GEN499: General Education Capstone

Behaviour & Information Technology, 2015 Vol. 34, No. 4, 330–340, http://dx.doi.org/10.1080/0144929X.2013.833650 How access gaps interact and shape digital divide: a cognitive investigation Shahla Ghobadi a∗ and Zahra Ghobadi b a Australian School of Business, University of New South Wales, Sydney, NSW 2052, Australia; bSchool of Computer Science and Information Systems, IKI University, Iran (Received 21 December 2012; accepted 7 August 2013 ) Inequalities in the access to and use of information and communication technologies (ICT) have become an important area of concern for over a decade. Yet, theoretical attempts to understand the dynamics behind shaping these inequalities are scarce.

This study draws upon the extent literature on digital divide and explains how the four diﬀerent access gaps (motivational, material, skills, and usage) interact and contribute to digital divide. Revealed causal mapping (RCM) is utilised to analyse the data collected from eight same-gender focus groups in four primary schools located in Iran. The revealed causal map demonstrates the interaction and linkages between diﬀerent access gaps. The ﬁndings provide a theoretical account of the dynamics behind shaping digital divide and generate insights into extending the concepts of access gaps. We establish a theoretical model that places an emphasis on the centrality of motivational-related factors such as ‘lack of interest in IT-related things’ and ‘lack of motivation to learn recent technology’ as well as skills-related factors such as ‘operating skills’, ‘anti-ﬁltering skills’, and ‘lack of IT background’.

Keywords: digital divide; Internet; access; access gaps; qualitative methodology; revealed causal mapping 1. Introduction Similar to many historical transitions, it has been argued that the Internet is the source of several beneﬁts as well as chal- lenges (DiMaggio and Hargittai 2001, Norris 2003). One of the important challenges pertaining to the phenomenon of the Internet is the so-called concept of digital divide (Gunkel 2003, Chen and Wellman 2004, Barth and Veit 2011, Van Dijk 2012). The key concern in digital divide research and policies is the growing gap between individuals, groups, and nations, which is due to an unequal allocation of Infor- mation and Communication Technologies (ICT) access and use (Vicente and Lopez 2010, Barth and Veit 2011,We i et al. 2011).

Digital divide literature has attempted to understand dig- ital divide outcomes and digital divide antecedents (factors that shape digital divide) (Mescha and Talmudb 2011, We i et al. 2011). In terms of outcomes, it is believed that digital divide contributes to several inequalities such as immate- rial, material, social, and educational types of inequality (Katz et al.2001, Katz and Rice 2002, Rice and Katz 2003, Van Dijk and Hacker 2003). The underlying reasons behind the growing digital divide (‘digital divide antecedents’) could be related to several factors such as international politics, inequality of positions, and power in social networks, and inequalities of skills, capabilities, and interests (Bonfadelli 2002, Cho et al. 2003, Chen and Wellman 2004, Rains 2008,Agarwal et al. 2009). In spite of calls for theoretical works that explore the underlying mechanisms shaping digital divide (Van Dijk 2006), attempts to understand these mechanisms are scarce.

In particular, most of the research on digital divide has remained at a descriptive level, emphasising the demo- graphics of income, education, age, gender, and ethnicity on material access to ICT (Hoﬀman et al.2000, Kalichman et al. 2002, Valadez and Durán 2007). A systematic inves- tigation of the interrelation of the underlying concepts behind shaping digital divide has yet to receive signiﬁcant attention. To address this void, this study seeks to develop a sound theoretical understanding of the dynamics behind digital divide. We build upon the theoretical lens of access gaps proposed by Van Dijk (2005, 2006). The model of access gaps explains the cumulative role of four types of access gaps including motivational access, material access, skills access, and usage access in shaping digital divide (Van Dijk and Hacker 2003, Van Dijk 2005). While this model provides an appropriate theoretical grounding for linking access gaps to digital divide (Hargittai and Hinnant 2008, Hargittai 2010, Van Deursen and Van Dijk 2011), it does not explain the potential interrelations between access gaps (e.g. if there is a link between ‘skills access’ and ‘moti- vational access’) as well as how these gaps interact and shape digital divide as a whole. For this, we seek to build upon the model proposed by Van Dijk (2005) and expand the focus to understand the ∗Corresponding author. Email: [email protected] © 2013 Taylor & Francis Behaviour & Information Technology331 dynamic interrelations between diﬀerent types of access gaps and digital divide. In particular, the research question that guides our enquiry asks: ‘how diﬀerent types of access gaps interact and shape digital divide?’ A qualitative methodology, revealed causal mapping (RCM) (Narayanan and Fahey 1990), is employed to dis- cover and analyse cognitive structures related to access gaps and interrelations among them. RCM is a subcategory of cognitive mapping where respondents reveal their causal assertions about a domain-speciﬁc phenomenon through interview or focus groups (Nelson et al.2000a, 2000b).

This methodology lends itself into construction of revealed causal maps that represent network of causal relations that are embedded in explicit statements. Thirty-two partic- ipants discussed access barriers over eight same-gender focus groups. The focus groups’ processes were guided by the model of access gaps proposed by Van Dijk (2005), but they were designed to evoke new concepts and linkages. Analysis on the collected data leads to identifying new concepts as well as causal linkages between access gaps and related concepts. We provide two valuable theoretical con- tributions. First, this study establishes a theoretical model that advances our understanding of the dynamics shaping digital divide. Second, the undertaken approach contributes to a more subtle understanding of the nature of diﬀerent access gaps. Speciﬁcally, access gaps are not independent concepts, but they are shaped through complex dynamic interaction with each other. Our results help individuals and policy-makers have a better understanding of the underlying mechanisms through which access gaps (1) are implicated and (2) contributed to inequalities in the access to and use of ICT. The remainder of this paper is organised as follows.

Section 2discusses the digital divide literature and provides the theoretical grounding for this study. Section 3outlines the research methodology including the data collection and analysis procedures. Results of data analysis are outlined in Section 4. We discuss research and practical implications in Section 5, prior to stating research limitations in Section 6, and oﬀering directions for future research in Section 7.

