation inside the SNS – they socialise, debate, and engage in several other activities, even killing time (as do lurkers). A detailed interpretation of each user type is described below, while Figure 2 shows how the clusters follow the specific criteria related to the level of participation (low/high) and modes of participation.

Figure 2 How the different user types link to participation level and modes of participation

Note: Since actives are active in different participation modes, actives are, in practice, placed in several different locations on this axis but are ‘high’ in participation in all activities.

Figure 3 Distribution of different SNS user types in terms of % (see online version for colours)

0

5

10

15

20

25

30

Socializers Debaters Lurkers Sporadics Actives

In p

erce

ntag

e

Recreational mode Informational mode

High participation

Low participation

LurkersSporadics

Debaters

Actives

Socialisers


Figure 3 gives the distribution in percentage of the five distinct user types. Lurkers represent the largest user group (27%), while debaters are the smallest user group identified (11%).

Table 3 shows an overview of the different user types with regard to demographics such as gender, age, and the level and proportion of users who have received higher education (college or university) (see RQ2). All reported differences between the clusters are significant in regard to Pearson’s chi-square (χ2) test; this is due in part to the large number of participants in the analysis. Table 3 SNS user types and socio-demographic differences

User types Median age Male in % Female in % Higher education in % (> 18 years)

Socialisers 15 31 69 33

Debaters 18 51 49 70

Lurkers 16 42 58 60

Sporadics 16 47 53 70

Actives 16 41 59 39

Note: To avoid results being confounded with age, only people older than 18 were included in the education analysis. Higher education refers to a college or university education.

In terms of age, debaters are older (median age: 18 years), while socialisers are the youngest user type (median age: 15 years). In terms of gender, girls are significantly more common among lurkers, socialisers, and actives. As one might expect, debaters are the most educated user type. However, sporadics are also educated. This could be related to the fact these users are characterised by a kind of a goal-directed instrumental use, using the community as an advanced form of e-mail.

Further, the user types are distributed unequally in the different SNSs, as shown in Table 4. Underskog.no attracts debaters (42%), and Biip.no attracts socialisers (32%); actives (24%) are most common on Nettby.no and quite rare on Underskog.no (6%). lurkers are the most equally distributed user type in the four SNS communities. Table 4 Distribution of user types in different SNSs in %

User types HamarUngdom.no Nettby.no Biip.no Underskog.no

Socialisers 22 15 32 3

Debaters 11 26 36 42

Lurkers 31 21 26 27

Sporadics 14 14 22 22

Actives 22 24 16 6

TOTAL 100% 100% 100% 100%

The following section is a detailed description of the different user characteristics related to each user type to answer RQ1. This section also includes results from the analyses of the answers to the open questions from users representing each user type.


5.1.1 Type 1: sporadics (19%)

The ‘sporadics’ are so named because they visit the community only from time to time, but not a frequent basis. These users have a low level of participation and tend more toward an informational mode since they, for the most part, check their status and see if somebody has contacted them (see Table 2). Sporadic users give few reasons for visiting the SNS. They are spread equally over the four SNSs and age groups. There is no large gender difference (see Table 3). A typical sporadic user is exemplified in the qualitative material, using the open-ended questions. One typical user is August, a 16 year old. He has 17 people in his profile but has been in contact with only one other person during the past week. August joined the community to keep in touch with friends. He is not interested in contributing UGC. He puts it this way: “It is just not my thing to submit a bunch of videos, etc., on the net”.

