SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Method
  5. Results
  6. Discussion
  7. References
  8. Appendices

This study explores the effects of Internet self-efficacy and search task specificity on the self-efficacy outcome and task perseverance of finding online health-related sites that contain attributes of website accountability, as established by the American Medical Association (AMA). In a mixed 2 (self-efficacy) x 2 (search task specificity) repeated-measures experimental design, participants conducted two search tasks (general and specific) that varied in the amount of task difficulty. When search task specificity was taken into account, there was an Internet self-efficacy and task specificity interaction according to which high Internet self-efficacy participants locate sites higher in website accountability in the general search task (the more difficult search task) than their low self-efficacy counterparts. There was no significant difference in website accountability for the specific search task (the less difficult task). High Internet self-efficacy participants also demonstrated more task perseverance than their low Internet self-efficacy counterparts.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Method
  5. Results
  6. Discussion
  7. References
  8. Appendices

The growth of the Internet provides an unprecedented opportunity for the public to access a cornucopia of health-related information. The vast compendium of health-related information on the Web appeases the public’s long-standing desires for more detailed medical information (Charles, Gafni, & Whelan, 1997) and a more participatory role in health management (Guadagnoli & Ward, 1998). Because approximately 30%–50% of medication misuse is a result of lack of information (Farley, 1995), the potential reliance on the Web as a source of information may contribute positively to public health. Information retrieved from online sources is often used by patients in discussions with health-care providers (Aspden & Katz, 2001), which suggests that the Web can empower individuals through enhanced interactions (Rice, 2001).

That this opportunity, as well as opportunities to promote education and affective well-being, may not be fully taken advantage of is the basis of much of the research on the manner in which self-efficacy influences the social benefits of technological utilization (Bandura, 2002). Underlying this line of research is the notion that access to information communication technologies (ICTs) is not entirely a function of physical access to resources, but also a function of the ability to utilize such resources effectively to achieve desired outcomes. According to social cognitive theory, the ability to utilize ICTs successfully to achieve desired outcomes is predicated in part by self-efficacy.

Bandura (1997) defines self-efficacy as an individual’s belief in his or her ability to perform a task successfully. Self-efficacy arises from the interpretation of information from primarily four information sources: mastery experience (e.g., previous performance), vicarious experience, social influence, and emotional states. Based on these four sources of information, individuals evaluate information about their capabilities and regulate their choices and efforts accordingly. Thus, self-efficacy can determine not only what endeavors to pursue, but also how much effort will be put forth toward these endeavors and the individual’s resilience toward adversity. With respect to ICTs, one may have access to ICTs, but low self-efficacy may hinder the ability of an individual to utilize them to reap their associated benefits, including locating online health-related information with accountability standards.

Silbert, Lundberg, and Musacchio (1997) defined accountability standards as a “set of quality moorings” to help “consumers and professionals alike to reasonably judge whether what they are reading is credible, reasonable, or useful.” These accountability standards include identification of authorship, attribution of sources, currency of content, and disclosure of site ownership and sponsorships. These accountability standards were subsequently expanded by the American Medical Association (Winker et al., 2000). Recent studies found an association between presence of accountability standards and information accuracy (Kunst, Groot, Latthe, Latthe, & Khan, 2002; Meric et al., 2002; Winker et al., 2000).

Thus, locating online health information with accountability standards is an important outcome in self-efficacy research because many health-related sites provide information that is of poor quality, inaccurate, and inconsistent with established professional guidelines (Berland et al., 2001; Griffiths & Christensen, 2000). To reap the benefits of ICTs, one must be able to parse out the quality of health information available online. Thus, this study is concerned with the effects of Internet self-efficacy on finding health-related information that is accountable within the AMA guidelines (Winker et al., 2000).

Internet Self-Efficacy

Bandura (1997) conceived self-efficacy as an individual’s self-perception that varies across circumstances, rather than as a global disposition that can be measured by a single omnibus scale. Thus, domain specific measures of self-efficacy should assess the different levels of task demands necessary for successful completion within a specific domain.

Per Bandura’s suggestion that self-efficacy scales be tailored to specific domains, several self-efficacy scales pertaining to the computer and Internet domain have recently been developed (Compeau & Higgins, 1995; Eastin & LaRose, 2000). Because self-efficacy is context-specific, these self-efficacy measures assess the competence to perform a range of tasks associated with the computer (Compeau & Higgins, 1995) and the Internet (Eastin & LaRose, 2000). Moreover, the judgment of what one can accomplish with the skills possessed, rather than the skill set in and of itself, is the basis of self-efficacy. Thus, these scale items differentiate between component skills (e.g., opening a Web browser) and the behaviors one can accomplish with the skill set (e.g., finding information on the Web). Studies utilizing such scales posit that people with higher computer/Internet self-efficacy are more likely to achieve the benefits associated with utilization of ICTs.

A self-efficacy outcome pertinent to online health management is finding health-related information that is accountable. Although previous research has found a direct association between self-efficacy and perceptions of source credibility in an information seeking context (Eastin & LaRose, 2000; Hofstetter, Zuniga, & Dozier, 2001), there are two issues these studies do not address. First, these studies, as well as other studies that have explored the relationship between self-efficacy and various ICT-related outcomes, do not take into account the influence that the varying types of search tasks have on outcome. The specificity of the search task can determine a search’s level of difficulty and the amount of effort needed to successfully accomplish the online task (Kim & Allen, 2002). Thus, task difficulty would represent a form of adversity for individuals with low self-efficacy. Previous research indicates that individuals with low self-efficacy tend to give up under adversity (Bandura & Schunk, 1981). Thus, the direct relationship between self-efficacy and the outcome of finding information that is perceived to be credible, as demonstrated by previous studies, may occur for only difficult Internet search tasks and not for comparatively less challenging search tasks.