2. Theoretical background 2.1. Digital divide and access gaps The term digital divide was ﬁrst coined in an oﬃcial pub- lication by the US Department of Commerce’s National Telecommunications and Information Administration in the middle of the 1990s (Van Dijk 2006). Initially, digital divide referred to the inequality of technological opportunities, in terms of the gap between ‘those who do and those who do not have access to new forms of information technology’ (Cammaerts et al. 2003). The technological orientation of this early digital divide deﬁnition led to several attempts in the equalisation of technological opportunities in the form of physical access to the ICT (Van Dijk 2006, Hohlfeldet al. 2008). The use of the term digital divide at the turn of the twenty-ﬁrst century highlighted ‘digital inequality in the Information Society’ on the political and academic agenda.

From 2001 onwards, the extant literature suggested the appearance of new expressions such as ‘redeﬁning the dig- ital divide’ and ‘going beyond physical access’ (Hargittai 2002, Gurstein 2003, Selwyn 2004). These expressions aim to reframe the overly technical concept of the digital divide and to pay speciﬁc attention to social, mental, and cul- tural factors that contribute to inequalities in the access to and the use of ICT (Warschauer 2002, Valadez and Durán 2007). In other words, technology access should be seen as a process that is driven by several social, psychologi- cal, and technological factors. As a result, digital divide gradually became understood as ‘the gap between “indi- viduals”, “households”, “businesses”, and “geographical” areas at diﬀerent socio-economic levels with regard to both their opportunities to access information and communica- tion technology and to their use of Internet for a variety of activities’ (OECD 2001, p. 5). This deﬁnition refers not only to the ICT access but also to the ability to develop an appro- priate use of the technology. For example, two individuals might have equal access to ICT, but limited IT skills can impede one from the appropriate and strategic use of tech- nology. Van Dijk (1999) proposed the multifaceted concept of access, and (2005) oﬀered a cumulative and recursive model of successive kinds of access to digital technolo- gies. This model (Figure 1) explains how digital divide is shaped by four divides of motivational, material, skills, and usage access. As shown, material access proceeds by moti- vational access and succeeds by skills access and usage access (Van Dijk 2005). The successive aspect of the model suggests that the eﬀective access to technology is depen- dent on four types of access, including motivational access, material access, skills access, and usage access. Accord- ingly, digital divide is the result of gaps in these four access areas. The recursive aspect of the model suggests that when the full process of technology appropriation is completed (usage access), a new innovation arrives and the process starts again. The concept of ‘motivational access’ refers to the wish to have a computer and to be connected to the ICT. The fac- tors explaining motivational access divide could be both of a social or cultural and a mental or psychological nature (e.g.

low levels of income, low levels of education, computer Figure 1. Access, digital divide, and outcomes (Van Dijk 2005). 332S. Ghobadi and Z. Ghobadi anxiety, and lack of time). The concept of ‘material access’ comprises physical access and other types of access that are required to reach complete disposal and connections such as conditional access (subscriptions, accounts, and pay-per-view). The major contributing factors to ‘mate- rial access’ can include: income, education, and occupation.

‘Material access’ is succeeded by having motivation to have ICT. The concept of skills access is divided into possessing three types of skills: (1) operational skills: the capacities to work with hardware and software, (2) information skills:

skills to search, select, and process information in computer and network sources, and (3) strategic skills: capacities to use computer and network sources as the means for par- ticular goals and for the general goal of improving one’s position in society (Van Deursen and Van Dijk 2009). Skills access succeeds by motivation to use ICT and access to ICT (through try and error work with ICT). Education is a critical factor that can aﬀect ‘skills access’. The concept of ‘usage access’ is about the diﬀerential use of ICT applications in daily practices. This could include both the actual use of ICT as well as ‘active versus passive use of ICT’. Active or creative use of ICT is about contributions to the Internet (e.g. publishing a personal website, creating a weblog, posting a contribution on an online bulletin board, and newsgroup or community). Usage is largely linked to demo- graphic characteristics of users and technical connections (e.g. social class, education, age, gender and ethnicity, and eﬀectiveness of the connection). ‘Usage access’ is also suc- ceeded by the motivation to use ICT, material access, and having appropriate skills. The general impression is that while physical access gaps are more or less closing in the developed countries, the skills gap (in particular, information skills) and usage gaps are growing (Mason and Hacker 2003, Van Dijk 2012). It should be taken into account that developing countries such as Iran and China still experience physical access gaps due to several factors such as low-speed Internet and limited access (Giroux 2009, Ameripouret al.2010, Cross 2010).

2.2. Interrelations between access gaps There are few studies that point to the existence of interac- tions between access gaps (Barzilai-Nahon 2006, We iet al.

2011). For example, based on the concept of diﬀerent types of access, We iet al. (2011) studied digital divide among school students by following the three-level digital frame- work proposed by Dewan and Riggins (2005). We iet al.

(2011) established a model that links three hierarchical lev- els of digital access divide (equivalent to the concept of ‘material access’), digital capability divide (equivalent to the concept of ‘skills access’), and digital divide outcome (knowledge outcome and skills outcome). Their ﬁndings shed light on the relationship between ‘material access’ and ‘skills access’. More speciﬁcally, their results showed the relationship between digital access divide and digital capability divide (e.g. students without home computers had lower computer self-eﬃcacy even when they had IT access in schools) as well as the relationship between digital capability divide and digital outcome divide (e.g. stu- dents with lower computer self-eﬃcacy had poorer learning outcomes).