5.1.2 Type 2: lurkers (27%)

Lurkers make up the largest user category. They are named ‘lurkers’ since they are quite low in participation and participate in activities that are more related to recreation. These users are somewhat involved in several activities, but only passively or to a small degree. In addition to “see if somebody has tried to contact me”, lurkers score high on “kill some time” (0.89) (see Table 2). In addition, lurkers are less likely to be contributors of UGC. lurkers are spread quite equally over all four communities (from 21% on Nettby.no to 31% on HamarUngdom.no) and consist of more females (58%) than males (42%). A typical ‘lurker’ identified in the qualitative analysis is June, a 16 year old from a small town. She has been a member of HamarUngdom.no for about three years but is starting to lose interest. She thinks that technology is important for entertainment but less so for keeping in touch with others. She has not socialised with anybody on the SNS this week. She has about 20 people in her profile. According to June, “It is always funny to watch a good video before the day starts”. Further, she is not a contributor, but more of a ‘free-rider’, who consumes others’ UGC.

5.1.3 Type 3: socialisers (25%)

These SNS users are the next biggest user type and are labelled ‘socialisers’ since their behavior is characterised by recreational in terms of ‘small talk’ with others, but the users’ participation level is high. They score high on ‘write letters and or messages’, ‘contact others’, and ‘look for a new friend’ (see Table 2). This pattern is typical of teenage girls (median age: 15 years; 69% of the sample), mainly on Biip.no (32%) and practically absent among the Underskog.no members (3%). A typical ‘socialiser’ in the qualitative data is Mary, a 14 year old who lives in a medium-sized town. She is an eager member of Biip.no and has been a member for almost a year. The internet is very important to her. She uses the community to keep in touch with friends, and she likes to make connections with new people. In the last week, she has been in touch with five people. She has about 30 contacts in her profile. Her response to the question, “What is your most important reason for using the SNS?” is, “It’s interesting. I have something to do in my spare time (…). I have contact with friends, write in friends’ guest books, comment on people’s pictures, send SMS, and submit pictures of myself and things”.


5.1.4 Type 4: debaters (11%)

Debaters are as high as socialisers in terms of participation level, characterised by being highly involved in discussions, reading, and writing contributions in general (see Table 2). In addition, this participation mode is related to a more informal practice as suggested in Figure 2. The debating pattern is unequally distributed among the communities: very frequent on Underskog.no and practically absent on Biip.no. The users are somewhat older than the medians of other user types, and there is practically no gender difference. A typical ‘debater’ is odd, a 42 year old who lives in Oslo. He has been a member of Underskog.no for more than a year. He is university educated and likes to discuss and express himself in writing. He depends on the internet for carrying out practical tasks and uses it mainly for instrumental reasons. He considers video contribution too time-consuming. He also regards his SNS as being more text- and picture-related than video-focused. He uses the SNS to keep updated on cultural events and new publications, as well as for social contacts, explaining this as follows: “I get informed about events, publications, and Net experiences; at the same time I am making bonds and having discussions with other people”.

5.1.5 Type 5: actives (18%)

‘Actives’ are so labelled because these users are engaged in almost all kinds of participation activities within the community, which includes being a member to “publish and share pictures” (see Table 2). The majority in this group are young females, as shown in Table 3. This user type is distributed unequally among the SNSs, being most common on Nettby.no and HamarUngdom.no. A typical Active user is Lena, a 16 year old who lives in the countryside. She has been a member of the HamarUngdom.no community for more than three years and is still very active. During the week before she answered the questionnaire, she was in contact with 20 people, almost everybody in her profile. When she is logged on, she engages in many activities, usually related to events and publishing music or videos. She uses the SNS mainly for social contact and communicates primarily with friends in her profile. She contributes videos and summarises her activity in this way: “I wish I had my own video camera. So far, I am just uploading and sharing videos that others have made”. Her willingness to be even more involved in UGC participation is high.

5.2 User types and social capital

Table 5 (as well as Figure 2) answers RQ4 and shows that actives are most active in the social field as well. Socialisers and debaters also have high activity here; they frequently have daily contact with their friends, and 10% have more than 150 friends in their profile. Socialisers and actives are also more frequently in touch with their family on a daily basis.

