Entrepreneur: Start & Grow Your Business

A social cognitive theory of Internet uses and gratifications: toward a new model of media attendance.


by LaRose, Robert^Eastin, Matthew S.

The addition of the Internet to the electronic media environment has renewed interest in the question of media attendance: the factors that explain and predict individual exposure to the media. Much of the research has been carried out by followers of the uses and gratifications tradition, who anticipated the medium as an exemplar of active media selection that could further validate the core tenets of that paradigm (Morris & Ogan, 1996; Newhagen & Rafaeli, 1996; Ruggerio, 2000). Instead, Internet research has introduced new conceptual and operational approaches and new variables that now challenge some of the basic assumptions, procedures, and findings of uses and gratifications. However, these findings have yet to be integrated into a comprehensive model of media attendance. Moreover, these relationships have been explored among college student samples and must now be extended to the general online population. The present research proposes and tests a model of media attendance inspired by Bandura's (1986) Social Cognitive Theory (SCT) that builds upon the conventional uses and gratifications approach by clarifying important explanatory constructs and identifying new ones.

Uses and Gratifications Meet the Internet

Numerous studies (e.g. Charney & Greenberg, 2001; Chou & Hsiao, 2000; Dimmick, Kline & Stafford, 2000; Eighmey & McCord, 1998; Ferguson & Perse, 2000; Flanagin & Metzger, 2001; Kaye, 1998; Korgaonkar & Wolin, 1999; LaRose, Mastro & Eastin, 2001; Lin, 1999; Papacharissi & Rubin, 2000; Parker & Plank, 2000; Perse & Greenberg-Dunn, 1998; Song, LaRose, Eastin & Lin, 2004; Stafford & Stafford, 2001) have applied uses and gratifications to the Internet. Collectively, these studies upheld one of the model's basic propositions (Palmgreen, Wenner & Rosengren, 1985), that gratifications sought explain individual media exposure. However, many Internet-related studies have also reconfirmed a basic weakness of uses and gratifications: They did not explain media exposure very well. Consistent with uses and gratifications studies of other media (cf. Palmgreen, Wenner, & Rosengren, 1985), the Internet studies that hewed most closely to the uses and gratifications tradition have explained less than 10% of the variance in Internet usage from gratifications (e.g., Ferguson & Perse, 2000; Kaye, 1998; Papacharissi & Rubin, 2000; Parker & Plank, 2000).

That the Internet is in many ways a unique medium has not escaped the attention of researchers. The time-honored list of gratifications derived from early television studies (notably, Greenberg, 1974; Rubin, 1983) has been expanded to explore unique facets of the Internet medium. For example, Papacharissi and Rubin (2000) proposed interpersonal communication gratifications, recognizing that communication functions like e-mail and chatrooms are common modes of Internet usage. Korgaonkar and Wolin (1999) found dimensions of information, interactive, and economic control. Other new gratification dimensions have included: problem solving, persuading others, relationship maintenance, status seeking, and personal insight (Flanagin & Metzger, 2001); Song et al.'s (2004) virtual community gratification; Charney and Greenberg's (2001) coolness, sights and sounds, career, and peer identity factors; and Stafford and Stafford's (2001) search and cognitive factors. Stafford and Stafford (2001) achieved a modest increase (to 21%) in the variance explained in Internet usage, mostly from the addition of a search factor (i.e., that accessing search engines was an important motivation for using the Internet) to more conventional information seeking and entertainment gratifications.

Others innovated with conceptual and operational definitions, creating what might be called prospective, or expected, gratifications. These ask respondents what they expect from the Internet in the future as opposed to those that they seek in the present or have obtained in the past. This is a departure from the gratifications sought/gratifications obtained (GS/GO) formulation that has long guided uses and gratifications (Palmgreen et al., 1985). Studies that have employed prospective measures (e.g., Charney & Greenberg, 2001; LaRose, Mastro, & Eastin, 2001; Lin, 1999) have consistently doubled, tripled, or quadrupled the amount of variance explained in Internet attendance behavior compared to conventional approaches.

