Internet Self-Efficacy and the Psychology of the Digital Divide

Complexity, knowledge barriers to initial Internet adoption, and comfort and satisfaction issues faced by new users may be construed as self-efficacy deficits. Self-efficacy is the belief "in one's capabilities to organize and execute the courses of action required to produce given attainments" (Bandura, 1997, p. 3). People who have little confidence in their ability to use the Internet, who are dissatisfied with their Internet skills or who are uncomfortable using the Internet may be said to have weak self-efficacy beliefs. Those with low self-efficacy should be less likely to perform related behaviors in the future (Bandura, 1982), in this case, adopt and use the Internet, than those with high degrees of self-efficacy.

Within social cognitive theory (Bandura, 1982; 1997) self-efficacy is a form of self-evaluation that influences decisions about what behaviors to undertake, the amount of effort and persistence put forth when faced with obstacles, and finally, the mastery of the behavior. Self-efficacy is not a measure of skill; rather, it reflects what individuals believe they can do with the skills they possess. For example, in discussing computer self-efficacy, Compeau and Higgins (1995) distinguished between component skills such as formatting disks and booting up the computer and behaviors individuals can accomplish with such skills, such as using software to analyze data. Thus, Internet self-efficacy focuses on what a person believes he or she can accomplish online now or in the future. It does not refer to a person's skill at performing specific Internet-related tasks, such as writing HTML, using a browser, or transferring files, for example. Instead, it assesses a person's judgment of his or her ability to apply Internet skills in a more encompassing mode, such as finding information or troubleshooting search problems.

The relationship between self-efficacy and personal computer use is perhaps intuitively obvious. Personal computers represent a complex and somewhat troublesome technology, requiring considerable skill and extensive training to operate successfully. Self-efficacy is essential to overcome the fear many novice users experience. Compeau and Higgins (1995) empirically verified the relationship between computer self-efficacy and computer use. Staples, Hulland, and Higgins (1998) found that those with high levels of self-efficacy in remote computing situations were more productive and satisfied, and better able to cope when working remotely.

The Internet requires development of a further set of skills that, to the novice user, at least, may be daunting. These include establishing and maintaining a stable Internet connection, learning how to navigate on the Internet, and searching it for relevant information. Internet self-efficacy may be distinguished from computer self-efficacy as the belief that one can successfully perform a distinct set of behaviors required to establish, maintain and utilize effectively the Internet over and above basic personal computer skills.

Social cognitive theory offers an alternative to socio-economic explanations of the Digital Divide (e.g., Hoffman, et al., 2000; NTIA, 1999); the latter are less convincing now that personal computer prices have fallen to the levels of VCRs and Internet services to the level of cable television subscriptions, expenditures that over half of US households manage. "Don't want it" rivals cost as a factor explaining non-use of the Internet in minority equipped with computers (NTIA, 1999), suggesting that users must experience the benefits of the Internet for themselves to close the Digital Divide. This realization, the formation of positive outcome expectations in social cognitive terms, occurs only if Internet use persists long enough for the benefits to become apparent. For that to happen, self-efficacy beliefs must first be established.

Early research on Internet self-efficacy focused on the performance of specific tasks such as entering World-Wide Web addresses, creating folders and bookmarks, mailing pages, using File Transfer Protocol (FTP) and telnet, constructing a hypertext index, and moving bookmarks (Nahl, 1996, 1997). Ren (1999) reported a measure of self-efficacy specific to searching for government information sources. Results were consistent with previous self-efficacy literature, with self-efficacy perceptions positively related to task performance (Nahl, 1996, 1997) and the amount of use (Ren, 1999).

The prior studies did not yield a measure of self-efficacy suitable for studying overall Internet usage, and rerpoted no information about reliability and validity. In Nahl (1997), scale items confounded distinct behaviors; a single item asked about e-mail, hypertext mark-up language (HTML) scripting, telnet, and file transfer protocol. Nahl's measure referred to specific subsidiary tasks (e.g., creating bookmarks) instead of overall attainments (e.g., obtaining useful information) and thus did not properly reflect the constructive definition of self-efficacy. Ren (1999) operationalized self-efficacy in a manner more consistent with its conceptual definition (e.g., search the Internet by yourself), but a single item measure was employed so its reliability could not be determined. Ren's measure applied to a specific behavioral domain (i.e., seeking government information) rather than overall Internet use, limiting its future application.