Further, debaters and actives more often report that political expression within the SNS is ‘very important’ to them. These user types view their SNS as a significant arena for community participation in terms of political activity. On the other hand, lurkers and sporadics are less interested in this type of activity. This result also indicates that the highest level of social capital inside an SNS is associated with actives, socialisers, and debaters.


Table 5 Social capital: social contact and civic activity within user types and between different user types in %

User types More than

150 contacts in user profile

Daily contact family

Daily contact friends

Contact with more than 30 people last

week on SNS

SNS is very important to me for political expression

Socialisers 10 13 61 6 10 Debaters 10 8 57 5 20 Lurkers 9 5 45 2 6 Sporadics 5 10 41 4 7 Actives 16 15 69 12 18

5.3 User types and UGC contribution

Table 6 presents results of the answers to RQ3 – an overview of the different user types in regard to the amount of UGC they contribute (several times a week and daily, added together) and the difference among the user types in terms of their willingness to publish videos on an SNS in the future. Table 6 Contribution level of UGC (several times a week or daily) within user types and

differences between user types in %

User types Text Photo Video Willingness to publish videos in the future

Socialisers 33 16 2 27 Debaters 55 12 5 34 Lurkers 26 7 5 22 Sporadics 27 8 2 20 Actives 51 24 6 40

Debaters are the most active group in terms of the contribution of text. This is not surprising, given that discussion seems to be their passion inside the community. However, actives and socialisers also actively contribute text. Lurkers are less active and are more of an observant type. Actives are most active in their photograph (24%) and video (6%) contributions; actives are also significantly more willing to publish videos in the future. It should be noted that reported video contribution is quite uncommon among all user types. In general, the highest level of participation in relation to UGC is seen among actives; debaters have the second highest level, and socialisers third. Lurkers and sporadics participate less. These results validate the typology.

6 Discussion

6.1 Five user types and participation inequality

Five distinct types of users have been identified based on their SNS behaviour patterns in terms of participation level and participation mode in four SNSs. The user types in this study reflect substantial differences in the underlying patterns of SNS usage, as well as


preferences and reasons for participating in a community. The actives engage in more or less all activities within the SNS community, whereas the debaters interact with others for the purpose of discussion. The socialisers primarily use the sites for communication and ‘small-talk’ activities, while the sporadics are, as are all other users, ‘socially curious’ in that they sporadically check to see whether anybody has contacted them. The lurkers might be the only user group that is somewhat asocial, as these users seem to log on to the sites for ‘time-killing’ and consuming entertainment rather than social interaction. Together with socialisers, lurkers is the least active in contributing text and photographs.

In our study, we found that the user population, not surprisingly, is made up of the younger part of the Norwegian population; this age skew is quite typical for most SNSs (e.g., Pfeil et al., 2009). Nevertheless, the types display large differences in usage patterns, reflecting different levels of motivation for participation. As expected, two main types of motivation for social participation in SNSs stand out. The first type is related to a recreational mode in terms of informal chatting, whereas the other is a more informational mode related to serious debate and discussion. Although both genders and most age groups are represented in both modalities, younger girls are more typically socialisers, while older boys and men are typically debaters. These modalities have existed on the internet for many years, supported by informal e-mailing, chat channels such as IRC and MSN, and blogs. Nowadays, these forms of interaction are being merged into a single type of system: the SNS. This also underscores the fact that we see a convergence of different types of applications in these systems such as mail, blog, chat, gaming, and homepage into one integrated service.

6.2 User types differences in diverse SNSs

The four SNSs in this study vary with respect to these two modalities of participation, highlighting the fact that SNS characteristics and design are important factors that influence and support different user types. In our study, all the different user types are found inside the diverse SNSs, but the socialisers are more frequent in the Biip.no community, whose members are younger, while Underskog.no is more often represented by debaters. Underskog.no has more highly educated users and a higher mean age than the other SNS. Underskog.no is also an SNS where members who share content are very culturally and academically oriented. In addition, Underskog.no is a very small SNS. The importance of SNS size relative to the level of participation among users has been suggested by others (e.g., Butler, 2001; Preece and Maloney-Krichmar, 2003). Smaller and locally bounded SNSs seem to involve more enthusiastic users who are willing to share and participate more frequently compared to, for example, YouTube (e.g., Nielsen, 2006), which is discussed in the next section.