Another example that points to the existence of interac- tions between access gaps is the literature on operational- ising and measuring digital divide (Barzilai-Nahon 2006, Vehovar et al.2006). For example, Barzilai-Nahon (2006) drew attention into causal interrelations between various indicators of digital divide (e.g. socio-demographic, acces- sibility, use, infrastructure access, aﬀordability, and social and governmental support). As an example, Barzilai-Nahon asserted that accessibility (‘material access’) may aﬀect Digital Divide Index directly, but it also has indirect impacts through use indicator (‘usage access’). Causal relations between indicators of digital divide concur with the concept of dynamic interactions between access gaps. While such studies point out interactions between access gaps, knowledge they provide is still fragmented with limited analyses based on and contributing to rigorous the- oretical perspectives. Against this backdrop, we focus our enquiry by asking ‘how diﬀerent types of access gaps (moti- vational, material, skills, and usage) interact and shape digital divide?’ We address this research question to develop a theoretical account that explicates the mechanisms that generate digital divide (discussed in the Result section).

3. Research methodology We employed a qualitative methodology, RCM (Narayanan and Fahey 1990), to investigate cognitive structures that explain how access gaps interact and shape digital divide.

RCM is a variant of cognitive mapping where respondents reveal their causal assertions about a phenomenon through interviews or focus groups (Narayanan and Fahey 1990).

RCM demonstrates the network of causal relationships that are embedded in explicit statements of individuals about their environment (Nelson et al.2000a) and, it is a well- recognised method among Information Systems researchers for studying cognitive perceptions in diﬀerent settings such as gender studies, knowledge management, and software development (Nelson et al.2000a, 2000b, Armstrong et al.

2007). RCM is therefore a suitable method for revealing the network complexity of the causal interrelations between diﬀerent access gaps. We use empirical descriptions from eight focus groups to provide inputs for analytical generaliaation (Yin 2009) or in another words generalising from empirical material to theory (Lee and Baskerville 2003). To ensure the quality of the descriptions and the validity of our generalisations, we follow Yin (2009) recommendations and use procedures that involve multiple sources of evidence as well as member checking. In terms of multiple sources of evidence, we col- lected data from 32 participants within four similar contexts (public primary schools), with 16 females and 16 males. Behaviour & Information Technology333 Following the guidelines proposed by Nelsonet al. (2000a), a three-step process was followed to evoke the revealed causal map. These steps were: (1) data elicitation (site and sample selection), (2) construction of revealed causal map, and (3) analysis and validation of the revealed causal map.

3.1. Data elicitation An empirical investigation of four primary schools (School A, School B, School C, and School D) in Iran was con- ducted. Data were collected through focus groups sessions, where individuals discussed the dynamics through which diﬀerent access gaps shape digital divide. In particular, 32 individuals participated from four pri- mary schools. Together with the Principal Manager of each school eight students were identiﬁed and recruited (eight students per school). In the following, either their mother or father was chosen for this study (random choice). Per school, four men and four women participated. There- fore, 16 men (from four schools) and 16 women (from four schools) participated in this study. Participation was voluntary and participants were assured of conﬁdentiality. After recruiting participants, one of the researchers con- ducted two same-gender focus groups in each school (in total: eight focus groups in four schools). Each session lasted from 60 to 90 minutes. The language of the focus groups was Farsi. The sessions were recorded and tran- scribed into a document. One of the researchers, ﬂuent in both Farsi and English, translated the documents to English. The length of the documents ranged from 5335 to 8343 words. Both researchers were ﬂuent in Farsi and this facilitated data collection and analysis. In average, respondents were 39.35 years old (females:

36.3 years old and males: 42.4 years old); 22% of the respon- dents had diploma, 49% had undergraduate education, and 29% had postgraduate education. The patterns of education in females and males were diﬀerent – 67% of the females had at least undergraduate education, whereas more than 90% of males had at least an undergraduate degree. The demographics (e.g. gender, education, and age) suggested that the sample can be regarded as the representative of a relatively young (average age of 39.35 years old) and edu- cated sector of the Iranian society (more than 78% of the population had at least an undergraduate education). The selected schools are located in Iran. Iran was chosen as a context, where the access to and the use of ICT encounter several challenges. There were three reasons that guide the choice of primary schools for data collection. First, the schools include a mixture of individuals (parents) with dif- ferent demographics such as age, education, and income for participants. Second, overcoming digital divide for par- ents of these students is in particular important, because the next working generation is brought up by them, and it is important for these parents to be active participants of the Information Society. Recent studies in this area have tar- geted the context of schools (We i et al.2011). Therefore, the context of data collection (primary schools) can be regarded as an important target for research on digital divide. Third, we had access to these primary schools and this choice facilitated data collection and member checking processes.

Respondents were asked about their experience in work- ing with ICT and the barriers they have faced regarding the access to and use of the ICT. In this research, we framed our questions to focus on the Internet, rather than ICT in general.

The focus groups were guided by the digital divide litera- ture and the model of access gaps, yet they were designed to evoke new concepts and linkages. The researcher asked open-ended questions such as: ‘What are the problems that you have faced regarding access to Internet?’, ‘What are the problems that you have faced in using Internet?’, and ‘Have you experienced any barrier that inhibits the eﬀective use of Internet?’ Based on the responses to these questions, probing questions were asked to elicit further information.

3.2. Construction of the revealed causal map Four steps were followed to derive the revealed causal maps from the transcribed interviews. This procedure (Nelson et al. 2000a) is detailed below:

Identifying causal statements and linkages in transcripts:

The transcriptions were systematically examined to iden- tify causal statements. Causal statements are statements that imply an explicit cause–eﬀect relationship. These state- ments can be identiﬁed by using speciﬁc keywords such as ‘so’, ‘if–then’, ‘because’, and so forth. Consistent with Nelson et al. (2000a), all the causal statements and linkages were recorded in the actual language of the interviewees.

Using the generated coding scheme, the identiﬁed causal statements were broken into ‘causes’ and ‘eﬀects’. For example, in a statement such as ‘It is challenging for my family to buy access to high-speed Internet. I think it is expensive for any average family. So it makes sense that I am not motivated to invest in the Internet,’ the phrase ‘I think it is expensive for any average family.’ was treated as the cause statement, and ‘I am not motivated to invest in the Internet’ was considered as the eﬀect statement.