In an effort to further understand psychological aspects of the Digital Divide, the present study builds on past research to develop a new measure of Internet self-efficacy. It assesses reliability and analyzes the construct validity of Internet self-efficacy by comparing it to measures of other constructs thought to be positively related, negative related or unrelated on theoretical grounds (Anastasi, 1988).

Hypotheses

H1: Internet self-efficacy will be positively correlated with prior Internet experience.

Self-efficacy judgments are in turn related to outcome expectations. Outcome expectations are estimates that a behavior will produce particular outcomes (Oliver & Shapiro, 1993) but depend upon how well one thinks her or she can perform the behavior (Bandura, 1977). Oliver and Shapiro (1993) found that the stronger a person's self-efficacy beliefs, the more likely he or she was to try to achieve the desired outcome. In the present context this means that Internet self-efficacy should be positively related to the expectation of positive outcomes of Internet use, such as meeting new people on the Internet.

Compeau & Higgins (1995) found that computer self-efficacy influenced expectations about the future outcomes of computer use such as job performance and personal accomplishment. In terms of the Internet, social outcomes would derive from social encounters on-line. Personal outcomes are what we can achieve personally through using the Internet, such as being entertained or obtaining information. Internet self-efficacy should be positively related to positive outcome expectations.

H2: Internet self-efficacy will be positively related to expected positive outcomes of Internet use.

Past research on computer self-efficacy indicated a significant positive relationship between computer efficacy and computer usage (Burkhart & Brass, 1990; Compeau & Higgins, 1995; Compeau & Higgins, 1999; Oliver & Shapiro, 1993). Internet use and Internet self-efficacy should also be directly related since we are more likely to attempt and persist in behaviors that we feel capable of performing.

The amount of stress a person feels performing a task is negatively related to self-efficacy (Bandura, 1977). Individuals experienced an increase in stress when attempting to perform behaviors they didn't feel confident performing (Stumpf, Brief, & Hartman, 1989). As stress increased, efficacy beliefs decreased due to self-doubt and emotional arousal when performing the behavior (Oliver & Shapiro, 1993). Performing a task successfully increased self-efficacy and decreased stress; conversely, failure or difficulty experienced in performing a task decreased self-efficacy and increased stress (Hancock, 1990).

Stress encountered while using the Internet can be understood in terms of the number of stressors encountered while online. Having trouble getting on the Internet or having the computer freeze up are common examples. When such problems are encountered they lower expectations about successful interactions with the Internet in the future. As the number of stressors encountered online increase, perceptions of success decrease and self-efficacy along with it.

H4: Internet stress will be negatively correlated with Internet self-efficacy.

Self-efficacy is one type of self-monitoring mechanism, but there are others. Self-disparagement occurs when an individual judges his or her performances as inferior to other performances. Self-disparaging people misrepresent their performance attainments or distort their recollection of past events as negative experiences (Bandura, 1977; Bandura, 1997). Self-disparaging people are depression-prone and typically dwell on their failures as evidence of their personal deficiencies while attributing their successes to external factors. In contrast, individuals with a high sense of efficacy accept success as an indication of their ability and attribute failure to external causes. Based on this relationship:

The hypothesized relationships also fit into a causally ordered theoretical framework. Self-efficacy beliefs are continually re-formed based on experience. Internet users therefore continually modify their Internet self-efficacy beliefs based on their experiences online. Using Internet self-efficacy as an antecedent to use (Bandura, 1997), the following relationships between Internet self -efficacy, Internet use, social, informational, and entertainment outcomes, and online stressors is proposed. While increased levels of self-efficacy increase Internet use, both self-efficacy and Internet use increase perceived social, informational, and entertainment outcome expectancies. Furthermore, while increased levels of self-efficacy will decrease perceptions of Internet stress, perceptions of stress will increase feeling of self-disparagement, and thus, decrease use.

Finally, to complete the construct validity argument, Internet self-efficacy should be unrelated to theoretically distinct concepts. It is conceptually important to distinguish self-efficacy from general measures of psychological well-being since a competing hypothesis would be that self-efficacy merely reflects a generally positive outlook on life, feeling good about oneself and one's social environment. Therefore, Internet self-efficacy should not be related to such general indicators of psychological well being such as depression, loneliness, perceived social support and general life stress.