The three other SNSs (i.e., HamarUngdom.no, Biip.no, and Nettby.no) have a greater mix of user types. This might be explained by the fact that these sites are free and open to everyone. These SNSs have no specific interest target and, therefore, may also have a more diverse distribution of user types. For example, Nettby.no is associated with the largest online newspaper in Norway and recruits a wide spectrum of users through this connection.


6.3 Theoretical implications

This paper draws upon theoretical threads to create a framework for a typology of SNS users according to their behaviour. To date, a typology of SNS users has not been explored in a way that draws upon a large source of SNS behavioural data combined with theoretical insights from Kozinets’s (1999) marketing model and Rogers’s (2003) adoption and innovation model. These and other studies should be compared to give more insight into the connections between the different user types presented (see Table 7 for a comparison). Table 7 Connections between different models of user types

This study’s typology

Kozinets’s typology

Nielsen’s ‘typology’

OFCOM typology

Rogers’s model Justification

Socialisers Minglers Intermittent contributors

Faithfuls Early majority Use social media in particular; can be viewed as the early majority since they may be seen as open to new ideas, active in the community, and influencing their neighbours.

Debaters Devotees Intermittent contributors

Functionals Early adopters Bloggers and debaters in SNSs can also be viewed as early adapters because of interests in A/V UGC.

Lurkers Tourists Lurkers (Late majority)

Account for the biggest user type in SNSs and in regard to UGC in general. Include people using media for lurking or time-killing.

Sporadics Alpha socialisers

and followers

Late majority (and laggards)

Newcomers and sporadic users of the particular media studied.

Actives Insiders Heavy contributors

Attention seekers

Innovators and early adopters

Use media frequently and are advanced compared to the rest of the user population.

The previous user typologies have been shown to connect in some ways to our meaningful categories of user types. This might suggest that previous research on typologies can help us understand and categorise the increasingly complex behaviours found on SNSs (see Table 7). Kozinets’s (1999) market segmentation model seemed to be the most similar typology when compared to the different user types in our SNS typology. However, Kozinets’s typology is related to the marketing and consumption of products or the provision of information about goods inside virtual communities, rather than being a general user typology for SNS. As Table 7 shows, we included two OFCOM types, both in the category of sporadics. It was not appropriate to identify lurkers in the OFCOM study. There may be a connection between attention seekers and actives,


because the latter type of users was the most eager to publish photographs (24%) in our study, as shown in Table 6.

Rogers’s categories are related to the time needed for different consumers to adopt (technological) innovations. Yet the present study identified user types by examining SNS usage in general, whether or not the adoption of innovations was involved. Furthermore, given that SNSs are themselves an innovation, our study investigated the behaviour of people who have already adopted an innovation. The application of different criteria for the identification of segments limits the possibilities for drawing comparisons between Rogers’s study and this one. However, the degree of use and the general usage pattern of the technology are important variables that describe the extent of diffusion of the innovation (Shih and Venkatesh, 2004). From this perspective, it is possible that innovators are present among the actives in our typology. Early adopters may be mainly associated with actives and debaters. If we use education as a criterion, debaters may be the most obvious category to associate with early adopters, because debaters are the highest educated user type. Finally, although the late majority and the laggards might not be using SNS at the moment, if we have to identify one of our types that should be associated with these groups, it would be the sporadics. However, the most likely situation is that laggards and the late majority have not yet joined SNS communities. These categories are given in brackets in Table 7. Nielsen’s (2006) typology is more related to a participation related to UGC contribution level in total and will be discussed in the next section.

