The vast quantities of online information have transcended some barriers to information, such as time and geography, to provide people with relevant, timely information that may increase their health and well-being [1]. The extent to which individuals are able to take advantage of available information and use the Internet to acquire benefits and opportunities [2-3] depends on their ability to navigate the online environment and requires diverse forms of online engagement [4-5]. Those who are unable to fully engage with these technologies are at a disadvantage to reap benefits that the online world may offer [1]. Digital inequalities describe the differing levels of ability to engage with the Internet among those with access [6], with research often emphasizing the differences in how individuals from lower socioeconomic position (SEP) incorporate the Internet into their lives compared to those with higher SEP [7-10]. However, focusing on a deficits approach in which low SEP individuals are compared to those with greater means may overshadow the myriad ways low SEP individuals benefit from the Internet. This study, from a randomized controlled trial involving low SEP individuals, used innovative, real-time Web tracking data to contextualize more fully the Internet information seeking of low SEP individuals.

Health information seeking may occur most often when need for a specific disease or medical decision making arises [1,11-14], but may not be indicative of everyday Internet use. Indeed, health information seeking serves as an important marker of disparities, with widespread evidence indicating that individuals from low SEP seek online health information less than their high SEP counterparts [11,15-17]. However, this seeking behavior must be placed in context of the entire spectrum of Internet engagement to fully understand the impact of digital inequalities.

Explorations of the ways low SEP individuals use the Internet, including the range and breadth of activities performed on the Internet for a variety of functions [3], may more fully contextualize the online environment for underserved groups. Beyond accessing health information, there is a need to participate in Internet-mediated economic, political, and social networks to take full advantage of what the online world has to offer and to meaningfully participate in an increasingly digital society [2,5,18]. “Capital-enhancing” (p. 606 [19]) information seeking represents exploring websites and topics that may enhance one’s life chances and offer upward mobility, such as helping with career advancement or consulting financial services, and serve to improve one’s life circumstances by increasing access to material and informational resources [20]. The inclusion of a range of topics that may impact health and well-being reflects a social determinants perspective, in which access to resources such as wealth, education, and adequate housing may impact health outcomes [21-22]. Specifically, capital-enhancing Internet use may be defined in this context as using the Internet to increase access to tangible and information-based resources that can address structural-level factors such as housing, education, employment, finances, and access to government resources.

Despite the potential benefits of such capital-enhancing information, research indicates that individuals with lower SEP take fewer opportunities to use the Internet comprehensively beyond amusement and communication, suggesting that the underserved may not fully take part in the new media environment [3]. Studies have found that, compared to individuals with higher income, low SEP individuals are less likely to use the Internet for capital-enhancing activities [19,23,24], instead using the Internet predominantly for entertainment purposes [20,23,25,26]. Such a “usage gap” between information and entertainment uses of the Internet may serve as an indicator of disparities, identifying who is not able to fully participate in and reap the benefits of the online world [18,19,23,26].

However, many studies reporting such a usage gap rely on self-report Internet use information [3,18,20,23,24], which is subject to the limitations of self-report data, such as bias and inaccuracy, particularly over longer periods of recall [27]. Relying on dichotomies of use (eg, never/ever use) of selected website categories from self-reported data may overgeneralize findings and may not accurately reflect or contextualize Internet usage in low SEP groups [3].

Additionally, providing a broad comparison between the Internet usage patterns of different income strata do not account for the differences in how low SEP individuals may engage with and learn from the Web in unique ways from their high SEP counterparts. Beyond the scope of entertainment versus information, scholars have suggested that a broader and more sophisticated use of the Internet, particularly engaging with the Web for diverse purposes, allows an individual a greater opportunity to acquire benefits and opportunities to meet individual and social goals [3]. The range and breadth of activities performed on the Internet, such as visiting a number of different types of sites that perform a variety of functions [3], may indicate that an individual is better situated to participate in a society that has transitioned many services online [4]. Use of the Internet for functions such as shopping, social network use, blogs, email, and reference sites are rarely captured in Internet studies and are even rarer if not nonexistent for low SEP individuals. Including an assessment of these activities from observed (tracked) Internet activity may provide useful context to Internet activities of interest, such as health information seeking, and if other types of Internet use serve to complement or substitute health seeking. Furthermore, they may indicate that the individual is more adept at using the Internet for a number of diverse functions.

Determining the detailed usage patterns of low SEP individuals may highlight the best ways to engage them in online activities that provide them with resources to improve their health or socioeconomic position.

The purpose of this paper is to build on our prior work [28] using a unique dataset containing directly tracked Internet use from Click to Connect (C2C), a randomized controlled trial that provided first-time at-home Internet access to the urban poor to explore their Internet behaviors. We were able to directly capture the Internet sites that participants visited to provide a detailed look at the Internet seeking behaviors of low SEP individuals. The nature of this study allowed us to capture all instances of information seeking in a natural setting. Previously, we examined the Internet browsing patterns of low SEP individuals and discovered that although Internet portals and social networking sites were the most used, the heterogeneity of the websites visited suggested the participants were using the Internet for a diverse range of functions [28]. In this study, we sought to provide a detailed description of the diverse ways that low SEP individuals use the Internet, including the range of functions individuals perform online, information seeking for specific beneficial functions, and entertainment use. Then, we explored how this use was associated with health information seeking within the context of everyday, home-based computer use in this low SEP sample.