Development of coding scheme : We then developed a coding scheme by grouping frequently mentioned words together and developing concept labels. Coding scheme summarises the meaning of a phrase in a word or word group (concept label). Based on the concepts that emerged from the phrases captured in the language of the partici- pants, one of the authors developed concept labels. As an example in the phrase. ‘I think it is expensive for any average family. So it makes sense that I am not motivated to invest in the Inter- net,’ the cause statement was coded as ‘High-Level Costs of Getting Access to ICT’ and the eﬀect statement was coded as ‘Motivation to Get Access to ICT.’ In another example, cause statement was coded as ‘No Work-Related Require- ments to Use ICT’ and eﬀect is ‘Operating Skills’: ‘I am a 334S. Ghobadi and Z. Ghobadi teacher in guidance school with very little IT knowledge.

My work doesn’t require me to do anything with computers.

So I don’t have that much exposure to the Internet to learn how to work with computer and the Internet.’ The appendix states some more sample quotes. The other author read the material to verify their face validity and assess the parsimony and coverage of the cod- ing scheme. Scott’s (1955) pi was calculated in order to estimate the reliability of the coding process. A heuristic for content analysis is to require a reliability coeﬃcient of approximately.75 or more when using pi or alpha (Holsti 1969). For this study, Scott’s (1955)pi was .93, indicat- ing an acceptable level of reliability. Where disagreement occurred, the discrepancies were resolved through discus- sion; thus the potential authors’ biases were controlled and managed.

Constructing raw revealed causal maps: Next, we replaced the causal statements with the appropriate concept labels as developed in the coding scheme. This resulted in developing eight revealed causal maps for each focus group session.

Aggregating revealed causal maps : The eight revealed causal maps were aggregated by adding together the con- cepts and linkages of each causal map. The union of all concepts and linkages from the individual maps were placed on the ﬁnal aggregate model. As the concepts emerged from the participants, the point of redundancy represented the point at which further data collection would not pro- vide additional concepts (Armstrong et al. 2007). The point of redundancy was computed by aggregating the concepts mentioned by each participant. No new concepts were elicited from the sixth focus group, so redundancy was reached by the sixth session. This suggested that the sample of eight focus groups was suﬃcient to capture the relevant concepts in the sample.

Analysis and validation of revealed causal map: Accord- ing to Armstrong et al. (2007), the two aspects of ‘content’ and ‘structure’ should be considered in analysing revealed causal maps. The content analysis consists of identifying and deﬁning the concepts contained in the domain under study. We used member checks to validate the concepts and linkages in the aggregate revealed causal maps. Simi- lar to internal validity in conﬁrmatory studies, the objective of comprehensive members’ checks is to test for interpre- tive accuracy and to check credibility and trustworthiness of the results (Guba and Lincoln 1985). This was done by going back to the original participants and asking their opin- ion about the concepts, constructs, and linkages that we represented on the maps. The structural analysis involves analysing the linkages between the concepts. It focuses on key measures from social network analysis: (1) reachability measure, (2) cen- trality measure, and (3) density measure. Reachabilityis an indicator of the total strength of the connection between constructs (Knoke and Kuklinski 1982); it is calculated as the sum of the direct and indirect eﬀects of one construct on another. For this research, we used a .2 reachability cut-oﬀ because it allowed for a maximum inclusiveness of con- cepts within the constraint of map readability. For example, the reachability of the linkage between ‘high-level costs of getting access to ICT’ and ‘motivation to get access to ICT’ is .05. ‘high-level costs of getting access to ICT’ has only direct eﬀect on ‘motivation to get access to ICT’. This link- age has be stated 11 times out of the total 202 stated linkages, and the reachability of this linkage is equal to.05 (11/202).

Centrality is an indicator of how central or important a con- struct is to the map (Nadkarni and Narayanan 2005); it is calculated by dividing the number of direct linkages involv- ing the construct to the total number of linkages in the map.

For example, centrality of ‘high-level costs of getting access to ICT’ is .04 and it is calculated as 1 (direct linkages to this construct)/28 (total linkages in the model). Constructs with higher levels of centrality are directly involved in more linkages of the map. Densityis an indicator of the inter- connectedness of the constructs in the map (Nadkarni and Narayanan 2005); it is calculated by dividing the number of links among constructs to the number of constructs in the map.

4. Results The research question asks: ‘How diﬀerent types of access gaps interact and shape digital divide?’ We address this question by analysing dominant themes that emerged from our empirical investigation. As a result, we present an empirically based theoretical model as in Figure 2. This is followed by a discussion of the ﬁndings with regard to the research question and the theoretical and practical contributions of the study.

4.1. Theoretical model We analysed the empirical data from eight focus groups sessions to investigate salient concepts that are cognitively associated with the barriers to access to and use of ICT.

This resulted in developing a theoretical model (Figure 2), which reveals 22 concepts and the linkages between them.

In the next section, we elaborate the theoretical insights.

4.2. Access types ‘Motivation to get access to ICT’, ‘operational skills’, ‘information and strategic skills’, and ‘active use of ICT’ concur with the literature on access gaps and digital divide (as discussed in the theoretical background section).

Besides, two new concepts emerged: ‘on and oﬀ access to ICT’ and ‘anti-ﬁltering skills’. Both of these concepts emerged due to the politics of the Internet in the examined context. This ﬁnding supports the extant literature, which refers to the censorship of the Internet as a socio-political tool (Ameripour et al. 2010). Respondents asserted that Behaviour & Information Technology335 Figure 2. Theoretical model (access gaps and digital divide).

challenges in getting material access are not only about getting an initial access to the ICT; but they also include ‘on and oﬀ access’ that occurs as the result of censorship policies during the year (e.g. ‘close to elections there are more control on the use of the Internet. So I can’t browse some of the websites that were ﬁne the week before’). On this background, respondents referred to another type of skills access, anti-ﬁltering skills, for dealing with censor- ship policies (e.g. ‘I learnt from a colleague that by installing software I can bypass the ﬁltering, but it doesn’t work all thetime and it needs updates’). In summary, Figure 2extends the concept of ‘material access’ and ‘skills access’.