A Social Cognitive Perspective of Uses and Gratifications

Prospective gratification measures were initially an innovative means of understanding the medium before it was widely distributed in the population (e.g., Lin, 1999). However, they are also consistent with a view of media attendance derived from Bandura's (1986, 1989) Social Cognitive Theory (SCT), which offers a theoretical explanation for the often-observed (for example, Papacharissi & Rubin, 2000) empirical relationship between media gratifications and media usage. SCT is familiar to media scholars in its earlier incarnation as Social Learning Theory (Bandura, 1977), as a theory of media effects. However, SCT is a broad theory of human behavior that may be applied to media attendance as well. SCT posits reciprocal causation among individuals, their behavior, and their environment. Within SCT, behavior is an observable act and the performance of behavior is determined, in large part, by the expected outcomes of behavior, expectations formed by our own direct experience or mediated by vicarious reinforcement observed through others. Thus, media usage is overt media consumption behavior (usage of the Internet in the present case), and it is determined by the expected outcomes that follow from consumption. Since expected gratification outcomes may be formulated from vicarious observation of others' behavior (Eastin, 2002) they can explain consumption both among prospective future users of the Internet (as in Lin, 1999) and current users.

Uses and gratifications can be understood in socio-cognitive terms. Where uses and gratification researchers have explored gratifications, SCT proposes expected outcomes and where uses and gratifications researchers posit needs, SCT proposes behavioral incentives. Expected positive outcomes of Internet exposure should cause further exposure. What people have gotten in the past from the Internet is an important part of the basis for their current expectations about it. However, expectations are also shaped by vicarious learning, based on observations of the experiences of others, and also self-efficacy (see below). However, it is the current expectation about outcomes of behavior that best determines behavior.

The expected outcomes are organized around six basic types of incentives for human behavior: novel sensory, social, status, monetary, enjoyable activity, and self-reactive incentives (Bandura, 1986, pp. 232-240) and these are theoretically constructed rather than statistical[y derived from exploratory factor analysis as in the uses and gratifications tradition. An analysis of these categories against Internet gratifications (LaRose et al., 2001) revealed that conventional uses and gratifications research underemphasized status and monetary incentives that had significant positive correlations with Internet usage (see a]so Charney & Greenberg, 2001; Flanagin & Metzger, 2001; Krgaonkar & Wolin, 1999). When expected outcome measures reflecting the full range of these incentive categories were subjected to exploratory factor analysis, a "new" virtual community dimension was uncovered that drew heavily on the status incentives lacking in conventional uses and gratifications research (Song et al., 2004). Despite few differences (notably the inclusion of measures of habit strength in gratification dimensions), other SCT incentive categories parallel conventional uses and gratifications dimensions. Activity incentives, predicated on the desire to take part in enjoyable activities, correspond to the entertainment gratifications. Self-evaluative incentives, which involve attempts to regulate dysphoric moods, parallel "pass time" or "boredom" gratifications. Novel sensory incentives include the search for novel information, and they are similar to information seeking gratifications. Social incentives stemming from rewarding interactions with others correspond to social gratifications.

However, the gratifications sought-gratifications obtained formulation (Palmgreen, et al, 1985) does not precisely match the concept of outcome expectations, or the subjective probability that a particular outcome will be obtained for future behavior. Gratifications sought reflect wished-for outcomes (e.g., I hope to find an e-mail from home) but not necessarily expectations of achieving the outcome through our present behavior (but the folks e-mailed yesterday, so I don't expect one today). Comparing gratifications sought with those obtained reflects the outcomes achieved in the past but not necessarily the likelihood that they will be repeated in the present by engaging in further media consumption. Rather, SCT assumes that outcome expectations are continually updated as a result of self-observation of our own experience and (vicarious) observation of the behavioral consequences that occur to others.

The socio-cognitive mechanism can be perceived in "A General Media Gratifications Model" (found in Palmgreen et al. 1985, p. 17) in which media consumption affects perceptions of gratifications obtained, which feed back to beliefs and expectations about media alternatives, which determine gratifications sought, which determine media consumption behavior. From the SCT perspective, the expectations about media alternatives--specifically the outcomes that our media consumption behavior produces, organized according to the incentives that motivate human behavior--are what determine further media consumption. GS and GO are thus imprecise, and perhaps superfluous, ways of describing the construct of outcome expectations. The SCT formulation differs in that these expected consequences are themselves the psychological origins of media behavior. The expected outcomes thus produce the "need" for media attendance.

