Attitudes and the Digital Divide: Attitude Measurement as Instrument to Predict Internet Usage

Donat, Elisabeth, Brandtweiner, Roman, Kerschbaum, Johann, Informing Science: the International Journal of an Emerging Transdiscipline

Introduction

"The last half of the 1900s has been characterized by the increasing importance of information and communication technologies (ICTs) in social and organizational life" (Sawyer & Rosenbaum, 2000). In the now established information society, it is crucial that people have access to the new media and know how to use the new ICTs. Without access to the internet and without the necessary skills that regularly go along with the attitude to use the new information technologies, people can neither inform themselves electronically nor can be informed by organisations and institutions using IT. This is why issues concerning the digital divide are of striking importance in this context, as pointed out by Elizabeth C. Boyd, "As information technology is fast becoming a major tool for disseminating and obtaining information, gaps between those who have access to this tool and those who do not is a major concern"(2002, p. 113).

Research on digital divide focuses first and foremost on the speed of various groups in adopting these new ICTs (Anderson, Bikson, Law, & Mitchell, 1995; Gehrke, 2004; Katz & Rice 2002; Katz, Rice, & Aspden 2001; Norris, 2001; Selhofer & Husing, 2002). The relation of demographics and internet usage is one of the major research questions in the field of digital divide. Phenomena like these can be subsummized under the so-called "first order digital divide" (Dewan & Riggins, 2005). The so-called "second order digital divide" addresses questions about different abilities to use the internet and draw advantages from its usage (DiMaggio, Hargittai, Neuman, & Robinson, 2001; Hargittai, 2002, 2003). Less research has been done in a third dimension of adopting new technologies, i.e. the field of attitudes towards the internet, although these attitudes play an important role in adoption and in learning how to use this new medium. Besides analyzing demographic characteristics of users and non-users we want to focus on the question of attitudes, especially of non-users in comparison to users, to acquire a deeper understanding of the process of "digital divide" and the adoption of new technologies.

Former studies investigating the digital divide used theories of diffusion and innovation as explanatory frameworks (Carveth & Kretchmer, 2002). Doubtlessly research like this provides useful and interesting information about the digital divide phenomenon. The explanatory strength of studies like this lies in explaining the causes and consequence of the first order digital divide. If we want to understand the second order digital divide, i.e. why people who theoretically have access to ICTs and Internet do not use these new media, we have to take a closer look at attitudes. This was already emphasized seven years ago by Burkett, Compton, and Burkett (2001). In another publication, they pointed out that "... the impact of computer attitudes on computer knowledge is still a key component to the understanding of information science" (Compton, Burkett, & Burkett, 2002, p. 219).

What do we know about attitudes towards the internet? To answer this question, unfortunately one has to refer to one of the several studies made that observe attitudes in populations of students (Jackson, Ervin, Gardner, & Schmitt, 2001; Li, Kirkup & Hodgson, 2001; Sam, Othman & Nordin, 2005) or even to one of the online-surveys addressing only current users. In our literature review we also found several studies about attitudes towards e-commerce - of course also focusing on a very specific section of the population. These target groups are specific and biased in a dimension of vital importance when discussing phenomena of digital divide: they certainly have a relatively high level of computer and internet literacy compared to the general population of a society. Therefore it is inappropriate to draw conclusions about internet attitudes and behaviour from such selective populations and to generalize from them. Studies using representative population samples are still scarce.

Research on attitudes has a very long tradition in social sciences, and techniques are becoming more sophisticated (Benninghaus, 1976, Krebs & Schmidt, 1993). Much effort has been made to refine questionnaires, scales and data analysis. Besides simply getting to know attitude patterns and components of a population, there is a rather implicit goal that lies behind these efforts: the wish to explain behaviour by attitudes has always been one of the major desiderata in empirical research. Solving this problem also addresses an issue of justification for social sciences: if behaviour is influenced by attitudes, various ways to influence behaviour are open to practitioners and politicians.

Getting to know general attitude patterns towards the internet via a representative sample can help to obtain knowledge beyond demographic characteristics about factors which might hinder people from getting online. The measurement of attitudes is a considerably more difficult problem than measuring the influence of demographics on people's internet usage behaviour. Whilst dealing with attitudes, we have to find elaborated indicators to measure the latent phenomenon "attitude". Demographics, like age, gender, and income, are more manifest in their nature compared to attitudes. We also have to rely on respondents' answers on stimuli (i.e. questions) when asking for demographics, but this may yield more reliable answers (besides effects of social desirability). The verbal formulation of attitudes demands a great deal of introspection and abstract thinking on the part of the respondents. Attitudes are of course also influenced by social desirability, which we tried to reduce by using a telephone survey and design.