4.3. Concepts and linkages The theoretical model has a density of 1.27, which indicates that participants saw strong connections between all of the concepts. The centrality of concepts is shown in their cir- cle. Upon visual investigation of 22 concepts, we labelled a category name for each of them. As shown in Table 1, the Table 1. Centrality of concepts.

Categories ConceptCentrality Motivational-related concepts Lack of interest in IT-related things .29 Lack of motivation to learn recent technology .18 Motivation to get access to ICT .18 Access-related concepts Access to ICT .14 Skills-related concepts Operating skills .14 Information and strategic SKILLS .14 Usage-related concepts Active use of ICT .14 Skills-related concepts Not having IT background .11 Anti-ﬁltering skills .11 Personal-related concepts Technology phobia .07 No-time for using ICT .07 Contextual-related concepts Regulations (governmental rules) .07 Access-related concepts On and oﬀ access to ICT .07 Personal-related concepts High-level costs of getting access to ICT .04 Gender-related expectations .04 Not being an IT person (personality) .04 No-urgency to use ICT .04 No good experience of using ICT .04 Work-related requirements No work-related requirements to use ICT .04 Routine work .04 Skills-related concepts Lack of critical thinking skills .04 Access-related concepts Low-speed Internet .04 336S. Ghobadi and Z. Ghobadi Table 2. Reachability matrix.

following concepts had the highest centrality to the map:

‘lack of interest in IT-related things’ (.29), ‘lack of moti- vation to learn recent technology’ (.18), and ‘motivation to get access to ICT’ (.18).

Reachability(a measure of the strength of relationships between two concepts) is represented by the number next to the arrows in Figure 2. For ease of understanding, we provide reachability information in Table 2. For example, reachability for the relationship between ‘lack of motiva- tion to learn recent technology’ and ‘operating skills’ is .11 and, this represents the sum of all the paths linking the two concepts. We can see that the strongest relationship (reach- ability at.19) is between ‘access to ICT’ and ‘active use of ICT’. Overlaying the information on reachability and cen- trality, we note eight concepts that play a key role in participants’ mental model. We categorise these eight key concepts as below:

(1) Motivational-related concepts: ‘lack of interest in IT-related things’, ‘lack of motivation to learn recent technology’, and ‘motivation to get access to ICT’.

(2) Access-related concepts: ‘access to ICT’.

(3) Skills-related concepts: ‘operating skills’, ‘anti- ﬁltering skills’, and ‘not having IT background’.

(4) Usage-related concepts: ‘active use of ICT’.

These ﬁndings indicate the importance of overcoming the four access gaps by increasing ‘motivation to get access to ICT’, facilitating ‘access to ICT’, improving ‘operating and anti-ﬁltering Skills’, and promoting ‘active use of ICT’. In addition, the ﬁndings place an emphasis on increasing indi- viduals’ interest in IT-related things as well as educating them to have some sort of IT-related knowledge. To achieve these, it is important to understand several factors that inﬂu- ence each of these concepts (as shown in Figure 2). For example, one of the contributing factors to ‘lack of interest in IT-related things’ is ‘no good experience of using ICT’.

Positive exposure to the Internet and creating attractive experiences can be positively attributed to increasing indi- viduals’ interest in using the Internet, as this will increase their interest improving their operating skills (as shown in Figure 2).

5. Discussion 5.1. Theoretical implications The extant literature on digital divide is limited in oﬀering theoretical frameworks that explicate the dynamics behind shaping inequalities to the access and use of the ICT (We i et al. 2011). In fact, most of the research in this area has primarily focused on the impact of demographics such as gender, race, education, income, and age on generating digital divide across individuals, organisations, and nations. A systematic investigation of the interrelation between the underlying concepts behind shaping digital divide has yet to receive signiﬁcant attention (Hoﬀman et al.

2000, Rice and Katz 2003, Gui and Argentin 2011). This study builds upon the previous literature (in particular, the Behaviour & Information Technology337 theoretical lens of access gaps, Van Dijk 2005)inorder to provide a ﬁner theoretical explanation for the dynam- ics shaping digital divide. Using an exploratory qualitative research method (RCM) enabled us to go beyond the model of access gaps and explore how four access gaps interact.

All together, our results shed light on the linkages between 22 concepts (out of which 7 concepts are in the clusters of access gaps, and the rest are related concepts). Overall, this study oﬀers three valuable contributions to the existing theory on digital divide: First, the model of access gaps by Van Dijk (2005)is extended into a more comprehensive and interactive the- oretical model that shows detailed mechanisms through which motivational, material, skills, and usage access gaps interact. In particular, building on the work of Van Dijk (2005), this study identiﬁed 22 inter-related concepts, which is a much broader conceptualisation of access gaps and dig- ital divide than has been previously suggested. While some of our suggested concepts have been discussed in other arti- cles (Rice and Katz 2003, Van Deursen and Van Dijk 2011), some of the concepts such as anti-ﬁltering skills, as well as some of the linkages have not. The ﬁndings conﬁrm that technology-related issues such as material access concepts are not neutral artefacts in society ( Eastin and Larose 2000, Barzilai-Nahon 2006), rather four types of access divides interact and shape digital divide. Beyond the context of this study (parents of students in the context of Iran), the theo- retical model can be applied to individuals in other contexts and countries. We believe testing this model in a wide vari- ety of contexts would help to establish the boundaries of its applicability. Second, the undertaken research methodology (RCM) provides a fertile theoretical grounding for future research on the dynamics of digital divide. This corresponds to the calls for theoretical-qualitative research studies in this area (Mason and Hacker 2003, Van Dijk 2006). The aggregate theoretical model (Figure 2) is rich in detail and provides future research possibilities. Third, this study yields some interesting results explain- ing how several individual, social, and national factors (e.g.