Second, previous related studies have not examined the effect of self-efficacy on website accountability. Instead, while these studies demonstrate a direct relationship between self-efficacy and the expectation of finding information that is perceived to be credible, the association is limited to expected outcomes, where one or more items from a self-reported questionnaire assess participants’ expectations of finding credible information online. Thus, these studies measure only expected outcomes rather than actual outcomes. An expected outcome “is a judgment of the likely consequences” (Bandura, 2002, p. 21) produced by a task, while an outcome is the actual realization of the consequences produced by the task. Expected outcomes, rather than actual outcomes, have been used in constructing and validating self-efficacy scales that have demonstrated the relationship between self-efficacy and finding credible online information (Eastin & LaRose, 2000; Hofstetter et al., 2001). It is feasible to ascertain the outcome of finding credible information online within an experimental setting where the website accountability standards are gauged. The current study addresses these shortcomings by examining the role of self-efficacy in two search tasks that differ in task specificity. This is achieved in an experimental study in which participants actively locate online health-related information.

Website Accountability as a Measure of Credibility

The credibility of a website is a perceived judgment of the believability of the source (Metzger, Flanagin, Eyal, Lemus, & McCann, 2003). Sources in credibility research have included media, organizations, and the individual spokesperson (O’Keefe, 2002); more recently, the individual website has been viewed as the source (Eighmey & McCord, 1998; Shon, Marshall, & Musen, 2000; Wathen & Burkell, 2002). What constitutes a website’s credibility is based on the individual receiver’s perceptions of the source’s believability. Because credibility judgments are subjectively based, they are not objective measures of the quality of information.

One solution is that rather than ascribing to the term credibility, we can gauge credibility based on objective measures ascribed by professionals, and, in particular for the topic of this study, by professionals of the medical community. With respect to online health-related information, one gauge of the believability of the source is the website accountability guidelines promoted by the AMA (Winker et al., 2000). These website accountability guidelines, sometimes referred to as benchmarks, were developed to assist consumers in assessing the quality of health-related websites. Emerging in response to concerns about the ubiquity of health-related information that does not meet medical standards (Berland et al., 2001; Griffiths & Christensen, 2000), these accountability measures are intended to assist consumers in looking for specific elements of the message to ascertain the credibility of the information provider. The AMA guidelines include notification of information currency (e.g., posting of date of content creation), content authorship, citation of references, selection of information (e.g., editorial board), and privacy policy practice. The specific elements in AMA website accountability guidelines (Winker et al., 2000) are similar to guidelines promoted by other health-related organizations (Eysenbach, Powerll, Kuss, & Sa, 2002) and earlier calls for quality published by the AMA (Silberg, Lundberg, & Musacchio, 1997).

While adherence to these guidelines in and of itself does not ensure protection against falsification of information, websites containing more AMA benchmarks (Silberg et al., 1997) were less likely to have inaccurate information than sites with fewer AMA benchmarks (Meric et al., 2002). For other related quality benchmarks, there is a statistically significant relationship between website quality markers and objective measures of information accuracy (Kunst et al., 2002). Thus, there is an association between information accuracy and presence of quality benchmarks (Kunst et al., 2002; Meric et al., 2002). Yet another reason that AMA benchmarks serve as markers of quality of information is that many sites do not have the AMA quality markers. For example, recent reviews of website adherence to AMA guidelines indicate that the majority of the sampled websites related to food allergies and cystic fibrosis did not indicate a date of last revisions (Anselmo, Lash, Stieb, & Haver, 2004; Stieb, Wang, & Haver, 2002). In contrast, adherence to these guidelines is required for all sites associated with the AMA (Winker et al., 2000).

While these guidelines have practical application, there is also a theoretical basis for them. Specifically, the literature on source credibility indicates that in the absence of knowledge of the source, judgments of the source’s credibility are made based on the efficacy of the content or message (Austin & Dong, 1994; Slater & Rouner, 1997). That is, the judgments of the message, such as the presence or absence of specific elements of a message, are the basis for the determination of the perceived credibility of the source. Such reliance on message cues to determine source credibility is also found in situations where little information is available about the source (Eagly & Chaiken, 1993). Recent research suggests that health information seekers have little knowledge of the source, as the majority of health information seekers do not have a specific site in mind during a Web search session (Pew, 2002).

Thus, it is feasible to ascertain the outcome of finding credible online information within an experimental setting by having participants select one website from their search session and subsequently have independent coders content analyze the selected site for the presence of elements of the AMA website accountability guidelines.

Self-Efficacy Outcome in ICTs Context—Website Accountability

Previous studies found a direct relationship between self-efficacy and the expectation of finding online information that is perceived to be credible (Eastin & LaRose, 2000; Hofstetter et al., 2001). Hofstetter et al. (2001) argued that for tasks that pertain to information seeking, credibility is associated with self-efficacy. Eastin & LaRose (2000) demonstrated that Internet self-efficacy is directly associated with outcome expectations related to attaining information, including procuring information that is perceived to be trustworthy. In both studies, self-efficacy has a direct effect on the expectation of finding credible health-related information. Such outcome expectations are a result of self-efficacy perceptions. Compeau and Higgins (1995) noted that “individuals with a weak sense of self-efficacy will be frustrated more easily by obstacles to their performance and will respond by lowering their perceptions of their capability.” Hence, those with lower self-efficacy will have lower expectations of finding online information that would be deemed credible. In addition, those with lower self-efficacy will expend less effort on the task. If one expects less, then one would expend less effort on the task. Moreover, the lack of effort toward the task also manifests itself in the face of obstacles, where those with lower self-efficacy tend to give up more quickly than those with higher self-efficacy. Thus, based on Bandura’s (1997) construct of self-efficacy and empirical studies on self-efficacy and outcome expectations of finding credible information, self-efficacy will influence the selection of sites with greater website accountability.