This imprecise match between outcome expectations, GS, and GS-GO formulations might explain the pattern of weak relationships observed between gratifications and media consumption (e.g., Papacharissi & Rubin, 2000). Unsuccessful attempts by researchers (Babrow & Swanson, 1988) to distinguish outcome expectations (derived from a related theory, the Theory of Planned Behavior (Ajzen, 1985) from gratifications perhaps indicated that the two are related constructs. However, the distinction between outcome expectations and gratifications is potentially consequential. Some gratifications sought could be negative predictors of media behavior (if we don't expect to achieve them), others positive ones, but in the aggregate are just possibly confounded ones. Comparing gratifications obtained with those sought compounds the problem (e.g., with gratifications that are obtained but not sought, those that are sought but never realistically expected) that may have no reliable relationship to media behavior. Outcome expectations cut through the ambiguity because they "reflect current beliefs about the outcomes of prospective future behavior but are predicated upon comparisons between incentives expected and incentives attained in the past" (LaRose et al., 2001, p. 399).

Extending Uses and Gratifications

SCT also suggests new concepts that may extend our understanding of uses and gratifications and their impact on media behavior. Two additional mechanisms of SCT, self-efficacy and self-regulation, are particularly heuristic.

Self-efficacy is belief in one's capability to organize and execute a particular course of action (Bandura, 1986). Self-efficacy is particularly relevant for novice users who have not yet acquired the requisite skills to obtain useful information and deal with the discontents of life online, from viruses to balky home Internet connections. It was directly related to Internet usage (Eastin & LaRose, 2000; LaRose et al., 2001), and also acted on usage indirectly, through expected outcomes. In other words, as Internet users become more self-efficacious, their expectations that they will obtain specific outcomes (e.g., finding useful information) also increase, and that encourages more usage. Prior experience with the Internet in turn causally preceded Internet self-efficacy (Eastin & LaRose, 2000), probably through the process of enactive mastery (Bandura, 1986), in which users gradually master complex tasks.

The SCT construct of self-regulation (Bandura, 1991) describes how individuals monitor their own behavior (self-monitoring), judge it in relation to personal and social standards (judgmental process), and apply self-reactive incentives to moderate their behavior (self reaction). Self-regulation is an important point of distinction between SCT and stimulus-response theories of human behavior in that self-generated influences free the individual from blindly following the dictates of external reinforcement. However, when self-regulation fails, increased media consumption may be expected. This issue has been conceptualized in terms of habit and deficient self-regulation (LaRose, Lin, & Eastin, 2003).

In simplest terms, a habit is a recurring behavior pattern. Habit is a well-established predictor of behavior (Oulette & Wood, 1998; Triandis, 1980). Although long overlooked in communication research (cf. Rosenstein & Grant, 1997; Stone & Stone, 1990), recent qualitative research suggests that a great deal of media behavior is habitual (Adams, 2000). Uses and gratifications researchers have associated habit with "ritualistic gratifications," such as passing time (Rubin, 1984), that are conceptually still part of an active selection process. However, recent research (e.g., Aarts, Verplanken, & van Knippenberg, 1998; Bargh & Gollwitzer, 1994) suggests that habit is a form of automaticity, a pattern of behavior (e.g., checking one's e-mail) that follows a fixed cognitive schema, triggered by an environmental stimulus (e.g., seeing one's computer desktop in the morning) or by recalling a goal (e.g., keeping up with one's associates), and performed without further self-instruction. This is outside the realm of active media selection presumed in uses and gratifications research At best, automatic media consumption behaviors were initially framed by active considerations that were eventually forgotten (cf. Stone & Stone, 1990): We carefully evaluated our options the first time we used e-mail but by the hundredth time we did not. Within SCT habit is a failure of the self-monitoring subfunction of self-regulation. Through repetition we become inattentive to the reasoning behind our media behavior, our mind no longer devotes attention resources to evaluating it, freeing itself for more important decisions.

Habit strength is not a catch-all for uses and gratifications but rather represents patterns of behavior established by past thinking about outcome expectations/ gratifications that is no longer repeated in the present. Habit strength is expected to influence ongoing behavior, independent of current active thinking about expected (gratification) outcomes. Habit should be causally determined by outcome expectations, which precede habit in time. Habit strength should be preceded by self-efficacy, since users are unlikely to be inattentive to a behavior they are still mastering. These relationships were confirmed in LaRose et al., 2003).