‘technology phobia’, ‘high-level costs of getting access to ICT’, and ‘regulations’) may intersect and, in turn gener- ate digital divide. The ﬁndings place a special emphasis on the centrality of motivational-related concepts such as ‘lack of interest in IT-related things’ and ‘lack of moti- vation to learn recent technology’, and to a lesser extent on ‘access to ICT’. As shown in Figure 2, the motivational access is shaped as the result of complex interactions among various concepts; this refers to a more complex realisation compared with the ﬁndings of prior research (Mescha and Talmudb 2011). While policies such as exposure to com- puters at the workplace or even at home can be eﬀective in increasing ICT-related skills and decreasing technology phobia (Vandenbroeck et al.2008), they are not suﬃcient in dealing with the complex phenomenon of digital divide (Kvasny and Keil 2005). ‘Regulations (government rules)’, ‘low-speed Internet’, and ‘high-level costs of getting access to ICT’ were shown to play a role. Surprisingly, the aggre- gate model of Figure 2shows that these concepts contribute to digital divide through inﬂuencing motivational issues.

For example, ‘regulations (government rules)’ indirectly contributes to ‘motivation to get access to ICT’ through ‘on and of access to ICT’ (as well as ‘low-speed Internet’) and ‘lack of interest in IT-related things’. These ﬁndings empha- sise the importance of promoting individuals’ motivation to use ICT as well as to lean recent technologies. Lack of motivation foments the development of negative attitudes to technology, and eventually deters individuals from adopting the Internet and developing ICT-related skills. Motivation drives individuals to get access to ICT and enhance their skills to deal with challenges such as regulations.

5.2. Practical implications The theoretical model (Figure 2) has insights for policy- makers in understanding what shapes digital divide, how access gaps are created, and how access gaps contribute to inequalities in the access to and the use of ICT. Having awareness of these concepts is important in (1) understand- ing digital divide at diﬀerent levels and (2) implementing the most appropriate methods for dealing with the existing challenges in this regard. The results underscore the importance of motivational- related (‘lack of motivation to learn recent technology’), access-related (‘access to ICT’), skill-related (‘operating skills’), usage-related (‘active use of ICT’), and context- related factors (e.g. ‘high-level costs of getting access to ICT’) for alleviating digital divide. New technologies fre- quently emerge; therefore, providing proper access to ICT and promoting the need for the use of ICT are added signif- icance. At the level of parents of primary-school students, digital divide may have various social-level eﬀects. For example, unequal patterns of access to and use of the Inter- net among parents are likely to exacerbate digital divide among the next-generation individuals (students) as well.

Using our results, government policy-makers can work with education service providers to design and implement pro- grammes that aim at limiting digital divide among adults and students. Strategies may include: IT training classes for parents (to tackle lack of IT background), promoting visibility of the use of ICT through incorporating ICT in administrative and educational activities (to tackle no-time for using ICT and not having a good experience from using ICT), school computing access, and reducing the burden for low-income households to purchase home computers (to tack cost-related challenges).

6. Limitations This study is one of the early empirical attempts to advance a theoretical account of the underlying mechanisms that con- tribute to digital divide. Therefore, it oﬀers a theoretical 338S. Ghobadi and Z. Ghobadi platform upon which subsequent studies on the digital divide can be built. However, as with any piece of research, this study has limitations. First, caution must be exercised when attempting to generalise our results to individu- als in other countries with diﬀerent institutional, cultural, and political environments. Another limitation arises with regard to the characteristics of the participants. Due to their relatively middle-age (on average ∼39 years) and their high levels of education, the data do not reﬂect the concerns of all individuals with diﬀerent ages and education character- istics. Finally, the model of this study does not distinguish the concerns of diﬀerent genders; rather we have aggregated all the cognitive insights in one single model. In the follow- ing section, we propose a number of directions for future research that can address the mentioned limitations.

7. Future directions We welcome future research that examines and extends our ﬁndings and proposes appropriate interventions to allevi- ate digital divide in diﬀerent cultures (e.g. Western versus Eastern) and even similar contexts (e.g. China and Iran). Examining younger population (e.g. high school or uni- versity students) can shed additional lights into the results of this study and produce further insights for studying digi- tal divide. We also acknowledge that genders have diﬀerent concerns regarding digital divide and factors that contribute to it (Cooper and Weaver 2003, Cooper 2006). In addition, it has been shown that diﬀerent races and sexes might have diﬀerent operating, information, or strategic skills (Gui and Argentin 2011). Given space constraints, in this research we do not choose to discuss how women and men diﬀer in experiencing digital divide and the reasons behind that. In future, we would like to investigate this issue and examine its various implications for researchers and policy-makers.

References Agarwal, R., Animesh, A., and Prasad, K., 2009. Research note – social interactions and the ‘digital divide’: explaining varia- tions in Internet use. Information Systems Research, 20 (2), 277–294.

Ameripour, A., Nicholson, B., and Newman, M., 2010. Convivial- ity of Internet social networks: an exploratory study of Inter- net campaigns in Iran. Journal of Information Technology, 25 (2), 244–257.

Armstrong, D.J., et al., 2007. Advancement, voluntary turnover and women in it: a cognitive study of work–family conﬂict.

Information & Management, 44 (2), 142–153.

Barth, M. and Veit, D., 2011. How digital divide aﬀects public e-services: the role of migration backgrounded. Wirtschaft- informatik proceedings, Zurich, Switzerland. Paper 118.

Association for Information Systems, 1–14.

Barzilai-Nahon, K., 2006. Gaps and bits: conceptualizing mea- surements for digital divide/s. The Information Society,22 (5), 269–278.

Bonfadelli, H., 2002. The Internet and knowledge gaps a the- oretical and empirical investigation. European Journal of Communication, 17 (1), 65–84. Cammaerts, B., et al., 2003.Beyond the digital divide. Brussels:

VUB Brussels University Press.

Chen, W. and Wellman, B., 2004. The global digital divide – within and between countries. IT & Society, 1 (7), 39–45.

Cho, J., De Zuniga, H.G., Rojas, H., and Shah, D.V., 2003. Beyond access: the digital divide and Internet uses and gratiﬁcations.

IT & Society, 1 (4), 46–72.

Cooper, J., 2006. The digital divide: the special case of gender. Journal of Computer Assisted Learning, 22 (5), 320–334.