6.4 Participation inequality: A 50-30-20 rule in locally bounded community

Based on our findings of participation in terms of UGC level of contributions, we would suggest a 50-30-20 rule for participation inequality:

• 50% explains the total of sporadics (19%) and lurkers (27%) and their overall passive participation level. These two user types contributed little to the community activity, and are users who more or less correspond with Nielsen’s (2006) lurkers.

• 30% of our sample, counting debaters and socialisers together, corresponds to Nielsen’s (2006) ‘intermittent contributors’, which contribute from time to time.

• 20% refers to the actives in our sample, and corresponds correspond with Nielsen’s (2006) ‘heavy contributors’. Actives are the heaviest contributors and account also for most of the action in the SNSs.

In comparison, Nielsen’s (2006) study reported lurkers to represent 90%, indicating a significant difference between our study, with a 50-30-20 participation, and the 90-9-1 rule presented by Nielsen. This difference might be explained by the fact that users in our sample mainly contribute with textural comment and amateur photos inside smaller and locally bounded communities, and not with videos on YouTube or encyclopedia articles in Wikipedia. Hence, the 90-9-1 rule might still be applicable to large, open communities such as YouTube or Wikipedia but not for smaller and locally bounded SNS communities. The latter communities obviously have lower thresholds for contribution, compared to high-profile sites such as Wikipedia or YouTube.

Another explanation for the, in general, higher level of participation in locally bounded communities compared to large-scale networks such as YouTube is that users share content with their peers and not with complete strangers. For example, in previous


studies, social presence dimensions have been associated with usage patterns (e.g., Shen and Khalifa, 2009). Similarly, social presence indicators were found to be positively correlated with the level of tagging in Flickr (Nov et al., 2008). In our study and in the SNS where video-sharing is technically supported (Underskog.no), we see that as many as 7% cite the purpose of their visit as uploading a video and that 25% contribute once a month or more. These findings suggest a high level of participation in small and local SNSs, even with video contribution, probably caused by experienced social presence in such sites.

6.5 Migration of user types?

Different modes and levels of participation might reflect different levels of skills and advancements in how users use SNSs. We might suggest that, over time, people develop more advanced communication skills or user type patterns in SNSs, as most people start out as sporadics when they are introduced to SNS, and in time progress to a higher level of usage, either as socialisers, debaters, or actives. For instance, the debater group is typically characterised by older users compared to the users among sporadics and socialisers. Future research should try to confirm such a migration and investigate how and in which directions these patterns develop over time.

6.6 Limitations

The generalisability of the present research is limited by

a the geographic scope of the sample (Norway)

b the more or less self-selected sample.

With respect to a, the Norwegian population is among the most eager SNS users in the world, and Norway is, according to Facebook, “the biggest Facebook nation in the world” when adjustments have been made for population size (Brandtzæg and Lüders, 2008). Therefore, our Norwegian sample should be regarded as interesting. Still, further research should be conducted that takes into account other nations as well, to identify any cross-nation variations. With respect to b, the SNS members who participated in our study were recruited through a link on the sites rather than through random sampling across the population and are mainly represented by young people. In addition, girls are overrepresented. This method of selection threatens the external validity of the study because we do not have full control over possible sources of systematic biases in our sample. However, this weakness of the study may be partially overcome, because the participants were recruited from four different SNSs. This constitutes an improvement over previous studies (Ellison et al., 2007) in which only one community was studied. We also tried to involve a broader segment of SNS users by rewarding participation in the survey with an entry to a raffle for which the prize was a travel voucher; this was thought to appeal to a wide range of people. However, except for the potential bias of self-selection discussed, we cannot see any other factors that should reduce the representativeness of the sample with respect to age or gender. Most SNSs are skewed in life stage/age and gender toward a majority of younger people (Pfeil et al., 2009) and females participating (Hargittai, 2007), which is reflected in our data sample.