Deficient Self Regulation is defined as a state in which conscious self-control is diminished. It has been proposed as an explanatory mechanism for so-called "Internet addictions," more properly called "problematic Internet use" (LaRose et al., 2003). In that research, deficient self-regulation was directly related to Internet usage and also contributed to usage indirectly, through habit strength. As deficient self-regulation comes into effect, media behavior tends to become an end unto itself and no longer subject to active consideration of its expected outcomes. One important exception are self-reactive outcomes, through which users counteract the negative affect that results from personal problems that intensify with excessive media usage, part of a self-reinforcing "downward spiral" into problematic or addictive usage.

Still, habit and deficient self-regulation have not been clearly distinguished. Addictions, including behavioral addictions, are a form of habitual behavior (Marlatt, Baer, & Kivlahan, 1988) so the two constructs overlap. At the operational level, measures of habit lacked sufficient reliability and exhibited some degree of multi-collinearity with deficient self-regulation so LaRose et al. concluded that they had not clearly distinguished the two. They proposed a possible theoretical distinction: Habit could represent the failure of self-monitoring, while deficient self-regulation might represent a failure of the judgmental and self-reactive subfunctions. Both the conceptual definition of deficient self-regulation and its operationalization (based on symptoms of pathological gambling and substance dependence, e.g., "I feel tense, moody, or irritable if I can't get on the Web when I want") betray an intense, even painful, self-awareness of media consumption. Thus, deficient self-regulation reflects a distinct state of mind from one in which media consumers are inattentive, explaining how both might have independent effects on media attendance. Yet, the two should be related in that persons with deficient self-control may also be expected to engage in habitual behavior.

Are College Students Typical Internet Users?

Research on Internet usage has focused on college students, often with the reassurance that scholars are interested in the lawful relationships among variables that should be observable among many groups, including purposive samples of college students. But there are also some important ways that college students differ from the general population and these are particularly salient from the SCT perspective.

College students, and particularly the freshmen who populate large introductory classes, have relatively high levels of depression (Rich & Scovel, 1987) and depression is known to inhibit effective self-regulation (Bandura, 1991), possibly exaggerating the effect of that variable. One reason that freshmen are depressed is separation from their family and friends, perhaps heightening the importance of social and mood-elevating self-reactive outcomes. Indeed, college students demonstrate an especially heavy reliance on the Internet for social interaction and fun activities (Pew Research Center, 2002). Will the same motives found among college students affect Internet attendance in broader populations?

Relationships among deficient self-regulation, Internet self-efficacy, habit strength, and Internet usage were explored previously in research on so-called "Internet addictions" among college students (LaRose et al., 2003) and can be summarized as follows:

H1: Internet self-efficacy will be directly related to Internet usage.

H2: Internet habit strength will be directly related to Internet usage.

H3: Deficient Internet self-regulation will be directly related to Internet usage.

H4: Deficient Internet self-regulation will be directly related to Internet habit strength.

H5: Internet self-efficacy will be directly related to Internet habit strength.

The causal relationship of prior Internet experience to Internet self-efficacy has also been verified in college student populations (Eastin & LaRose, 2000). Self-efficacy is conceived as the product of progressive mastery of behavior that increases with experience. If habit is simply a recurring behavior pattern, then the amount of prior experience with the Internet should be directly related to habit strength. Theoretically, repetition fosters a growing inattentiveness to behavior, undermining self-monitoring. Triandis (1980) pointed out that the best predictor of behavior is past behavior. Indeed, this truism may be largely responsible for the scant attention paid to habits in media research, since it has the ring of a tautology: One measure of behavior predicts another measure of behavior (and so what?). However, we propose that the impact of prior experience on current behavior is explainable entirely through the causal paths among socio-cognitive constructs.

H6: Prior Internet experience will be directly related to Internet self-efficacy.

H7: Prior Internet experience will be directly related to Internet habit strength.

To complete the model of Internet usage, the role of outcome expectations (a.k.a uses and gratifications) also must be specified. LaRose et al. (2003) examined a single type of outcome expectation, self-reactive outcomes, which uses and gratifications researchers would perhaps recognize as "pass time" gratifications. They found that self-reactive outcome expectations were causally related to Internet usage and also acted on Internet usage through deficient self-regulation. Self-reactive outcome expectations were themselves preceded by Internet self-efficacy and followed by habit strength.

However, other types of outcome expectations should also explain Internet usage. LaRose et al. (2001) found that the gratifications of the Internet, reconceptualized as outcome expectations reflecting other incentive categories recognized by SCT, were positively related to Internet usage. However, these relationships were not arrayed in an overall causal model and self-reactive and activity outcomes were combined to attain satisfactory reliability. Thus, the self-reactive and activity constructs will be separated to match the conceptual distinction between these two incentive categories.