Cooper, J. and Weaver, K.D., 2003. Gender and computers:

understanding the digital divide. Mahwah, NJ: Lawrence Erlbaum.

Cross, K., 2010. Why Iran’s green movement faltered: the limits of information technology in a rentier state. SAIS Review,30 (2), 169–187.

Dewan, S. and Riggins, F.J., 2005. The digital divide: current and future research directions. Journal of the Association for Information Systems, 6 (12), 298–337.

DiMaggio, P. and Hargittai, E., 2001. From the ‘digital divide’ to ‘digital inequality’: studying Internet use as penetration increases. Princeton: Center for Arts and Cultural Policy Studies, Woodrow Wilson School, Princeton University.

Eastin, M.S. and Larose, R., 2000. Internet self-eﬃcacy and the psychology of the digital divide. Journal of Computer- Mediated Communication, 6 (1). Doi: 10.1111/j.1083-6101.

2000.tb00110.x Giroux, H., 2009. The Iranian uprisings and the challenge of the new media: rethinking the politics of representation [online].

Available from: http://www.uta.edu/huma/agger/fastcapitali sm/5_2/Giroux5_2.html [Accessed 12 July 2011].

Guba, E.G. and Lincoln, Y.S., 1985. Naturalistic inquiry. Newbury Park, CA: Sage.

Gui, M. and Argentin, G., 2011. Digital skills of Internet natives: diﬀerent forms of digital literacy in a random sample of north- ern italian high school students. New Media & Society, 13 (6), 963–980.

Gunkel, D.J., 2003. Second thoughts: toward a critique of the digital divide. New Media & Society, 5 (4), 499–522.

Gurstein, M., 2003. Eﬀective use: a community informatics strat- egy beyond the digital divide. First Monday[online], 8 (12), 1. Available from: http://ﬁrstmonday.org/ojs/index.php/fm/ article/viewArticle/1107/1027&quot%3B&gt%3B [Accessed 29 October 2012].

Hargittai, E., 2002. Second-level digital divide: diﬀerences in peo- ple’s online skills. First Monday, 7 (4). Doi: http://dx.doi.org/ 10.5210%2Ffm.v7i4.942.

Hargittai, E., 2010. Digital na (t) ives? Variation in internet skills and uses among members of the ‘net generation’.

Sociological Inquiry, 80 (1), 92–113.

Hargittai, E. and Hinnant, A., 2008. Digital inequality diﬀer- ences in young adults’ use of the internet. Communication Research, 35 (5), 602–621.

Hoﬀman, D.L., Novak, T.P., and Schlosser, A., 2000. The evo- lution of the digital divide: how gaps in Internet access may impact electronic commerce. Journal of Computer- Mediated Communication, 5 (3). Doi: 10.1111/j.1083- 6101.2000.tb00341.x.

Hohlfeld, T.N., Ritzhaupt, A.D., Barron, A.E., and Kemker, K., 2008. Examining the digital divide in K-12 public schools:

four-year trends for supporting ICT literacy in Florida.

Computers & Education, 51 (4), 1648–1663.

Holsti, O.R., 1969. Content analysis for the social sciences and humanities. Don Mills, ON: Addison-Wesley.

Kalichman, S., Weinhardt, L., Benotsch, E., and Cherry, C., 2002. Closing the digital divide in HIV/AIDS care: development of Behaviour & Information Technology339 a theory-based intervention to increase Internet access.AIDS Care, 14 (4), 523–537.

Katz, J.E. and Rice, R.E., 2002. Social consequences of Internet use: access, involvement, and interaction. Cambridge: The MIT Press.

Katz, J.E., Rice, R.E., and Aspden, P., 2001. The Internet, 1995–2000 access, civic involvement, and social interaction.

American Behavioral Scientist, 45 (3), 405–419.

Knoke, D. and Kuklinski, J.H., 1982. Network analysis. Beverly Hills, CA: Sage.

Kvasny, L. and Keil, M., 2005. The challenges of redressing the digital divide: a tale of two us cities. Information Systems Journal, 16 (1), 23–53.

Lee, A.S. and Baskerville, R.L., 2003. Generalizing generaliz- ability in information systems research. Information Systems Research, 14 (3), 221–243.

Mason, S.M. and Hacker, K.L., 2003. Applying communication theory to digital divide research. IT & Society, 1 (5), 40–55.

Mescha, G.S. and Talmudb, I., 2011. Ethnic diﬀerences in inter- net acces. Information, Communication & Society, 14 (4), 445–471.

Nadkarni, S. and Narayanan, V., 2005. Validity of the struc- tural properties of text-based causal maps: an empirical assessment. Organizational Research Methods , 8 (1), 9–40.

Narayanan, V.K. and Fahey, L., 1990. Evolution of revealed causal maps during decline: a case study of admiral. In: A.S. Huﬀ, ed. Mapping strategic thought. Chichester, UK: John Wiley & Sons, 109–134.

Nelson, K.M., Nadkarni, S., Narayanan, V., and Ghods, M., 2000a. Understanding software operations support expertise:

a revealed causal mapping approach. MIS Quarterly, 24 (3), 475–507.

Nelson, K.M., Nelson, H.J., and Armstrong, D., 2000b. Revealed causal mapping as an evocative method for information sys- tems researched. Hawaii international conference on system sciences. Hawaii: IEEE.

Norris, P., 2003. Digital divide: civic engagement, information poverty, and the Internet worldwide . New York: Taylor & Francis.

OECD, 2001. Bridging the digital divide: issues and policies in OECD countries. Paris: OECD.

Rains, S.A., 2008. Health at high speed broadband Internet access, health communication, and the digital divide. Communica- tion Research, 35 (3), 283–297.

Rice, R.E. and Katz, J.E., 2003. Comparing internet and mobile phone usage: digital divides of usage, adoption, and dropouts.

Telecommunications Policy, 27 (8), 597–623. Scott, W.A., 1955. Reliability of content analysis: the case of nominal scale coding. Public Opinion Quarterly , 19 (1), 321–325.