ICT-related studies that assess the effects of self-efficacy (Ford, Miller, & Moss, 2001; Nahl, 1996; Thompson, Meriac, & Cope, 2002; Tsai & Tsai, 2003) have utilized experimental settings in which participants actively performed an online task. While these studies have examined search efficiency and accuracy and other related performance variables, none of the studies have examined website accountability as a task outcome. In general, these studies explore the effect of self-efficacy on performance by comparing individuals who are high vs. low on self-efficacy (e.g., Nahl, 1996). Thus, we can expect that the accountability of the website that high-self-efficacy individuals find in their search tasks will be higher than the accountability ratings of the websites that low-self-efficacy individuals find in their search. These observations are stated in the following hypothesis:

H1: Website accountability will be higher for websites selected by high Internet self-efficacy individuals than for low Internet self-efficacy individuals.

Influence of Task Specificity on Self-Efficacy Outcomes

Self-efficacy studies in the ICT context often examine the effect of self-efficacy on outcomes for only one specific online search task. For example, one study examined the effects of self-efficacy for locating information on psychologists (Thompson et al., 2002), while another study examined the effects of self-efficacy for acquiring science information online (Tsai & Tsai, 2003). These and other related studies do not take into account that some online search tasks are inherently more difficult to complete successfully than others.

When the effects of self-efficacy are examined within one search task, any differences found for performance and subsequent outcomes do not reflect the difficulty of the task. Hence, the performance and outcomes of low- and high-self-efficacy individuals for one search task may provide an incomplete picture of the influence of self-efficacy. It is, thus, important to consider the influence of self-efficacy in terms of search tasks of varying difficulty. For comparatively less difficult search tasks, while high-self-efficacy individuals should demonstrate better performance and achieve better outcomes than low-self-efficacy individuals, the difference may not be significant. In contrast, for more difficult search tasks, the self-efficacy outcomes may be significant. Exploring the influence of task difficulty and self-efficacy on outcomes can reveal the extent to which self-efficacy influences desired outcomes in the ICT context. That is, rather than coming up with the general observation that self-efficacy influences the adoption and utilization of ICTs to retrieve accountable information, we can determine if the relationship differs according to search task, thus further defining the conditions by which self-efficacy is most influential. This, in turn, will provide a better understanding of which online health-related information-seeking goals can be achieved by high- compared to low-self-efficacy individuals.

The search task is one of the most important elements to understanding electronic information seeking (Ingwersen, 1992). Previous research has found that search task specificity influences search patterns and behavior because varying tasks require different information-seeking skills (Marchionini, 1989; Saracevic & Kantor, 1988). A search task is a goal in that it “is the manifestation of an information seeker’s problem and is what drives information-seeking actions” (Marchionini, 1995, p. 36). Task can be characterized by the specificity of the goal, which is defined as the variability of appropriate answers available to achieve the goal (Marchionini, 1995). Search task specificity has numerous related terminology, including “broad” and “specific” tasks (Saracevic & Kantor, 1988), “open” and “closed” tasks (Marchionini, 1989), “known-item search” and “subject search” (Drabenstott, 1984), and “general” and “specific” (Qiu, 1993), the latter of which are the terms used in the current study. A general search task pertains to a more abstract idea of the subject of the search task. In contrast, a specific task is more concrete in that it specifies a particular information element to be sought.

As the specificity of the task decreases, more effort is required to complete the task (Marchionini, 1995). This suggests, as defined in the current study, that a general search task would require more effort than a specific search task. The amount of effort required to complete a task presents a barrier to task completion. Thus, the specific search task would be least aversive to successful task completion, and the general search task would be most aversive. When self-efficacy is taken into account along with task specificity, differences in website accountability for high and low self-efficacy should be more pronounced in the more difficult search task, such that individuals high in self-efficacy would locate websites with higher website accountability than those low in self-efficacy. This interaction effect is stated in the following hypothesis:

H2: There will be an Internet self-efficacy by task specificity interaction, such that high-self-efficacy individuals will locate sites higher in website accountability than low-self-efficacy individuals in the more difficult search task, but there will be no significant difference in the easier search task.

Influence of Self-Efficacy and Search Type on Task Perseverance

Self-efficacy beliefs determine how long individuals will persevere when confronted with obstacles. Bandura (2002) noted that the management of information on the Internet is a complex task, one in which self-efficacy can determine the successful utilization of the electronic environment. Previous research has demonstrated that under obstacles to task completion, low-self-efficacy individuals tend to give up and/or exert less effort (Bandura, 1982, 1997; Bandura & Schunk, 1981; Nahl, 1996; Wood & Bandura, 1998). Specific to online information seeking, Nahl (1996) found that for a class with no previous Internet experience, students in the lower third percentile for self-efficacy dropped out of a Web-related course. Given such adversity, utilization of ICTs and its associated benefits may not be realized for those low in self-efficacy. Specific to the outcome of finding health-related information that is accountable, certain search tasks may present more adversity than others.

In the current study, the amount of time spent completing the task is a measure of perseverance, similar to the Nahl (1996) study. Based on previous findings on the effects of self-efficacy on perseverance, self-efficacy has a main effect on perseverance (total time spent on search). However, when self-efficacy is taken into account together with task specificity, differences in the total time spent for search tasks for high- and low-self-efficacy individuals should be more pronounced in the more difficult search task. As a result, those low in self-efficacy would exert less effort and end the search task sooner than their high self-efficacy counterparts. Thus, there should be a main effect of self-efficacy on performance (total time spent on search) and an interaction between self-efficacy and task specificity.

H3: High-self-efficacy participants will spend more time on the search task than low-self-efficacy participants.

H4: There will be an Internet self-efficacy by task specificity interaction, such that high-self-efficacy individuals will spend more time searching online than low-self-efficacy individuals in the general search task, but there will be no significant difference in the specific search task.