Self-efficacy should precede each of the the various types of outcomes and habit strength should follow them. Users need to learn how to successfully obtain social, status, activity, novelty, and monetary gratifications as much as self-reactive ones. But once they achieve satisfactory means for attaining those outcomes, they should become increasingly inattentive to specific behaviors that support them.

Past research expended a great deal of effort distinguishing gratification dimensions, but from the SCT perspective expected outcomes are a unitary construct, representing the mechanism of learning through experience. Thus, we represent gratifications, now understood to be different types of expected outcomes grouped into six incentive categories, as first-order concepts that are part of a second-order construct, outcome expectations.

H8: Expected a) activity, b) social, c) status, d) novel, e) self-reactive, and f) monetary Internet outcomes will be directly related to Internet usage.

H9: Internet serf-efficacy will be directly related to expected Internet outcomes.

H10: Expected Internet outcomes will be directly related to Internet habit strength.

An important exception is noted for the relationship between self-reactive outcomes and deficient self-regulation. Self-reactive outcomes bear a unique relationship to deficient self-regulation. The use of the media to adjust internal states should be the main type of incentive susceptible to triggering the spiral of excessive usage and dysphoria thought to lead to problematic media usage. Thus,

H11: Self-reactive outcomes of Internet usage will be positively related to deficient Internet self-regulation.

These hypothesized relationships are represented in the path diagram shown in Figure 1. The final path analysis (see Results, below) was identical except for one path that was proposed but not statistically significant, another that narrowly missed significance (both indicated by light dotted lines) and another that was not initially proposed but uncovered during analysis (indicated by a heavy dotted line), in the interest of space only the final path diagram is shown. For the sake of clarity the word "Internet" has been omitted from the labels in figures and tables.

[FIGURE 1 OMITTED]

Research Methods

Procedure

To obtain a diverse sample of the general population of adult Internet users, respondents were recruited by mail from two Midwestern communities to complete an online survey in April and May of 2002. Both communities included a major university and surrounding counties. A commercial mailing list vendor provided a random sample of household addresses in the designated communities. The initial mailing included a letter advising respondents of the purpose of the study and their rights as human subjects. Half the letters requested that the survey be filled out by a male head of household and the other half by a female head of household, if such a person were available. Also included in the envelope was a nominal cash incentive and a postcard with the URL and a respondent ID for the survey. Internet users were instructed to use the card and ID number the next time they went on the Internet. Non-users were instructed to indicate their gender and year of birth and return the card by mail so that response rates could be calculated and the results compared to U.S. Census data.

Respondents

Of the 1100 solicitations sent, 170 (15%) bad addresses were returned; leaving a total usable sample of 930. A total of 331 responded to the solicitation. One hundred and seventy-two Internet users completed the survey and 159 returned the non-Internet user postcard (36% total response rate). Recent research assessing response rates (Yun & Trumbo, 2000) indicates that the present rates were consistent with methods employing Web surveys. A rule of thumb for structural equation modeling, and also for regression analysis, is that there be 10 respondents for each link/ relationship in the model. The proposed model has 16 links so the sample size was deemed adequate. There were no response difference by city and thus data were collapsed. As a total sample (N = 331) participants were 55% male and 45% female. The general population in the counties surveyed was 50% female (U.S. Census, 2002). Six percent of the participants were between the ages of 18-24 (census population = 17%), 48% were between the ages of 25-44 (census population = 30%) 34% were between 45-65 years old (census population = 40%), and finally, 13% were over the age of 65 (census population = 14%). The respondents were thus a somewhat biased sample of the respective populations from which they were drawn. However, a diverse sample of adult respondents was obtained and therefore this sample was deemed suitable for the purpose of this study which was to examine relationships between variables.