H6: Internet self-efficacy will be unrelated to depression, loneliness, perceived social support and life stress.

Methods

The participants were 171 undergraduate students from an introductory communication class at a large Midwestern university. A convenience sample was deemed appropriate for the purposes of scale construction and validation since college students are a population with wide variation in Internet experience, including both heavy Internet users and many novice users. Of those who participated in the survey, 35 percent were freshman, 22 percent were sophomores, 18 percent were juniors, and 25 percent were seniors. Of these, 60 percent were male, 40 percent were female, and the mean age was 21 years old.

Questionnaires were administered in class at two separate times to maximize participation in the survey. Respondents picked up the questionnaire on the first day of class each week and returned it the second day of class that same week. Respondents were offered extra credit for participating in the study. An alternattive form of extra credit was provided for those who chose not to participate.

Items for the Internet self-efficacy scale were suggested by Compeau and Higgins (1995), the GVU 10th survey (GVU, 1999), and Nahl (1996). Items from these scales were adapted to the conceptual definition of Internet self-efficacy by phrasing them as individuals' judgments of their ability to use the Internet to produce overall attainments, as opposed to accomplishing specific sub-tasks. An eight-item measure of Internet self-efficacy was developed. A Likert-type agree-disagree scale was used to assess the participants' confidence that they could use the Internet in each of the ways specified, where 7 corresponded to "strongly agree" and 1 to "strongly disagree." Confirmatory Factor Analysis (CFA) was conducted on the eight items to assess internal consistency and factor loadings using program PACKAGE (Hunter & Gerbing, 1982).

Substantial factor loadings and a standardized Cronbach alpha of .93 were obtained (Table 1), indicating internal consistency. Each of the scale items, factor loadings, means, and standard deviations can be found in Table 1.

First, Previous Internet Experience was measured with one item ranging from less than two months (scored 1) to over 24 months (scored 5). Borrowing from Charney & Greenberg (in press), three outcome expectancy constructs measuring social and personal (including entertainment and information) outcome expectancies were created. The five-item Social Outcome (a= .86) scale assessed the perceived likelihood of developing relationships over the Internet.¹ A four-item Personal Entertainment Outcome (a = .87) scale measured the likelihood finding entertainment on the Internet.² The six-item Personal Information Outcome (a = .83) measure assessed the likelihood of finding immediate information on the Internet.³ For each of these measures, the likelihood (rated as very likely (7) to very unlikely (1)) of an expected outcome was multiplied by the corresponding evaluation of that outcome (rated very good (+3) to very bad (-3)), following the expectancy-value formulation recommended by Ajzen and Fishbein (1980).

Internet Stress was a four-item measure (a = .61) developed for this study from previous work evaluating Internet frustrations (Charney & Greenberg, in press) and problems encountered on the Internet (GVU, 1999). Respondents were asked to rate their likelihood of experiencing each type of stressful Internet behavior (e.g., trouble getting on the Internet)⁴ on a seven-point scale that ranged from very likely (7) to very unlikely (1). Self -disparagement consisted of three Likert-type items rated from strongly agree (7) to strongly disagree (1) (a= .71).⁵ This measure assessed self-perceptions of Internet- related performance.

Internet Use was measured with four items. Two items, ranging from no use (1) to more than five hours of use (5) assessed Internet use on a typical weekend and weekday, respectively. One item scored, from 0 to 7, assessed the number of days the respondent went online during a typical week; and one item ranging from no hours (1) to over 20 hours (7) assessed time spent surfing during a typical week.

Life Stresses were measured with 49 items drawn from the Kanner, Coyne, Schafer, and Lazarus (1981) Hassles Scale (a = .93). Subjects reported on the frequency with which they had encountered the daily life stresses (e.g., car maintenance, crime) in the previous month, on a four-point scale (None, Somewhat severe, Moderately severe, Extremely severe). Depression was measured with the 20-item Center for Epidemiological Studies Depression (CES-D) scale (Radloff, 1977) (a= .91). The 20-item UCLA Loneliness Scale (Russell, Peplau, & Cutrona, 1980) was used to assess general loneliness (a= .90). Finally, sixteen (out of 40) representative items from the Interpersonal Support Evaluation List⁶ (Cohen, et al., 1985, a = .81) were used as a measure of social support.