Method

  1. Top of page
  2. Abstract
  3. Introduction
  4. Method
  5. Results
  6. Discussion
  7. References
  8. Appendices

Design

This study is a mixed 2 (self-efficacy) X 2 (search task specificity) repeated-measures design. Internet self-efficacy is the between-subjects factor with two levels represented by high and low Internet self-efficacy. Task specificity is the within-subjects factor, measured by general search tasks and specific search tasks. The sequence of the two searches was randomly assigned to minimize order effects. The dependent variables are website accountability and total time spent per search (as a measure of perseverance). A control variable (Internet reliance) was implemented in the analysis to account for potential influence that Internet reliance may have on the ability to find credible information. Recent research on the predictors of perceived credibility has identified three audience factors associated with perceived credibility—media reliance, knowledge, and relevance (Metzger et al., 2003). In the web credibility literature, perhaps the most studied of these variables is media reliance, where the literature has found a strong relationship between medium reliance and medium credibility (Johnson & Kaye, 1998, 2000, 2002). Media reliance, or the degree to which an individual depends on a medium to achieve a specific gratification, is an audience factor that is predictive of perceived credibility.

Procedure

Prior to initiation of searches, participants filled out a scale to assess their Internet self-efficacy. Participants then physically searched the Web for health-related information. The participants were instructed to search through as many websites as necessary until they found a website containing information they would feel comfortable giving to a family member or friend who had requested their assistance, which is a fairly common occurrence (Pew, 2003b; Widman & Tong, 1997). The final website selected by a participant was the one used for content analysis.

Three desktop computers with high-speed Internet connections running the Netscape® Communicator browser were used for conducting Web searches. As participants surfed the Web, a noninvasive tracking software (NetSnitch™) captured in a log file the Web addresses of the sites visited and the time of the visits. Various precautions were taken to ensure that all participants had the same search conditions. These included having all participants begin their search sessions at the university’s home Web page, resetting the browsers to clear the “history,” and setting the browsers not to cache documents.

Participants

84 students (33 men and 52 women, mean age = 21.64 years) from a major university on the west coast of the United States volunteered to participate in the study. The gender proportion in this study is similar to the student population for the major. In exchange for their participation, students were given a gift certificate to the university bookstore. The majority of participants (65.5%) were of Caucasian descent. Asians constituted 16.7% of the sample, followed by Hispanics with 8.3% and African Americans with 3.6%. With respect to searching for health-related information, undergraduates are similar to the overall online population. Specifically, 75% of online youths (ages 15–24) have sought health-related information online (Kaiser Family Foundation, 2001), which is comparable to the 80% of American Web users who have used the Web to locate health-related information (Pew, 2003b). Internet users also tend to be highly educated, with 37% holding undergraduate or graduate degrees and another 34% having some college education (Pew, 2003a). In addition, women are more likely to be online health information consumers than men (Pew, 2003b), and Caucasians constitute 71% of the online population (Pew, 2003a).

Dependent Variables

The dependent variables are a within-subjects measure of the “website accountability” associated with the website that participants selected from their two health-related search tasks, and total time spent for each search. For website accountability, participants selected one site from each search task that they deemed to provide the best information. The selected websites were then independently content analyzed by two trained graduate students. Based on the AMA guidelines (Winker et al., 2000), the coding instrument assessed the presence/absence (1 = presence, 0 = absence) of the following elements of website accountability: information currency, content authorship, references of information, selection of information (e.g., editorial board), and privacy policy. Krippendorff's alpha (Krippendorff, 2003) was calculated for each of the six variables that comprised the Web site accountability measure. Krippendorff’s alpha ranged from .68 to 1.0, which is within the acceptable level of reliability recommended by Krippendorff (2003). The average Krippendorff alpha was .90. The items in the content analysis instrument were summed to create the variable website accountability for the general search (M = 1.7, SD = 1.05) and the specific search (M = 1.50, SD = 1.11). See Appendix A for items in the coding instrument.1

For the dependent variable, total time spent per search, the noninvasive tracking device provided the total time participants spent for the general search task (M = 6.75, SD = 3.19) and the specific search task (M = 6.51, SD = 3.09).

Independent Variables

Internet Self-Efficacy

Several scales have been developed to assess self-efficacy in the computer and Internet domains. Some of the computer self-efficacy scales were developed prior to the adoption of the Internet (e.g., Compeau & Higgins, 1995; Murphy, Coover, & Owen, 1989), and are thus not applicable to this study on self-efficacy outcomes for an information-seeking task. In the current study, items assessing Internet self-efficacy are from a previously validated scale (Eastin & LaRose, 2000). The scale consists of eight items that tap into beliefs regarding completion of general online tasks (Cronbach’s α= .91). Participants responded to the Internet self-efficacy scale prior to conducting their search tasks. In defining high and low Internet self-efficacy groups, participants were divided at the median (Median = 4.19, M= 4.20, SD= 1.22). This approach is commonly used in self-efficacy research in the ICT context (Nahl, 1996). See Appendix B for items in the Internet self-efficacy scale.

Task Specificity

There were two searches based on task specificity. The general search task asked participants to locate any tobacco cessation strategy, while the specific search task asked participants to locate a specific tobacco cessation method. The search topic of tobacco cessation strategies/products was chosen because of the health topic’s pertinence to the general public. No specific tobacco product was identified because the population of college students is known to use a variety of tobacco products other than cigarettes, including cigars, pipes, chewing tobacco, and snuff (Rigotti, Lee, & Wechsler, 2000). Prior to the general search task, participants were given a handout containing the following instructions:

A family member/friend of yours wants to quit smoking, but he doesn’t know what would be a good strategy. You want to help this family member by finding a good strategy on the Web. Search through as many Web sites as necessary until you have located the site with the information you feel you can give to this family member. Because this information is very important to this family member, you want information that is of high quality. When you have located this site, browse through it, print the Web site, and raise your hand to notify the research assistant.

Prior to the specific search task, participants were given a handout containing the following instructions that directed them to locate information on nicotine gum, the nicotine patch, nicotine nasal spray, or the nicotine inhaler:

The family member/friend who wants to quit smoking has recently heard that ________ is a good method for people who want to quit smoking, but he wants to find more information about it before trying it out. He wants to locate this information on the Web but he is unfamiliar with surfing the Web, and has asked you to help locate this information. Because this information is very important to this family member, he has asked that you find information that is of high quality. Search through as many Web sites as necessary until you have located the site with the information you feel you can give to this family member. When you have located this site, browse through it, print the Web site, and raise your hand to notify the research assistant.