The non-Internet users (N = 159) were 48% male and 52% female and their mean age was 52 years old. Given current estimates of Internet penetration (54%, NTIA, 2002) we estimated that respondents at 504 (out of 938) of the valid addresses had access to the Internet, and thus, could have completed the online survey. Therefore, we estimate that the 172 people who completed the survey represent an Internet user response rate of 34%. Of those, 41% were female and 58% were male (with 1% not indicating their gender) with an average age of 42 years old. Eighty-nine percent were Caucasian, 5% were African American, 2% were Latino and the remaining 4% were Asian, Pacific Islander, Native American, or other. Forty-two percent of the sample had average household incomes under $50,000; the remaining 58% had incomes greater than $50,000. Educationally, participants ranged between 9-22 years beyond kindergarten (Mean = 16, S.D. = 2.61). Six respondents were removed from the sample for incomplete data, yielding a final sample of 167. A common rule of thumb in structural equation modeling is to have 10 cases for each link in the model. The proposed model had 16 links and so the size of the sample was deemed adequate.

Compared to an NTIA (2002) study completed in fall of 2001, the present sample tended to be older, better educated (including more with advanced degrees and fewer with high school diplomas), and disproportionately male. There were no income differences. There were few significant relationships between demographic variables and the main explanatory variables (females had lower self-efficacy, r = .26, and less deficient self regulation, r = .20, than males, and age was negatively, r = -.15, related to self-efficacy), and no significant correlations with the dependent variable so sample bias was not deemed an important issue. Within Social Cognitive Theory, demographic differences are attributed to explanatory variables (e.g., females have lower Internet self-efficacy due to the nature of their past experiences with the Internet).

Operational Measures

The questionnaire was introduced with the following description of Internet use: "Internet use includes sending or receiving electronic mail, visiting chatrooms, participating in discussion groups and visiting locations on the World-Wide Web." No distinction was made between work and leisure activity.

The usual procedure for analyzing gratifications in the uses and gratifications tradition is to conduct an exploratory factor analysis of the gratification items. However, in the present research a priori theoretical assumptions about the nature of the expected outcomes were available, in the form of the incentive categories recognized in SCT. Items were collected from prior uses and gratifications studies, rephrased as outcome expectations (i.e., "using the Internet how likely are you to ..." on a scale of 1-7, where 1 was very unlikely and 7 very likely, cf. Ajzen, 1985). These statements of outcome expectations were classified into SCT incentive categories by consulting the conceptual definitions found in Bandura (1986, pp. 233) and supplemented with items reflecting status and monetary incentives that were underrepresented in uses and gratifications research (cf. LaRose et al., 2001). Six categories of expected outcomes, one representing each incentive category, were subjected to confirmatory factor analyis. The means and standard deviations of the scales and their component items, along with confirmatory factor analysis results and their alpha coefficients, are found in Table 1.

Previous research (LaRose et al., 2003) left the distinction between habit strength and deficient self-regulation unclear. Accordingly, new items were developed by drawing upon theoretical works describing habitual behavior (Aarts et al, 1998; Bargh & Gollwitzer, 1994; Oulette & Wood, 1998) and LaRose et al.'s previous description of deficient self-regulation. The pool of items was subjected to an exploratory principal components factor analysis using varimax rotation. Two interpretable factors emerged, also shown in Table 1. These factors seemed to reflect the distinction between the self-observation subfunction of self-regulation on the one hand and the judgmental process and self-reactive subfunctions on the other hand.

The Internet Self-Efficacy Scale (Eastin & LaRose, 2000) was replicated (Table 1). Also from that study, a measure of Internet experience was computed by asking the number of years and months it had been since the respondent first started using the Internet.

The dependent Internet usage variable was the sum of the total number of minutes spent on the Internet in the typical weekday, the typical weekend day, and the day prior to the survey. An inspection of the distributions of responses to these items revealed that outliers were present and so a log10(1 + value) transform was applied to each one before summing the three items. The resulting composite index had a Cronbach alpha of .66 (M = 5.17, S.D. = 1.59).

Data Analysis

Pearson product-moment correlation coefficients and exploratory factor analyses were calculated using SPSS version 11.5 (SPSS, Inc., 2002). Structural equation analysis was completed with Amos version 4.0 (Arbuckle, 1999).