Zero order correlations were used to test each of the hypothesized relationships. LISREL 8.3 was used to test the proposed path model (Jöreskog & Sörbom, 2000). .

Results

Finally, Hypothesis 6 was supported. Life hassles (r = -.06, p = .381), depression (r = -.12, p = .122), loneliness (r = -.06, p = .418) and perceived social support (r = .09, p = .240) were not related to Internet self-efficacy (Table 2).

The initial model which specified the development of self-efficacy through use and outcome expectations was not consistent with the data (c²(17) = 78.94, p < .001). A revised model shown in Figure 2 was found to be consistent with the data (c²(7) = 13.27, p > .05). In it, prior experience was related to Internet self-efficacy (ß = .36) which in turn was related to use (ß = .54), self disparagement (ß = -.61) as well as social (ß = .13) and informational (ß = .18) outcome expectancies. Use was related to both social (ß = .35) and informational (ß = .21) outcome expectations and self-disparagement (ß = - .15).

The predictive power of this model is indicated by the R²statistics shown in Figure 2. From this model, 13 percent of the variance in Internet self-efficacy was explained. Further, 41 percent of the variance in Internet use was explained, while 20 percent and 12 percent of the variance was explained in social and informational outcome expectancies, respectively. Thirty-seven percent of the variance in self-disparagement was explained.

Discussion

Prior Internet experience was the strongest predictor of Internet self-efficacy. Up to two years' experience may be required to achieve sufficient self-efficacy. Prior research showed that new users who had been on the Internet for two years or less encountered more stressful problems online and were also less satisfied with their Internet skills than veteran users (GVU, 1999). In the present research there was also a demarcation at the two-year point, Internet self-efficacy was much lower in the first two years than later (t= -2.37, p < .027). .

The path model (see Figure 2) supported a theoretical model constructed from Bandura, (1997). The model demonstrated that Internet use was directly affected by self-efficacy judgments. Usage and self-efficacy subsequently increased outcome expectations.

However, while self-efficacy was found to affect outcome expectations directly, the relationships were not as powerful as those found with usage. The specificity of the Internet self-efficacy measure could be at issue. The measure included only single items representing the achievement of social or informational outcomes on the Internet, such as using it to gather data or to obtain help from a discussion group. It appears that the ability to obtain these types of outcome expectancies through the Internet is something that gradually develops over time (Pew Research Center, 2000), so Internet self-efficacy specific to relationship formation and management may play a role in its attainment.

Likewise, there were no items in the self-efficacy measure referring to the use of the Internet for entertainment purposes. It also may be that the entertainment outcomes of the Internet are so easily attained that they do not require special skills, so that no mediation through self-efficacy is required. That said, the relationship between Internet self-efficacy and Internet outcome expectancies has provided researchers with new areas from which to begin validating and expanding this self-efficacy measure.

The expected relationship among self-efficacy, Internet stress, self-disparagement and Internet use was not observed. The revised model presented self-disparagement as an antecedent to Internet use. Given the low reliability obtained for Internet stress , researchers should continue to explore this relationship as presented in Figure 1.

Further research should explore self-efficacy measures specific to achieving particular types of outcomes through the use of the Internet. The importance of distinguishing general and task-specific self-efficacy has been discussed with respect to computer usage (Marakas, Yi & Johnson, 1998) and can be expected to be an important issue in Internet-related studies as well. Social cognitive theory also distinguishes coping self- efficacy, or beliefs in one's ability to deal with specific stress-inducing problems. In this case, these would be the various technical (e.g., inability to establish a connection) and socio-technical (e.g., receiving unwanted e-mail) difficulties that result from Internet use. The process of initially establishing Internet access, whether through one's own computer or a public access node, is also a distinct skill set that is beyond the scope of the Internet self-efficacy measure developed in the present study.

Future research should investigate the interplay among Internet self-efficacy, stress and on-line support. Social support should relieve stress (Cutrona & Russell, 1987). The amount of perceived and actual technical support available has been found to increase computer self- efficacy (Compeau & Higgins, 1995).