To assess that the two tasks differed in specificity, a single 7-point Likert-like item assessed the difficulty of each search task. The difficulty of a search was assessed with the following statement: “I had a hard time finding this Web site.”

Control Variable

A control variable was implemented to account for potential influences on the outcome of website accountability. The extant credibility literature suggests that audience factors can influence credibility perceptions. Specifically, media reliance is an audience factor that is associated with perceived credibility (Metzger et al., 2003). Media reliance is the degree to which an individual depends on a medium to achieve a specific gratification. Although not related to health information, recent studies in the online context have found that reliance on the Web is significantly associated with credibility (Johnson & Kaye, 1998, 2000). Reliance (M = 4.65, SD = 1.75) was assessed by how likely participants would be to use the Web when they needed health information. The following statement was used, anchored by “strongly disagree” (1) and “strongly agree” (7): “When I need information on a health issue, I would go to the Internet,” which is modified from a previous study that examined media reliance for political information (Johnson & Kaye, 2000). While there are other potential control variables, including knowledge and relevance, only media reliance is used in the analysis, as it is perhaps the most studied of these variables in the web credibility literature.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Method
  5. Results
  6. Discussion
  7. References
  8. Appendices

Manipulation Check

Paired sample t-test for participant self-assessments of the relative difficulty of the respective searches indicated that the general search (M= 4.20, SD= .18) was significantly more difficult than the specific search (M= 3.15, SD= .16), t(83) = 8.12, p < .05.) This result suggests that the task specificity manipulation worked as intended.

Hypothesis 1

A 2 x 2 mixed ANCOVA was performed on the dependent variable website accountability (repeated measure). The between-subjects independent variable was Internet self-efficacy (high and low). The control variable was reliance on the Web for health-related information. Results meet the assumptions of sphericity and homogeneity of variance. The first hypothesis predicted that high Internet self-efficacy participants would find websites with higher website accountability than their low Internet self-efficacy counterparts. The main effects for Internet self-efficacy on website accountability was not significant, F(1,82) = 3.08, p= .08, ηp2= .04. Hypothesis 1 was thus not supported.

Hypothesis 2

The second hypothesis predicted an Internet self-efficacy and task specificity interaction, where high-self-efficacy participants would find a website in the general search task (the more difficult task) to have higher website accountability than their low-self-efficacy counterparts. There would be no significant difference in the specific search task (the less difficult task). Thus, H2 predicts a disordinal, nonsymmetrical interaction. The task specificity x Internet self-efficacy interaction was significant, F(1,82) = 4.0, p < .05, ηp2= .05, indicating that the change in website accountability for the high self-efficacy group was significantly different from the change in website accountability for the low self-efficacy group. Specifically, for the low self-efficacy group, the website accountability in the general search task (M = 1.38, SD = 1.1) is not significantly different from the website accountability in the specific search task (M = 1.52, SD = 1.1), t(41) = .658, p= .51. In contrast, the website accountability in the general search task (M= 2.02, SD= .89) is significant higher than the website accountability in the specific search task (M= 1.48, SD= 1.1) for the high-self-efficacy group, t(41) = 2.53, p < .05. Although the ANCOVA showed that the means were significantly different, the effect size was small to modest. See Figure 1 for the graphical depiction of this interaction effect.

image

Figure 1. Interaction between task specificity and Internet self-efficacy on website accountability.

Download figure to PowerPoint

Hypotheses 3 and 4

The third hypothesis predicted a main effect of self-efficacy on total time spent performing search. Main effects were found for self-efficacy on task performance, F(1,82) = 4.02, p < .05. High Internet self-efficacy participants (M= 7.15) spent more time searching for health-related information for both search tasks than their low Internet self-efficacy counterparts (M= 6.11). Hypothesis 3 was thus supported.

The fourth hypothesis predicted an interaction between self-efficacy and total time spent on search tasks. The total time spent on search x Internet self-efficacy interaction was significant, F(1,82) = 4.40, p < .05, ηp2= .05, indicating that the change in total time spent for search for the high-self-efficacy group was significantly different from the change in total time spent for search for the low-self-efficacy group. For the high-self-efficacy group, although the total time spent in the general search task (M= 7.51, SD= 3.49) is higher than the total time spent on the specific search task (M= 6.80, SD= 3.59) for the high-self-efficacy group, the difference was not significant, t(41) =−.94, p= .35. In contrast, for the low-self-efficacy group, the total time spent in the general search task (M = 5.52, SD = 2.26) is significantly less than the total time spent in the specific search task (M = 6.71, SD = 2.78), t(41) = 2.38, p < .05. Hypothesis 4 was thus supported. Although the ANCOVA showed that the means were significantly different, the effect size was small to modest. See Figure 2 for the graphical depiction of this interaction effect.

image

Figure 2. Interaction between task specificity and Internet self-efficacy on total time spent on search.

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Method
  5. Results
  6. Discussion
  7. References
  8. Appendices

The impetus for much of the research on the effects of self-efficacy in the ICT context has been the concern that people with low self-efficacy may not fully realize the personal benefits associated with utilization of ICTs. The current study is concerned with one particular self-efficacy outcome—finding accountable online health information. Specific to online health, the AMA has established information accountability guidelines for Web publishing that are similar to the message credibility attributes in the credibility literature. While previous research has demonstrated that self-efficacy influences the expected outcome of finding credible online information, this study examined the association between self-efficacy and an actual outcome rather than self-reported outcome expectations. The attainment of the outcome of finding credible information was measured by an independent content analysis of website accountability of the site participants located for two search tasks.