Further, the identification of the user typology in SNSs merely reflects a snapshot of user behaviour taken at a particular time, whereas usage behaviour and SNS options are constantly changing. The boundary between one user type and another is also fluid; some users are at the cluster centre, while others may be on the border between two or even three clusters. However, the qualitative data collected from typical users at the cluster centres validate the results. A validation of the typology was also reflected in the results reporting statistically significant differences in UGC production and social contact among different user types.

Finally, our cluster analysis does not produce a definitive solution; instead, the analyst usually has a choice among several probable solutions. Our solution with five clusters resulted in groups of adequate sizes – no fewer than 11% of the cases and a maximum of 27% of the cases (see Table 2). Five and four clusters are common when constructing a typology. In addition, the results depend on the set of questions asked. Other criteria may also be useful, and it is not claimed here that all possible user behaviour is described. However, the alternative ‘other’ was not chosen very frequently among the sample participants.

7 Conclusions

This study has outlined a proposed typology that can be used to describe the ways in which SNSs users participate. SNSs have become a predominant environment for social interactions and involvement online in the past few years, which might indicate new forms of participation inequality or a new digital divide. However, few quantitative studies have been performed that extensively explore usage behaviour in SNSs. The user typology identified in this study will help provide a nuanced understanding of an increasingly fragmented population of SNS users and allows much more subtlety in targeting SNS users, rather than simply lumping them all together in a single category. Our findings contribute to a new way to identify participation inequality. A typology will aid the design of successful SNSs by enabling the requirements of different user types to be formulated. A user typology could help requirement engineers identify stakeholders in their systems during the requirement analysis and thus include the diversity of users and their requirements. An SNS that desires to nurture a particular user type can be designed more successfully. For example, debaters could be nurtured by designing an SNS that allows members to share and discuss different topics more easily. In other words, SNS companies will need to craft a different design strategy if they want a target audience with a higher proportion of socialisers than debaters.

By revealing the distinct characteristics of the user types, we have shown not only the differences in usage behaviour and degree of motivation but also how this pattern is related to different kinds of SNSs and the genders and ages of the sites’ members. In terms of participation inequality, we also suggest a new 50-30-20 rule for participation in smaller and locally bounded online communities. Further, our findings could contribute significantly to the evaluation of SNSs. By being aware of and understanding different user types, we will be able to spot the roots of existing problems or seeds of success among different user types.

Further, it is important for the information and communication technologies (ICT) industry to determine how different patterns of SNS use are linked to different user models, so that new SNS services can be personalised and public services can overcome


the digital divide by promoting services that target different user types. Finally, a user typology can help scholars see how various types of usage are associated with different social outcomes.

7.1 Future research

Our study opens up a number of avenues for further research. First, a study should be conducted to determine whether our findings about the robustness of the SNS user typology are correct. A slightly modified questionnaire-tool can be found online (http://www.tinyurl.com/6xxobl), which makes it possible for others to reproduce the study. Second, the user typology should be analysed in a long-term study, as both technology options and usage behaviour may evolve over time. For example, meta-analyses of computer-mediated communication indicate that internet users progress from initially asocial information gathering toward more social activities (Walther et al., 1994). Third, interviews could be conducted or observations made in an attempt to cross-validate our findings. Finally, the investigation of user typologies in SNSs as such typologies apply to SNSs in different countries would be worthwhile to identify cross-cultural differences.

Acknowledgements

The research leading to these results received funding from the CITIZEN MEDIA project (038312) FP6-2005-IST and the RECORD-project (180135/S10), supported by the VERDIKT programme of the Research Council of Norway. The authors would like to thank all the SNSs and the users’ participation in this study.