Social cognitive theory recognizes instances of reciprocal causation that cannot be assessed through one-time cross-sectional surveys. For example, perceived outcomes of behavior directly affect the future performance of behavior, so a reciprocal causation path from behavior to expected outcomes should be examined. Likewise, the successful performance of a behavior should have a direct reciprocal effect on self-efficacy perceptions. Longitudinal research would supply the time series data needed to test predictive validity and the reciprocal causality that should exist between Internet self-efficacy and Internet use.

Our measure and conceptualization of Internet self-efficacy should help guide future efforts to close the digital divide. Social cognitive theory suggests four mechanisms that can be used to formulate and understand intervention strategies. The sources of self-efficacy that should be investigated include enactive mastery, vicarious experience, verbal persuasion and physiological responses (Bandura, 1997; Oliver & Shapiro, 1993).

Enactive mastery, gained by reflecting upon one's own successful past performances, is by far the most powerful source of self-efficacy. Enactive mastery of complex behaviors such as Internet use can be bolstered by steadily building upon the successful attainment of subskills that are relatively easy to master. However, infrequent trips to standalone computer labs or short-term immersion courses are unlikely to be effective.

Vicarious experience gained by observing others as they master the Internet can be both positive and negative. Vicarious experience is generally thought to be less effective than direct (enactive) experience with one important exception: the observation of failure on the part of similar others can have a particularly devastating effect on self-efficacy judgments. So, when testing intervention strategies aimed as closing the digital divide, research should pair novice users with Internet experienced peer tutors (including on-line); this could be an effective method to increase Internet self-efficacy judgments and subsequent use. The common practices of holding group computer labs in educational settings and using the "buddy system" to share computer resources can have a negative effect on self-efficacy in populations where Internet skills are generally low. Here, observing the failures of peers is likely to discourage (i.e., decrease self-efficacy judgments) those struggling with Internet use and may also negatively affect those who have achieved early success. Individualized instruction would thus be preferable for new Internet users. Failing that, labs could be redesigned with partitions or staggered seating to restrict information about the failures of peers. Novice Internet users can also be persuaded to have greater self-efficacy through verbal feedback about their performance, if delivered by competent and credible evaluators. Feedback about the capability of the novice user is highly effective. However, verbal feedback must be constructive in order to increase self-efficacy. Telling new Internet users that they can succeed only through hard work or that they need to work harder is likely to lower self-efficacy in the long run since that conveys the message that the user must have been deficient to begin with to require such hard work to succeed.

Our alternative formulation of the digital divide problem is by no means intended to minimize the role played by race and class discrimination in creating unequal access to the Internet. Indeed, there are likely to be race and class differences in Internet self-efficacy as well. Research that examines the suggested intervention strategies within specific ethnic and socioeconomic groups is needed. We hope that our self-efficacy conceptualization will alert reformers to the possibility that although providing computers and network connections eliminates the physical barriers to access, psychological barriers may still remain.

However, by vigorously publicizing the Digital Divide as an important social problem while simultaneously defining it in terms of race and class, there is the risk that deficient computer skills will come to be viewed as stereotypical of the groups that are presently below the divide. Social cognitive theory warns us that when this happens the stereotyped group tends to adopt the stereotype as a standard for their own self-comparisons (Bandura, 1997), lowering their self-efficacy and imposing a further psychological barrier to successful Internet use. Thus, as researchers attempt to uncover the underlying barriers influencing the divide, it might be more productive to conceptualize the divide in terms of the barriers shared by all novice users of the Internet.

The validity of a construct cannot be established by a single study. Without longitudinal data it is hard to distinguish cause and effect ordering (Pedhazur, 1982) and the reciprocal causation mechanisms specified by social cognitive theory could not be examined. The convenience sample used restricts the generalizability of the results. Prior Internet experience was a single item measure and the Internet stress measure had a marginal level of internal consistency, calling into question the reliability of those results. Finally, only a single measure of Internet self-efficacy was employed. Construct validation procedures following the multi-trait multi-method approach (Anastasi, 1988) require the development and comparative analysis of multiple measurement methods using alternative approaches to self-efficacy measurement (Lee & Bobko, 1994).

The present study represents a further step in understanding the role that Internet self-efficacy plays in the use of the Internet. Finally, research suggested on the development of self-efficacy judgments (e.g., enactive mastery, vicarious experience, verbal persuasion and physiological responses) would help to further validate the Internet self-efficacy scale presented in this study as well as increase our overall understanding of Internet use.