The results suggest that Internet self-efficacy influences the outcome of finding accountable online information, but it is contingent on search task specificity. Contrary to H1, which posited that those with high-self-efficacy would find websites with higher website accountability than their low-self-efficacy counterparts, no significant differences were found. When task difficulty was taken into account, there was a self-efficacy by task specificity interaction such that high-self-efficacy individuals will locate websites with higher website accountability than low-self-efficacy individuals in the more difficult search task. However, while high-self-efficacy participants found sites with higher Web site accountability than low-self-efficacy counterparts in the more challenging search (the general search task), the difference was not as pronounced as in the less difficult search task (the specific search task). This finding suggests that while both low- and high-self-efficacy individuals can locate health-related online information, this benefit of utilization of ICTs can be primarily realized for the context in which the search task is more challenging.

In the specific search task, which the manipulation check indicated to be less difficult than the general search task, participants had in mind exactly what they were searching for—a tobacco cessation product. However, in some health-related situations, people do not have a preconceived notion of the exact nature of the information they are searching for, such as in self-diagnosis situations. This scenario was created in the general search task when participants were asked to find any information that they thought would help a family/friend to quit tobacco consumption. Thus, while many self-efficacy studies in the ICTs context have found a direct relationship between self-efficacy and outcomes, these findings are often for one task, such that significant differences between low- and high-self-efficacy individuals may not be present in a task that requires less effort and skill. It is important to indicate that Bandura (2002) noted that for tasks that are too easy and require little skill, self-efficacy would not be a determinant of outcomes. In the current study, the specific search task was deemed by all participants to be moderately difficult.

Hypotheses 3 and 4 pertained to the performance of achieving the outcome of finding credible information online. While there are several performance variables associated with finding credible information online, this study was concerned with only one performance variable—the amount of time spent for each search. Previous research demonstrated that, under adversity to task completion, low-self-efficacy individuals tend to give up and/or exert less effort (Bandura & Schunk, 1981; Nahl, 1996). Thus, low-self-efficacy participants would spend less time searching than high-self-efficacy participants, as posited in H3. However, when task difficulty is taken into consideration, the performance differences would be more dramatic, as posited in H4. The main effect of self-efficacy on performance and the interaction between self-efficacy and task specificity on performance was found. Low-self-efficacy participants spent less time in the more challenging search task (the general search task) than the less challenging search task (the specific search task). It appears that for the more challenging task, low-self-efficacy individuals give up sooner than high-self-efficacy individuals, who spent more time in the more difficult search task than in the less challenging task.

The findings of this study are consistent with previous research on the effects of self-efficacy on ICTs outcomes. However, this study distinguishes itself from previous research in that it examines outcomes (rather than outcome expectations) and takes into account the influence of task difficulty. By doing so, the study can further define the online conditions in which self-efficacy plays a critical role in task success. It appears that high-self-efficacy individuals benefit more than low-self-efficacy individuals, but the benefits are more pronounced when the task is more complex.

A limitation of this study involves generalization. Generalizing the results to people who are confronted with medical decisions that may require further treatment or diagnostic information should be done only with caution. Although the literature suggests that searching for health information for a family member is a common practice, there may be differences between people searching for information for themselves and those searching for information for a family member. Thus, if the information is for individual use, participants may demonstrate more effort, which would be reflected in the quality of performance, which causally precedes outcomes. In addition, while this study utilized media reliance as a covariate, other covariates should be explored, including knowledge and relevance. Future research on self-efficacy outcomes in the ICT context should test other health-related online tasks. One task particularly pertinent to online health-seekers is medical diagnosis. Finally, other online self-efficacy outcomes related to health should be explored—including use of chat rooms and Internet community outlets to obtain information, comparison of website to other health-related websites, and consultation with physicians.