Selwyn, N., 2004. Reconsidering political and popular under- standings of the digital divide. New Media & Society, 6 (3), 341–362.

Valadez, J.R. and Durán, R.P., 2007. Redeﬁning the digital divide: beyond access to computers and the internet. The High School Journal, 90 (3), 31–44.

Van Deursen, A.J.A.M. and Van Dijk, J.A.G.M., 2009. Improv- ing digital skills for the use of online public information and services. Government Information Quarterly , 26 (2), 333–340.

Van Deursen, A.J.A.M. and Van Dijk, J.A.G.M., 2011. Internet skills and the digital divide. New Media & Society, 13 (6), 893–911.

Van Dijk, J.A.G.M., 1999. The network society, social aspects of new media. Thousand Oaks, CA: Sage.

Van Dijk, J.A.G.M., 2005. The deepening divide: inequality in the information society. Thousand Oaks, CA: Sage.

Van Dijk, J.A.G.M., 2006. Digital divide research, achievements and shortcomings. Poetics, 34 (4), 221–235.

Van Dijk, J.A.G.M., 2012. The evolution of the digital divide. In:J.

Bus et al. eds.Digital enlightenment yearbook. Amsterdam, Netherlands: IOS Press, 57–75.

Van Dijk, J. and Hacker, K., 2003. The digital divide as a complex and dynamic phenomenon. The Information Society, 19 (4), 315–326.

Vandenbroeck, M., Verschelden, G., and Boonaert, T., 2008. E-learning in a low-status female profession: the role of motivation, anxiety and social support in the learning divide. Journal of Computer Assisted Learning , 24 (3), 181–190.

Vehovar, V., Sicherl, P., Hüsing, T., and Dolnicar, V., 2006. Methodological challenges of digital divide measurements.

The Information Society, 22 (5), 279–290.

Vicente, M.R. and Lopez, A.J., 2010. A multidimensional analysis of the disability digital divide: some evidence for Internet use.

The Information Society, 26 (1), 48–64.

Warschauer, M., 2002. Reconceptualizing the digital divide. First Monday [online], 7 (7). Available from: http://www.ﬁrstmon day.dk/issues/issue7_warschauer/ Wei, K.K., Teo, H.H., Chan, H.C., and Tan, B.C.Y., 2011. Concep- tualizing and testing a social cognitive model of the digital divide. Information Systems Research, 22 (1), 170–187.

Yin, R.K., 2009. Case study research: design and methods.

Thousand Oaks, CA: Sage. 340S. Ghobadi and Z. Ghobadi Appendix. Concept level classiﬁcation scheme.

Concept DeﬁnitionExample High-level costs of getting access to ICT High and challenging costs associated with getting access to ICT (e.g. buying computer, buying Internet accounts) Being an average-family, It is not easy for us to buy high-speed Internet for home. It is just expensive and we can’t aﬀord it easily Technology phobia Fear and anxiety in working with technology I don’t feel comfortable when working with computers; I don’t feel conﬁdent about it Not having IT background Not having any IT-related university background Well, I teach sociology at high school. I haven’t been trained at university to know how hardware and software work Lack of motivation to learn recent technology When individuals who are not motivated in learning new technologies To be honest, I am not interested to learn new technologies such as IPhones. I prefer to stay with traditional ways of communication No-time for using ICT When individuals do not have time to work with computers and the Internet I have two children, both going to school. I also work. After work, I have to engage with my family, and so there is not much time for Internet or such things Gender-related expectations When there are gender-related expectations that impede the use of ICT I am the mother of an 8-year old child, besides I work. When I am back home, I need to cook for family and play with my child Not being an IT person (personality) When a person does not see herself/himself into IT Most of the things that I enjoy are not related to computers. I prefer to do painting, than sitting in front of the computers!

No-urgency to use ICT When using ICT does not contribute to the important things of someone’s life If I don’t check my email two weeks, nothing happens No good experience of using ICT When a person has not yet experienced something interesting and useful with the use of ICT I don’t remember that using IT has made me say ‘wow’ Lack of interest in IT-related things When a person is not interested in IT-related things I am not quite interested in IT. That is all Motivation to get access to ICT The motivation to buy a computer and an Internet account to be connected I like to have a good PC at home, to have a high-speed Internet at home Regulations (governmental rules) Governmental rules that control and limit the use of the Internet There are always new governmental control strategies and limitations on the use of the Internet On and oﬀ access to ICT When a person is sometimes connected to the Internet/or has PC, and sometimes is not connected Sometimes, my connection is fast and nice. But not always... Some websites can be shut down for a period or for always; sometimes the Internet is so slow that it is simply not connected Access to ICT Having access to computer and the Internet We have computer and a reasonably ﬁne access to the Internet No work-related requirements to use ICT When the work environment does not require working with computer and the Internet I know many places provide computers for their employees. But I don’t really need a computer for my work Routine work When the job responsibilities are rather routine, than requiring creative ways for accomplishing tasks Checking sales and writing a weekly report do not need much thinking or searching for information from other channels Lack of critical thinking skills When someone perceive that s/he does not have critical thinking skills You know our education systems appreciate being agreeable and simply memorising information Operating skills Skills for working with computers and the Internet It is a basic requirement to know how to work with computer and the Internet Information and strategic skills Skills to search, select, process information in computer, and use ICT for competitive advantage I know my brother is just great in looking for what he wants in the Internet. But I am not as good as him Anti-ﬁltering skills Skills to bypass ﬁltering If I want to check my Facebook, I need to know updated anti-ﬁltering software very well Active use of ICT Making use ICT Once I updated some information in the Wikipedia Low-speed internet When the speed of the Internet is low and frustrating Do you think how much It takes for me to download photos? It is so frustrating. Copyright ofBehaviour &Information Technologyisthe property ofTaylor &Francis Ltd and itscontent maynotbecopied oremailed tomultiple sitesorposted toalistserv without the copyright holder'sexpresswrittenpermission. However,usersmayprint, download, or email articles forindividual use.