Footnotes

1. Find companionship, meet new friends, maintain relationships, get in touch with people I know, and meet someone in person whom I met on the Internet.

2. Feel entertained; Find a way to pass time; Relieve boredom; Have fun.

3. Find current information like time, weather, stock prices and sports scores; Get information about products and services; Get immediate knowledge of big news events; Get information I can trust; Find information that is new to me; Encounter controversial information; Find information to complete a course assignment.

4. The other items were have trouble finding what I am looking for, have my computer freeze up, and get blocked by password protection

5. I feel my computer skills are inadequate; The things I can do on the Internet really don't amount to much; I can never accomplish what I want on the Internet.

6. Scored 1 for True and 0 for false with items indicating a lack of social support reflected. The items were: There is at least one person I know whose advice I really trust; There is really no one who can give me objective feedback about how I'm handling my problems; There is someone whom I feel comfortable going to for advice about sexual problems; I feel that there is no one with whom I can share my most private worries and fears; No one I know would throw a birthday party for me; There are several different people with whom I enjoy spending time; Most people I know don't enjoy the same things that I do; I feel that I'm on the fringe in my circle of friends; If I were sick and needed someone to drive me to the doctor, I would have trouble finding someone; There is no one I could call on if I needed to borrow a car for a few hours; If I needed a quick emergency loan of $100, there is someone could get it from; If I needed some help in moving to a new home, I would have a hard time finding someone to help me; In general, people don't have much confidence in me; Most of my friends are more successful at making changes in their lives than I am; I think that my friends feel that I'm not very good at helping them solve problems; I am closer to my friends than most other people.

References

Ajzen, I., & Fishbein, M. (1980). Understanding attitudes and predicting social behavior. Englewood Cliffs, NJ: Prentice-Hall.

Anastasi, A. (1988). Psychological testing. New York, New York: Macmillan Publishing Company.

Bandura, A. (1977). Self-efficacy:Toward a unifying theory of behavioral change. Psychological Review, 84, 191-215.

Bandura, A. (1997). Self-efficacy: The exercise of control. New York: W.H. Freeman.

Bandura, A. (1982).Self-efficacy mechanisms in human agency.American Psychologist, 37, 122-147.

Benton Foundation (1999). Digital Divide Network.[Online] Available:http://www.digitaldividenetwork.org/

Burkhart, M., & Brass, D. (1990). Changing patterns of patterns changing: The effects of a change in technology on social network structure and power. Administrative Science Quarterly, 35, 104-127.

Charney, & Greenberg, (in press). Uses and gratification of the Internet. Communication, technology and science. In C. Lin & D. Atkin (Eds.), Communication, technology and society: New media adoption and uses.Hampton Press.

Cohen, S., Mermelstein, R., Kamarck, T., & Hoberman, H. (1985). Measuring the functional components of social support. In I. G. Sarason & B. R. Sarason (Eds.), Social support: Theory, research and applications (pp. 73-94). The Hague, Holland: Martines Niijhoff.

Compeau, D., & Higgins, C. (1995). Computer self-efficacy: Development of a measure and initial test. MIS Quarterly, 19, 189-211.

Compeau, D., & Higgins, C. (1999). Social Cognitive Theory and individual reactions to computing technology: A longitudinal study. MIS Quarterly, 23, 145-158.

Cutrona, C. E., & Russell , D. W. (1987) The provisions of social support and adaptations to stress.Advances in Personal Relationships, 1, 37-67.

GVU (Graphic, Visualization and Usability Center). (1999) GVU"s Tenth Annual WWW User's Survey.Atlanta, GA:Georgia Institute of Technology. [Online] http://www.cc.gatech.edu/gvu/user_surveys/

Hancock, V. (1990). Overcoming technophobia. Educational Leadership, 47, 81-85.

Hoffman, D. L,. & Novak, T. P. (1998). Bridging the racial divide on the Internet. Science, April 17.