Note
  • 1

    The content analysis instrument includes AMA website accountability guidelines as well as other website quality markers. However, only those items that pertain to the AMA guidelines were used to create the website accountability indices.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Method
  5. Results
  6. Discussion
  7. References
  8. Appendices
  • Anselmo, M. A., Lash, K. M., Stieb, E. S., & Haver, K. E. (2004). Cystic fibrosis on the Internet: A survey of site adherence to AMA guidelines. Pediatrics, 114(1), 100103.
  • Aspden, P., & Katz, J. E. (2001). Assessments of quality of health care information and referrals to physicians: A national survey. In R. E.Rice & J. E.Katz (Eds.), The Internet and Health Communication: Experiences and Expectations (pp. 99106). Thousand Oaks, CA: Sage Publications, Inc.
  • Austin, E. W., & Dong, Q. (1994). Source v. content effects on judgments of news believability. Journalism Quarterly, 71(4), 973983.
  • Bandura, A. (1982). Self-efficacy mechanism in human agency. American Psychologist, 37(2), 122147.
  • Bandura, A. (1997). Self-Efficacy: The Exercise of Control. New York: W.H. Freeman.
  • Bandura, A. (2002). Growing primacy of human agency in adaptation and change in the electronic era. European Psychology, 7(1), 216.
  • Bandura, A., & Schunk, E. H. (1981). Cultivating competence, self-efficacy, and intrinsic interest through proximal self-motivation. Journal of Personality and Social Psychology, 41(3), 586598.
  • Berland, G. K., Elliott, M. N., Morales, L. S., Algazy, J. I., Kravitz, R. L., Broder, M. S., Kanouse, D. E., Muñoz, J. A., Puyol, J.-A., Lara, M., Watkins, K. E., Yang, H., & McGlynn, E. A. (2001). Health information on the Internet: Accessibility, quality, and readability in English and Spanish. The Journal of the American Medical Association, 285(20), 26122621.
  • Charles, C., Gafni, A., & Whelan, T. (1997). Shared decision-making in the medical encounter: What does it mean? Social Science and Medicine, 44(5), 681692.
  • Compeau, D., & Higgins, C. (1995). Computer self-efficacy: Development of a measure and initial test. MIS Quarterly, 19(2), 189211.
  • Drabenstott, K. M. (1984). Subject Searching in Library Catalogs: Before and After the Introduction of Online Catalogs. Dublin, OH: Online Computer Library Center.
  • Eagly, A., & Chaiken, S. (1993). The Psychology of Attitudes. Orlando, FL: Harcourt Brace.
  • Eastin, M. S., & LaRose, R. (2000). Internet self-efficacy and the psychology of the digital divide. Journal of Computer-Mediated Communication, 6(1). Retrieved January 24, 2006 from http://jcmc.indiana.edu/vol6/issue1/eastin.html
  • Eighmey, J., & McCord, L. (1998). Adding value in the information age: Uses and gratifications of sites on the World Wide Web. Journal of Business Research, 41(3), 187194.
  • Eysenbach, G., Powerll, J., Kuss, O., & Sa, E.-R. (2002). Empirical studies assessing the quality of health information for consumers on the World Wide Web. Journal of the American Medical Informatics Association, 287(20), 26912700.
  • Farley, D. (1995). FDA’s RX for better medication information. FDA Consumer, 29(9), 510.
  • Ford, N., Miller, D., & Moss, N. (2001). The role of individual differences in Internet searching: An empirical study. Journal of the American Society for Information Science and Technology, 52(12), 10491066.
  • Griffiths, K. M., & Christensen, H. (2000). Quality of web based information on treatment of depression: Cross sectional survey. British Medical Journal, 321(7275), 15111515.
  • Guadagnoli, E., & Ward, P. (1998). Patient participation in decision-making. Social Science Medicine, 47(3), 329339.
  • Hofstetter, C. R., Zuniga, S., & Dozier, D. M. (2001). Media self-efficacy: validation of a new concept. Mass Communication & Society, 4(1), 6176.
  • Ingwersen, P. (1992). Information retrieval interaction. London: Taylor Graham.
  • Johnson, T. J., & Kaye, B. K. (1998). Cruising is believing?: Comparing Internet and traditional sources of media credibility measures. Journalism & Mass Communication Quarterly, 75(2), 325340.
  • Johnson, T. J., & Kaye, B. K. (2000). Using is believing: The influence of reliance on the credibility of online political information among politically interested Internet users. Journalism & Mass Communication Quarterly, 77(4), 865879.
  • Johnson, T. J., & Kaye, B. K. (2002). Webelievability: A path model examining how convenience and reliance predict online credibility. Journalism and Mass Communication Quarterly, 79(3), 619642.
  • Kaiser Family Foundation. (2001). Generationrx.com: How young people use the internet for health information. Menlo Park, CA: Kaiser Family Foundation.
  • Kim, K., & Allen, B. (2002). Cognitive and task influences on Web searching behavior. Journal of the American Society for Information Science and Technology, 53(2), 109119.
  • Krippendorff, K. (2003). Content analysis: An introduction to its methodology (2nd ed.). Beverly Hills: Sage.
  • Kunst, H., Groot, D., Latthe, P. M., Latthe, M., & Khan, K. S. (2002). Accuracy of information on apparently credible websites: survey of five common health topics. British Medical Journal, 324(7337), 581582.
  • Marchionini, G. (1989). Information seeking strategies of novices using a full-text electronic encyclopedia. Journal of the American Society for Information Science, 40(1), 5466.
  • Marchionini, G. (1995). Information seeking in electronic environments. Cambridge: Cambridge University Press.
  • Meric, F., Bernstam, E. V., Mirza, N. Q., Hung, K. K., Ames, F. C., Ross, M. I., Kuerer, H. M., Pollock, R. E., Musen, M. A., & Singletary, S. E. (2002). Breast cancer on the world wide web: Cross sectional survey of quality of information and popularity of websites. BMJ, 324(7337), 324577.
  • Metzger, M. J., Flanagin, A. J., Eyal, K., Lemus, D. R., & McCann, R. M. (2003). Credibility in the 21st century: Integrating perspectives on source, message, and media credibility in the contemporary media environment. In P.Kalbfleisch (Ed.), Communication yearbook (Vol. 27, pp. 293335). Newbury Park, CA: Sage Publications.
  • Murphy, C., Coover, D., & Owen, S. (1989). Development and validation of the computer self-efficacy scale. Educational and Psychological Measurement, 49(4), 893899.
  • Nahl, D. (1996). Affective monitoring of Internet learners: Perceived self-efficacy and success. Proceedings of the 59th ASIS Annual Meeting, 33, 100109.
  • O’Keefe, D. J. (2002). Persuasion: Theory and research (2nd ed.). Thousand Oaks, CA: Sage Publications.
  • Pew. (2002). Vital decisions: How internet users decide what information to trust when they or their loved ones are sick. Washington, DC: Pew Internet & American Life Project.
  • Pew. (2003a). The ever-shifting internet population. Washington, DC: Pew Internet & American Life Project.
  • Pew. (2003b). Internet health resources. Washington, DC: Pew Internet & American Life Project.
  • Qiu, L. (1993). Analytical searching vs. browsing in hypertext information retrieval systems. Canadian Journal of Information and Library Science, 18(4), 113.
  • Rice, R. E. (2001). The Internet and health communication: A framework of experiences. In R. E.Rice & J. E.Katz (Eds.), The internet and health communication (pp. 4773). Thousand Oaks, CA: Sage.
  • Rigotti, N. A., Lee, J. E., & Wechsler, H. (2000). US college students’ use of tobacco products. Journal of the American Medical Association, 284(6), 699705.
  • Saracevic, T., & Kantor, P. (1988). A study of information seeking and retrieving: Users, questions, and effectiveness. Journal of the American Society for Information Science, 39(3), 177196.
  • Shon, J., Marshall, J., & Musen, M. A. (2000). The impact of displayed awards on the credibility and retention of Web site information. Journal of the American Medical Informatics Association, 794798.
  • Silberg, W. M., Lundberg, G. D., & Musacchio, R. A. (1997). Assessing, controlling, and assuring the quality of medical information on the Internet: Caveant lector et viewor—Let the reader and viewer beware. Journal of the American Medical Association, 277(15), 12441245.
  • Slater, M. D., & Rouner, D. (1997). How message evaluation and source attributes may influence credibility assessment and belief change. Journalism and Mass Communication Quarterly, 73(4), 974991.
  • Stieb, E. S., Wang, J., & Haver, K. E. (2002). Do food allergy websites satisfy AMA guidelines for health information on the Internet. The Journal of Allergy and Clinical Immunology, 109(1), s309.
  • Thompson, L. F., Meriac, J. P., & Cope, J. G. (2002). Motivating online performance. Social Science Computer Review, 20(2), 149160.
  • Tsai, M., & Tsai, C. (2003). Information searching strategies in Web-based science learning: The role of Internet self-efficacy. Innovations in Education and Teaching International, 40(1), 4350.
  • Wathen, C. N., & Burkell, J. (2002). Believe it or not: Factors influencing credibility on the Web. Journal of the American Society for Information Science and Technology, 53(2), 134144.
  • Widman, L. E., & Tong, D. A. (1997). Requests for medical advice from patients and families to health care providers who publish on the World Wide Web. Archives of Internal Medicine, 157(2), 209212.
  • Winker, M. A., Flanagin, A., Chi-Lum, B., White, J., Andrews, K., Kennett, R. L., DeAngelis, C. D., & Musacchio, R. A. (2000). Guidelines for medical and health information sites on the Internet: Principles governing AMA Web sites. The Journal of the American Medical Association, 283(12), 16001606.
  • Wood, R., & Bandura, A. (1998). Social Cognitive Theory of organizational management. Academy of Management Review, 14(3), 361384.
About the Author
  1. Traci Hong is an Interdisciplinary Women’s Health Research (IWHR) Faculty Scholar at the Department of Community Health Sciences, Tulane University School of Public Health and Tropical Medicine. Her research is on women’s health, cardiovascular diseases, and the uses of technologies in health contexts.