Results

Pearson product-moment correlations among the independent and dependent variables are shown in Table 2 and the results of structural equation modeling are shown in Figure 1. The model shown was a good fit to the data ([chi square] = 62.3, df = 34, RMSEA = .994, CFI = .071). As hypothesized, Internet usage was directly predicted by expected Internet outcomes ([rho] = .29), Internet habit strength ([rho] = .26), and deficient Internet self-regulation ([rho] = .15). The individual activity (r = .40), monetary (r = .27), novel (r = .36), social (r = .44), self-reactive (r = .46), and status (r = .53) expected outcome categories all had significant (p < .001) zero-order correlations with usage. Expected Internet outcomes were a second-order factor consisting of status, activity, self-reactive, social, novel sensory, and monetary incentives, preceded by Internet self-efficacy ([rho] = .55). Internet habit strength was predicted by expected Internet outcomes ([rho] = .26). Deficient self-regulation predicted Internet habit strength ([rho] = .39) and was itself causally determined by self-reactive outcomes ([rho] = .43). Finally, Internet self-efficacy was also predicted by prior Internet experience ([rho] = .38). However, the hypothesized relationships between prior experience and Internet habit strength was not significant ([rho] = .03, p = .63). The relationship between Internet self-efficacy and habit strength ([rho] = .14, p = .069) narrowly missed significance.

An inspection of the modification indices suggested a causal link from self-efficacy to novel expected outcomes. Since this outcome category represents information seeking on the Internet and the task of seeking useful information is likely to require a high degree of confidence in one's ability, this was accepted as a logical extension to the model. Correlated error terms between self-reactive outcomes and both activity (r = .44) and social outcomes (r = .18), and between novel and monetary outcomes (r = .37), not shown, were added to improve fit. The model explained 42.2% of the variance in the dependent variable, Internet usage.

Discussion

The present results both affirm the uses and gratifications paradigm and extend it to a theory of media attendance grounded in Social Cognitive Theory. A basic implication of uses and gratifications, that media exposure may be predicted from media gratifications, was upheld. Indeed, by instituting new operational measures of expected gratifications, it was possible to predict media consumption to an unprecedented degree. However, new variables from SCT improved the explanatory power of gratifications, here reconstrued as outcome expectations.

Expected activity outcomes, which closely parallel entertainment gratifications in uses and gratifications research, and social outcomes/gratifications were significantly related to usage, as they had been in prior uses and gratifications research involving college students (e.g., Kaye, 1998; Papacharissi & Rubin, 2000), but with more variance explained by the current expected outcomes formulation. Unlike the previous studies, novel outcomes (paralleling information gratifications), self-reactive outcomes (parallel to "pass time" gratifications) were also significantly related to usage, when conceived as expected outcomes of Internet usage. Monetary outcomes, somewhat overlooked in previous research, were also significantly related to usage. Status outcomes, a gratification/outcome dimension identified by SCT but underrepresented in prior uses and gratifications research, had the highest zero-correlations with Internet usage of all. The perceived ability of the Internet to improve one's lot in life thus emerges as a powerful motivating factor for the use of the medium.

Uses and gratifications research, including Internet studies, have tended to subsume habit in other gratifications dimensions, usually under either an entertainment or "pass time" factor. Here, it emerged as a powerful and independent predictor of media exposure even after the effects of gratifications sought/expected outcomes had been accounted for. This finding supports the conceptualization of habit strength as a distinct construct from gratifications/expected outcomes. The correlation between habit strength and expected outcomes perhaps indicated the availability of memories of past active media selection processes, in the form anticipated by uses and gratifications research, that had become dormant with repetition. In this vein, among newer Internet users (those who had been online less than three years) the correlations between expected outcomes and usage were higher than among those with more experience. For example, the correlation between activity outcomes and usage was .54 for new users, compared to .34 for the more experienced ones. This could well indicate that the newer users were making active media selection decisions on the basis of expected outcomes while veteran users had lapsed into more habitual modes of Internet consumption.

The relationship of habit and deficient serf-regulation was further clarified. Habit perhaps indicates a failure of the first of the three subfunctions of serf-regulation proposed by SCT, self-observation. As such, this aspect of unregulated media behavior is closely related to notions of automaticity (Bargh & Gollwitzer, 1994). Deficient self-regulation derives from the failure of the judgmental and self-reactive subprocesses of self-regulation. It reflects a conscious failure of self-control wherein individuals struggle with themselves to judge their own behavior against appropriate standards and to apply incentives to moderate their consumption. The findings supported the proposed theoretical relationship between the two constructs wherein deficient self-regulation adds to habit strength.

Internet self-efficacy causally preceded Internet usage and was in turn determined by prior Internet experience as it was in previous college student samples (Eastin & LaRose, 2000). The hypothesized relationship between self-efficacy and habit strength was not observed, falling just short of statistical significance. It appears that self-efficacy acts on habit strength through expected outcomes. The inattentiveness to one's own behavior that signals habit formation thus could be more determined by a gradual cessation of active thinking about outcomes/gratifications rather than inattentiveness accompanying task mastery.