References Barnes, S.J., Bauer, H.H., Neumann, M.M. and Huber, F. (2007) ‘Segmenting cyberspace: a

customer typology for the internet’, European Journal of Marketing, Vol. 41, No. 1, pp.71–93. Bishop, J. (2007) ‘Increasing participation in online communities: a framework for

human-computer interaction’, Computers in Human Behaviour, Vol. 23, No. 4, pp.1881–1893. Blashfield, R.K. and Aldenderfer, M.S. (1978) ‘The literature on cluster analysis’, Multivariate

Behavioral Research, Vol. 13, No. 1, pp.271–295. Boyd, D. and Ellison, N.E. (2007) ‘Social network sites: definition, history and scholarship’,

Journal of Computer-Mediated Communication, Vol. 13, No. 1, article 11, available at http://www.jcmc.indiana.edu/vol13/issue1/boyd.ellison.html.

Brandtzæg, P.B. (2010) ‘Towards a unified media-user typology (MUT): a meta-analysis and review of the research literature on media-user typologies’, Computers in Human Behavior, Vol. 26. No. 5, pp.940–956.

Brandtzæg, P.B. and Heim, J. (2008) ‘User loyalty and online communities: why members of online communities are not faithful’, INTETAIN, Proceedings of the 2008 ICST Second International Conference on Intelligent Technologies for Interactive Entertainment, 8–10 January, ACM Digital Library, Playa del Carmen, Cancun, Mexico, available at http://www.portal.acm.org/citation.cfm?id=1363200.1363215.

Brandtzæg, P.B. and Lüders, M. (2008) ‘eCitizen 2.0: the ordinary citizen as a supplier of public-sector information’, Report for the Department of Modernisation and Administration in Norway, SINTEF, Oslo, Norway, available at http://www.hellotxt.com/l/G6gh.


Brandtzæg, P.B., Heim, J. and Kaare, B.H. (2010) ‘Bridging and bonding in social network sites – investigating family-based capital’, Int. J. Web Based Communities.

Brandtzæg, P.B., Heim, J., Karahasanović, A., Skjetne, J., Dahl, H., Warholm, L., Vanattenhoven, J., Nielsen, T.K., Vermeir, L., Pierson, J., Lievens, B. and Jacobs, A. (2009) Non-professional Users as Co-creators of Media Content, P6-2005-IST-41 (038312), D1.1.3.

Bunn, M.D. (1993) ‘Taxonomy of buying decision approaches’, Journal of Marketing, January, Vol. 57, pp.38–56.

Butler, B.S. (2001) ‘Membership size, communication activity, and sustainability: a resource-based model of online social structures’, Information Systems Research, Vol. 12, No. 4, pp.346–362.

Ellison, N., Steinfield, C. and Lampe, C. (2007) ‘The benefits of Facebook ‘friends’: exploring the relationship between college students’ use of online social networks and social capital’, Journal of Computer-Mediated Communication, Vol. 12, No. 4, pp.1143–1168.

Gore, P.A. (2000) ‘Cluster analysis’, in Tinsley, H.E. and Brown, S.D. (Eds.): Handbook of Applied Multivariate Statistics and Mathematical Modelling, Academic Press, New York.

Hargittai, E. (2007) ‘Whose space? Differences among users and non-users of social network sites’, Journal of Computer-Mediated Communication, Vol. 13, No. 1, article 14, available at http://www.jcmc.indiana.edu/vol13/issue1/hargittai.html.

Heim, J., Brandtzæg, P.B., Endestad, T., Kaare, B.H. and Torgersen, L (2007) ‘Children’s usage of media technologies and psychosocial factors’, New Media & Society, Vol. 9, No. 3, pp.49–78.

Horrigan, J.B. (2007) ‘A typology of information and communication technology users’, Pew Internet Report, May, USA.

Johnsson-Smaragdi, U. (2001) ‘Media use styles among the young’, in Livingstone, S. and Bovill, M. (Eds): Children and their Changing Media Environment: A European Comparative Study, Lawrence Erlbaum Associates, Mahwah, NJ, London, pp.131–141.