    Address: Tulane School of Public Health and Tropical Medicine, 1440 Canal St., Ste. 2301, New Orleans, LA 70112 USA

Appendices

  1. Top of page
  2. Abstract
  3. Introduction
  4. Method
  5. Results
  6. Discussion
  7. References
  8. Appendices

Appendix A

Content Analysis for Structural and Message Features

1) Site Domain Name  .GOV .ORG .EDU .COM

Is the domain name “.gov” (e.g., www.cdc.gov), “.org” (ama.org) or “.edu” (www.iub.edu)?

Site Ownership

2) Site Authorship  YES  NO

Does the name of the organization/person running the site appear on the site? This is usually found on the top or the bottom of a Web page.

3) Site Contact Information  YES  NO

Is there an email, telephone, or mailing address by which site visitors can contact someone affiliated with the site?

Advertisement

4) Sponsorship  YES  NO

Does the name of the organization/person “sponsoring” the site appear on the site? Other key words to denote sponsorship include“brought to you by, “sponsored by,”“sponsor,”“Courtesy of.”If any of these key words appear on the site, check yes; If none of these terms appear, check no.

5) Advertisement  YES  NO

Is there at least one advertisement on the Web site? This can be a banner ad which appears either on the top or the bottom of the page or any graphic image. Alternatively, it can also be a hyperlink with text suggesting the sale of a product.

An advertisement would be a product or service provided in exchange for money. The product/service should NOT be directly associated with the site. For example, a banner ad on Nicorette on the Nicorette site would NOT be considered an advertisement. Advertisements should be distinct from site sponsorship. That is, if the site sponsor has a banner ad, do not include it in this variable.

Information

For the following questions in this section (“information”), carefully read through all the text that provides information on a health-related topic.

6) Information Currency  YES  NO

Is there a date that indicates when the information was last updated? This usually appears at the bottom of the page.

7) Information Authorship  YES  NO

Is there an author attributed to the information provided. This can be a byline for an article, opinion, or a “written by.”

8) Information Reference  YES  NO

Is there at least one citation of reference for the information provided? When a scientific/medical claim is made, is it attributed to a source which is referenced?

9) Selection of Information  YES  NO

Is there evidence that a medical board or editorial board decides what information is posted on the site? This can be a statement stating such, or a link to “medical board” or “editorial board.”

10) Testimonials  YES  NO

Is there a story or account given by an individual (either anonymous or specified by name) regarding the health-related topic? A quote in the main text of the site does not constitute a testimonial. A quote that appears alone (i.e., not buried in the main text) does count as a testimonial.

11) Quotations  YES  NO

Does a quote regarding the health-related topic that is attributed to an individual appear in the main text of the site? The individual may be anonymous. Quotes are text found between “text text.”

12) Statistics  YES  NO

Are statistics regarding the health-related topic used on this site? This can be percent, ratios, etc. However, it is NOT $.

Third Party Endorsements

13) HON Network  YES  NO

Go to http://www.hon.ch/HONcode/Conduct.html and enter the URL of the site into the HON database. This will indicate whether the site has been accredited by the HON Network (Health on the Internet Network).

14) Other Third-Party endorsements  YES  NO

Is there an emblem/seal, notice, or other indication that the site is endorsed by a third-party organization that attests for the integrity of the site? Examples of third-party endorsements: TRUSTe, VIPPS, BBB, or related organizations.

15) Privacy Notice  YES  NO

Is there a privacy policy statement or a link to one? A privacy policy statement is a comprehensive description of a site’s practices regarding information collected. It can appear under various labels including “privacy statement,” privacy policy,”“privacy,”“security,”“online privacy practices,”“our policies,” or other similar labels.

Site Features

16) Navigation  YES  NO

Is there a Menu or list that serves as a directory for the Web site?

Appendix B

Internet Self-Efficacy Scale from Eastin & Larose (2000)

I feel confident . . .

understanding terms/words relating to Internet hardware.

understanding terms/words relating to Internet software.

describing functions of Internet hardware.

troubleshooting Internet problems.

explaining why a task will not run on the Internet.

using the Internet to gather data.

learning advanced skills within a specific Internet program.

turning to an online discussion group when help is needed.