The lack of a causal connection between experience and habit strength, despite a significant zero-order correlation (r = .30, p < .001), belies the rival "tautology hypothesis" about the role of habit in human behavior; namely, that habit is "just" prior behavior and that relationships between two measures of behavior are not theoretically meaningful. An alternative causal model in which a direct link from experience (prior behavior) to usage (current behavior) was hypothesize also failed to produce a significant relationship, and did not diminish the causal link between habit and usage, again despite a highly significant zero-order correlation (r = .35, p < .001) between experience and usage. These findings suggest that habit strength indeed has an impact on behavior that can be described in social cognitive terms.

The Internet emerges from the present study as something of a distinctive medium, but perhaps not in ways previously described. It is not primarily a social medium (e.g., Papacharissi & Rubin, 2000), but also a medium through which enjoyable activity, self-reactive, monetary, novel (i.e., informational), and, above all, status incentives can be obtained. That the Internet is a medium of social interaction is indisputable, but a question now arises as to the purpose of the social interaction. Prior research surrounding the Internet Paradox (Kraut, Patterson, Lundmark, Kiesler, Mukophadhyay, & Scherlis, 1998) focused on social interaction as a means of securing social support and thereby improving psychological well-being. Now it appears that social status, not social support, might be the prime mover in Internet usage. Perhaps by finding like-minded individuals on the Internet and expressing ourselves in those venues we enhance our social status. Or, recalling Turkle's (1995) Life on the Screen ethnography, perhaps the Internet is a means of constantly exploring and trying out new, improved versions of our selves. From this we should begin to empirically explore online personal development (as self or with virtual others) as well as social maintenance (a support mechanism).

Limitations

The generalizability of the present research is limited by the geographic scope of the sample. The sample contained disproportionately small representations of young people and males. As a one-shot survey study, the direction of causation cannot be established. Indeed, within SCT reciprocal causation is recognized. For example, self-efficacy is a precondition for successful performance of a behavior, but successful performance also increases self-efficacy.

Implications for Further Research

Internet usage was broadly defined. Future research might distinguish Internet applications (e.g., e-mail vs online chat), functions (e.g., entertainment vs news Web sites), or settings (e.g., work vs leisure). However, in keeping with the operational procedures recommended in SCT, it will be important to match the explanatory constructs (e.g., expected e-mail outcomes, e-mail self-efficacy, e-mail habits, etc.) to achieve satisfactory results.

Habit strength, deficient self-regulation, and self-efficacy might extend to other forms of media attendance. Many media consumption behaviors (e.g., tuning in the evening news) would seem to be habit-prone. While few mass media require skills as complex as the Internet, there are perhaps parallel media self-efficacy constraints. Anyone who has ever given up recording a favorite show because programming the recorder was "too much hassle" may be said to suffer from a self-efficacy deficit. Television addiction has been described (by Kubey & Csikszentmihalyi, 2002) in the same terms of behavioral addiction that underlie deficient self-regulation. Here, we found the construct useful in explaining media attendance in a normal population of media consumers; perhaps that would extend to other media as well.

The present research suggests new departures from uses and gratifications traditions. Redefining gratifications as expected outcomes may have merit on both conceptual and operational levels. Secondly, gratification dimensions from previous research may have neglected some potentially important variables, particularly the status that media attendance may confer. Third, habit strength appears to be a distinct construct from gratifications, as early conceptualizations (e.g., Palmgreen et al., 1985, p. 17) observed, but later research neglected.

More fundamentally, the present findings suggest that active selection of media that best meet personal needs is not the sole mechanism explaining media attendance. Active selection dominates when new media alternatives appear or when personal routines are disrupted. Self-efficacy beliefs about one's ability to utilize alternative media channels may also contribute to media selection. Thereafter, repeated consumption is increasingly habitual and automatic as we turn our attention elsewhere. Once habits are established, users no longer think through whether one alternative or another is a better way of obtaining a particular outcome. Users still monitor their overall level of Internet usage and apply self-reactive incentives to adjust the amount to appropriate levels, as defined by personal or social norms. But some users may lose the power to self-regulate, perhaps through a process of operant conditioning (cf. LaRose et al., 2003) to self-reactive outcomes, and in extreme cases they might develop a media dependency. SCT provi