Hoffman, D. L.,Novak, T. P., & Schlosser, A. E. (2000). The evolution of the digital divide: How gaps in Internet access may impact electronic commerce. Journal of Computer Mediated Communication [Online] 5(3).Available: http://www.ascusc.org/jcmc/vol5/issue3/hoffman.html

Hunter, J. E., & Gerbing, D. W. (1982). Machiavellian beliefs and personality: Construct invalidity of the Machavellianism dimension. b Journal of Personality and Social Psychology, 43, 1293 � 1305.

Jöreskog, K., & Sörbom, D. (2000). LISREL 8.30. Scientific Software International Inc.

Kanner, A. D., Coyne, J. C., Schaefer, C., & Lazarus, R. S. (1981). Comparisons of two modes of stress measurement: Daily hassles and uplifts versus major life events. Journal of Behavioral Medicine, 4, 1�39.

Katz, J. E., & Aspeden, P. (1996, October).Motivations for and barriers to Internet usage: Results of a national public opinion survey.Paper presented to the 24^th Annual Telecommunications Policy Research Conference, Solomons, MD.

Katz, J. E.,& Aspden, P. (1997). A nation of strangers? Communications of the ACM, 40 (12), 81-86.

Lee, C., & Bobko, P. (1994).Self-efficacy beliefs: Comparison of 5 measures. Journal of Applied Psychology, 79, 364-369.

Lewis, C. (1985). An exploratory study of the influences of self-efficacy on the personal computer learner. Dissertation Abstracts International, 46. Cited in Oliver, T. A., & Shapiro, F. (1993). Self-efficacy and computers. Journal of Computer-Based Interactions, 20, 81-85.

Nahl, D. (1996). Affective monitoring of Internet learners: Perceived self-efficacy and success. Journal of American Society for Information Sciences, 33, 100-109.

Nahl, D. (1997). User-centered assessment of two Web browsers: Errors, perceived self-efficacy, and success. Journal of American Society for Information Sciences, 34, 89-97.

National Telecommunications and Information Administration (NTIA), (1999). Falling through the Net: Defining the digital divide. [Online] http://www.ntia.doc.gov/ntiahome/fttn99/contents.html

Nielsen NetRatings, (1999). Nielsen NetRatings. [Online] http://www.nielsennetratings.com.

Oliver, T. A., & Shapiro, F. (1993). Self-efficacy and computers. Journal of Computer-Based Interactions, 20, 81-85.

Pedhazur, E. (1982). Multiple regression in behavioral research. Fort Worth: Harcourt Brace College Publishers.

Pew Research Center (2000). Tracking online life: How women use the Internet to cultivate relationships with family and friendsRetrieved May 10, 2000 from the World Wide Web: http://www.pewinternet.org/reports/toc.asp?Report=11

Radloff, L. (1977). The CES�D Scale: A self-report depression scale for research in the general population. Applied Psychological Measurement, 1, 385�401.

Ren, W. (1999). Self-efficacy and the search for government information. Reference & User Service Quarterly, 38, 283-291.

Russell, D., Peplau, L., & Cutrona, C. (1980). The revised UCLA Loneliness Scale: Concurrent and discriminant validity evidence. Journal of Personality and Social Psychology, 39, 472�480.

Staples, D. S., Hulland, J. S., & Higgins, C. A. (1998). A self-efficacy theory explanation for the management of remote workers in virtual organizations. Journal of Computer-Mediated Communication [On-line], 3 (4). Available: http://www.ascusc.org/jcmc/vol3/issue4/staples.html

Stumpf, S. A., Brief, A. P., &Hartman, K. (1989). Self-efficacy expectations and coping with career-related events. Journal of Vocational Behavior, 31, 91-108.

Van Dijk, J., andHacker, K. (2000, June). The digital divide as a complex and dynamic phenomenon. Paper presented at the 50th Annual Conference of the International Communication Association, Acapulco, Mexico, 1-5 June.

Wellman, B., & Gulia, M. (1999).Virtual communities as communities:Net surfers don't ride alone.In Smith, M. A. & Kollock, P. (Eds.),Communities in Cyberspace (pp. 167-194). NY: Routledge.

About the Authors

Robert LaRose is a Professor in the Telecommunication Department at Michigan State University. He holds a Ph.D. in Communication Theory and Research from the Annenberg School at the University of Southern California.
Address: Department of Telecommunication, Communication Arts and Science #408, Michigan State University, East Lansing, MI 48823, 517-432-4528

JCMC 6 (1) September 2000

Message Board