Kollock, P. and Smith, M. (1996) ‘Managing the virtual commons: cooperation and conflict in computer communities’, in Herring, S. (Ed.): Computer-Mediated Communication: Linguistic, Social, and Cross-Cultural Perspectives, John Benjamins, Amsterdam, pp.109–128.

Kozinets, R. (1999) ‘E-tribalized marketing? The strategic implications of virtual communities of consumption’, European Management Journal, Vol. 17, No. 3, pp.252–264.

Kraut, R., Patterson, M., Lundmark, V., Kiesler, S. et al. (1998) ‘Internet paradox – a social technology that reduces social involvement and psychological well-being?’, American Psychologist, Vol. 53, pp.1017–1031.

Nielsen, J. (2006) ‘Participation inequality: encouraging more users to contribute’, Jakob Nielsen’s Alertbox, 9 October 2006, available at http://www.useit.com/alertbox/participation_inequality.html (accessed on 27 May 2009).

Nov, O., Naaman, M. and Ye, C. (2008) ‘What drives content tagging: the case of photos on Flickr’, Proceedings of CHI 2008: ACM Conference on Human Factors in Computing Systems, April, Florence, Italy, pp.1097–1100.

OFCOM (2008) ‘Social networking. a quantitative and qualitative research report into attitudes, behaviours and use’, Office of Communication, London.

Ortega Egea, J.M., Menéndez, M.R. and González, M.V.R. (2007) Diffusion and Usage Patterns of Internet Services in the European Union Information Research, Vol. 12, No. 2, available at http://www.informationr.net/ir/12-2/paper302.html (accessed on 27 May 2009).

Pfeil, U., Arjan, R. and Zaphiris, P. (2009) ‘Age differences in online social networking – a study of user profiles and the social capital divide among teenagers and older users in MySpace’, Computers in Human Behaviour, Vol. 25, No. 3, pp.643–654.

Preece, J. (2000) Online Communities: Designing Usability, Supporting Sociability, Wiley, Chichester.

Preece, J. and Maloney-Krichmar, D. (2003) ‘Online communities’, in Jacko, J. and Sears, A. (Eds.): Handbook of Human-Computer Interaction, pp.596–620, Lawrence Erlbaum Associates Inc. Publishers, New York.


Preece, J. and Shneiderman, B. (2009) ‘The reader-to-leader framework: motivating technology-mediated social participation’, AIS Transactions on Human-Computer Interaction, Vol. 1, No. 1, pp.13–32.

Quan-Haase, A. and Wellman, B. (2004) ‘How does the internet affect social capital?’, in Huysman, M. and Wulf, V. (Eds.): IT and Social Capital, MIT press, New York.

Rogers, E.M. (1962, 2003) Diffusion of Innovations, Free Press, New York. Shah, D., Kwak, N. and Holbert, R. (2001) ‘‘Connecting’ and ‘disconnecting’ with civic life:

patterns of internet use and the production of social capital’, Political Communication, Vol. 18, No. 2, pp.141–162.

Shen, K.N. and Khalifa, M. (2009) ‘Design for social presence in online communities: a multidimensional approach’, AIS Transactions on Human-Computer Interaction, Vol. 1, No. 2, pp.33–54.

Shih, C. and Venkatesh, A. (2004) ‘Beyond adoption: development and application of a use-diffusion model’, Journal of Marketing, Vol. 68, No. 1, pp.59–72.

Tan, P., Steinbach, M. and Kumar, V. (2006) Introduction to Data Mining, Pearson, Addison Wesley, New York.

Walther, J.B., Anderson, J.F. and Park, D. (1994) ‘Interpersonal effects in computer-mediated interaction: a meta-analysis of social and anti-social communication’, Communication Research, Vol. 21, No. 2, pp.460–487.

Wellman, B. and Gulia, M. (1999) ‘Virtual communities as communities: net surfers don’t ride alone’, in Smith, M.A. and Kollock, P. (Eds.): Communities in Cyberspace, pp.167–194, Routledge, New York, NY.