Theoretical Considerations

Talking about attitudes presupposes that there are attitudes to talk about. Although the question of non-attitudes is a very unpleasant issue for social scientists, our analysis would be incomplete and just a piece of empiricism (Mende, 2005) without addressing this question before discussing the basics of attitude theory. Social sciences have seen attitude theories employing mechanisms of rational thinking (Ajzen, 1993) and of a more spontaneous response towards attitude objects (Fazio, 1990). But what if there is actually no awareness of the phenomenon "internet"? We therefore asked respondents whether they actually knew what the internet is, and followed Katz and Rice (2002) in their considerations on issues of awareness. In leaving this "door" open to our respondents, we acknowledge the possibility of having no differentiated opinion on this issue.

There is not enough room to open up the whole discussion on attitude theory and measurement, so we will address only some basic considerations and open questions on these issues. A commonly agreed upon definition was proposed by Ajzen (1993), "An attitude is an individual's disposition to react with a certain degree of favorableness or unfavorableness to an object, behavior, person, institution, or event - or to any discriminable aspect of the individual's world." This definition points to a key element of attitudes: an evaluative dimension. Using scales to evaluate attitudes is a common way to obtain information about respondents' evaluations. Furthermore, attitudes are multidimensional in the sense that they include three components: a cognitive, an emotional, and a behavioural component (Rosenberg, Hovland, McGuire, Abelson, & Brehm, 1969). The cognitive component includes perceptions and knowledge of the attitude object, typically represented via stereotypes. The emotional or affective component represents feelings towards the attitude object, and the behavioural or conative component addresses questions of reacting towards the attitude object. Through accepting this definition of attitudes, one employs a multidimensional model of attitudes which can serve as a useful heuristic to structure analysis and data measurement.

As is often the case with heuristics, in practice we find exceptions; for example, the question whether attitudes are really best represented by a tripartite model or whether it would be preferable to construct a bipolar model (consisting of a cognitive and an affective component only). This question becomes even more important as the conative component is the most difficult to measure, when thinking about unidimensional and precise operationalizations. Ajzen (1993) suggests using behavioural intentions as an indicator for the conative component. In measuring hypothetical behavioural intentions, the problem becomes even more abstract and difficult for respondents, and there is wide discussion whether we should rely on such non-committal, one moment measurements. Although there are recommendations (Ajzen, 1993) to stay as close as possible to the behaviour in question and to use a narrow time frame, we also tried to convey information on concrete, current behaviour and asked non-users for perceived barriers when it comes to internet usage.

According to commonly accepted knowledge on attitudes (Taylor, Peplau, & Sears, 1994) we expect attitudes towards internet to be cognitively complex and evaluatively simple. Especially when people do not know exactly what the internet is, they may have some basic feelings (good/bad) about it without many supporting cognitions. Although there should be some kind of coherence between the three attitude components, we also know that the relationship between them can be rather complex. Attitudes can influence behaviour, but we also infer our attitudes from our behaviour. Furthermore, two people reporting nearly the same beliefs on an attitude object can hold quite different cognitions, emotions, and conative predispositions towards it (Rosenberg et al., 1969). Nevertheless, altogether people seek some kind of consistency between cognitions, affect, and behaviour and apply mechanisms of dissonance reduction to reach this consistency.

Our paper aims to introduce some new possibilities of measuring attitudes towards the …

Questia, a part of Gale, Cengage Learning. www.questia.com
Publication information: Article title: Attitudes and the Digital Divide: Attitude Measurement as Instrument to Predict Internet Usage. Contributors: Donat, Elisabeth - Author, Brandtweiner, Roman - Author, Kerschbaum, Johann - Author. Journal title: Informing Science: the International Journal of an Emerging Transdiscipline. Volume: 12. Publication date: Annual 2009. Page number: 37+. © 2008 Informing Science Institute. COPYRIGHT 2009 Gale Group.

This material is protected by copyright and, with the exception of fair use, may not be further copied, distributed or transmitted in any form or by any means.