Data for this study were drawn from “Click to Connect: Improving Health Literacy Through Computer Literacy” (C2C), a randomized controlled trial funded through the National Cancer Institute to understand computer- and Internet-related challenges, barriers, and facilitators among a low SEP population. Intervention details may be found elsewhere (see [28,29]). Recruitment efforts were designed to recruit low-income, urban poor who are seldom represented in national surveys [29]. For the purposes of this study, we included data from the intervention group participants of all three waves of implementation who were given computers and Internet access during the course of the trial. Human participant approval for this study was granted by the Institutional Review Board (IRB) for the Dana-Farber Cancer Institute.

The demographic characteristics of the sample can be found in Table 1, which illustrates the focus of C2C on low SEP groups compared to national communication surveys. The majority of the sample was black (55.1%, 65/118), with 80 of 118 (67.8%) older than 35 years, and 73 of 118 (61.8%) female. The sample was low income, with 33.1% (39/118) reporting a household income of less than US $10,000 per year and 90 of 118 (76.2%) reporting less than a high school degree.

This study represents in-depth research of natural online behaviors of low SEP individuals over a period of several months that draws from directly tracked Internet data. Through this method, we were able to place health information seeking, capital-enhancing information seeking, entertainment use, social network use, and other diverse forms of Internet use within the context of the total Web use experience of low SEP individuals, data that are not often captured in such detail for this group. Participants sought information on a number of domains; 64% visited all the categories of the Internet use index over the study period and each additional category of computer use was associated with double the rate of health information seeking. Higher use of all individual types of capital-enhancing seeking was associated with increased rates of Internet health information seeking, with the highest increases seen in education and governmental website use. When all capital-enhancing categories were combined, the highest users of capital information sought health information at 13 times the rate of low users. Furthermore, both medium and high entertainment users were significantly associated with higher health information seeking compared to those who used entertainment sources to a low extent, and high social network site use was associated with higher health information seeking compared to low use.

Broader use of the Internet may provide individuals with skills to become more active online consumers [3]. Indeed, once online, low SEP individuals in the current study used the Internet for a diverse range of functions, from information gathering to sharing data to using search engines to springboard to other types of information. Importantly, this increased use was associated with increases in health information seeking. Although this behavior may present differently from that of a high SEP seeker, it points to the diverse ways in which low SEP individuals use the Web to find numerous types of content, including health information. This also indicates that even those who have not had much prior Internet use experience are able to quickly acclimate to the Web and use it for diverse functions beyond just entertainment once they are provided with access.

Evidence shows that health outcomes are patterned by access to material resources, education, and occupation [22]. By engaging in capital-enhancing information seeking, participants may gain an increase in resources that improves their SEP and subsequently improves their health. Thus, capital-enhancing functions may be important to consider when understanding how low SEP groups are using the Internet. Although all capital categories were significantly associated with higher Internet health information seeking, the notable rates of seeking associated with news, government, and education uses of the Internet may highlight the importance of these functions working to complement one another for the goal of greater well-being. For example, the literature has documented how online governmental information seeking and using the Internet as a civic resource and forum strongly influences civic engagement [36]. It may be that those who are more likely to research current events, governmental and political information, and engage with their communities may be more prone to research information pertaining to other avenues to well-being; participants may also visit government sites to access services such as food stamps or housing assistance.

In all, these associations suggest that higher levels of Internet use for functions such as searching for a job, financial resources, or educational programs correspond with higher levels of searching for health information. Past literature has found that in certain contexts, financial information served as a competing concern to health information for low SEP individuals [37]. In contrast, this study supports a conceptualization of complementation of seeking for types of beneficial sites instead of substitution. In other words, instead of the search for money or adequate housing replacing health information seeking due to competing needs, different types of capital-enhancing information may work together to provide a more comprehensive set of resources to increase well-being.

Although participants visited entertainment websites more frequently than health or capital websites reflecting past literature, we also observed entertainment use of the Internet was positively associated with greater health information seeking. This is an important distinction; instead of entertainment use taking precedence over higher-order activities, individuals who spend more time online may do so in varied topic areas as they gain more confidence in using the Web [18] and as complex skills required for Web 2.0 uses develop [38]. This may also indicate that once individuals gain more exposure to the Internet, their perceived utility of the Web for different functions increases. Such findings are an important distinction, particularly in a media environment that often portrays low SEP individuals as using entertainment content to the extent that it overshadows other forms of Internet use. For example, a recent New York Times article claimed that increased access to technology creates a “time-wasting gap” in which the entertainment uses of the Internet by low SEP individuals eclipse its use for positive, educational uses, particularly for youth [39]. In contrast, our study illustrates that Internet behaviors must be placed in context of the larger seeking environment; although entertainment content was highly accessed, high entertainment users were also more likely to use the Internet to gather health information.

Although both health and capital information seeking represented only a small total of all websites visited compared to categories such as social networking, Internet portals, and entertainment, it may be indicative of the nature of the sites’ structure. For example, sites with constantly changing content, such as celebrity gossip, and particularly user-generated content, such as social network sites, may require more frequent interaction to remain up-to-date with activity. However, static sites may only require one visit to gather needed information, such as referencing a health diagnosis. Other sites, such as job sites, may be visited only as a certain need arises. The frequency with which these dynamic sites, such as social network sites, are accessed provide a promising platform for future eHealth content delivery [28]. This may point to more diverse use of social network sites for patient support and delivery of important health information, particularly because they are already a familiar platform for many Internet users. Social network sites also represent an important source of information exchange and a growing number of Internet users are turning to these sites to post about their own health activities, follow friends’ health experiences, or find health information [40].