Health literacy and the internet: An exploratory study on the 2013 HINTS survey

Computers in Human Behavior 58 (2016) 240e248

Contents lists available at ScienceDirect

Computers in Human Behavior journal homepage: www.elsevier.com/locate/comphumbeh

Full length article

Health literacy and the internet: An exploratory study on the 2013 HINTS survey Shaohai Jiang*, Christopher E. Beaudoin Department of Communication, Texas A&M University, 102 Bolton Hall, MS 4234 TAMU College Station, TX 77843-4234, USA

a r t i c l e i n f o

a b s t r a c t

Article history: Received 20 October 2015 Received in revised form 2 January 2016 Accepted 6 January 2016 Available online xxx

Health literacy rates among American and European adults remain low, with almost half of adults having only basic levels in 2012. In this digital era, the Internet has been recognized as an important medium for improving health literacy. However, little is known about the mechanisms that underlie its impact on health literacy. With a general basis in the Cognitive Mediation Model, this study empirically tested a model that included motivation for health-related Internet use, health-related Internet use, perceived health information overload, and health literacy. Structural equation modeling was used to analyze the US-based Health Information National Trends Survey (HINTS) 2013 dataset. The results support for all the paths in our posited model. The effects of motivation for health-related Internet use on health literacy were completely mediated by health-related Internet use and perceived health information overload. The findings extend the Cognitive Mediation Model to the context of health literacy and provide significant implications for the design and dissemination of online health information. Recommendations are made for future research, including further validation of the five-item scale of health literacy. © 2016 Elsevier Ltd. All rights reserved.

Keywords: Motivation Internet use Information overload Health literacy Cognitive mediation model

1. Introduction Low health literacy rates pose a significant problem for population health. In the United States, levels are low, with nearly half of American adults having only basic health literacy in 2012 (Champlin & Mackert, 2015) and 20% of Medicare patients having low health literacy and 29% marginal health literacy (Mitchell, Sadikova, Jack, & Paasche-Orlow, 2012). Similar levels have been found in Europe. According to the European Health Literacy Survey, almost 1 in 2 (47%) adults have insufficient or problematic health literacy (Sørensen et al., 2015). The Internet may pose a formidable communication channel for helping advance health literacy in the United States. During the past decade, the Internet has become an important medium in terms of population health, with its strengths well documented, including availability of a wide range of information, reduced cost, increased access, ability to overcome time and space, realetime interaction, tailoring of content, and anonymity (Murero & Rice, 2006; Rice, 2006). The progressive development of these advantages is congruent with the rapid growth in Internet use for health

* Corresponding author. E-mail addresses: [email protected] (S. Jiang), [email protected] (C.E. Beaudoin). http://dx.doi.org/10.1016/j.chb.2016.01.007 0747-5632/© 2016 Elsevier Ltd. All rights reserved.

purposes, with 78.8% of U.S. citizens reporting having searched for health information on the Internet (National Cancer Institute, 2014). The positive outcomes of such health-related Internet use have been borne out specific to healthy literacy, including how people's online health information use can spur improvements in their medical knowledge, sense of patient empowerment, selfpresentation during medical encounters, and self-management skills (Lomanowska & Guitton, 2014; Mano, 2014; Murero & Rice, 2006; Rice, 2006). Despite this documentation of the benefits that Internet use has for health literacy, empirical results have, at times, been inconsistent. For example, a recent study investigated the associations between health literacy, health information access and Internet usage among patients in private and public health clinics, finding a nonsignificant relationship between Internet use and health literacy (Gutierrez, Kindratt, Pagels, Foster, & Gimpel, 2014). Even more concerning, Schulz and Nakamoto (2011) contended that Internet use can raise problems with health literacy given that inaccurate online information can result in hasty, ill-informed, and dangerous health decision making. Thus, what type of health information is available online and whether such material is adapted to a target population are critical issues in health education and promotion (Guitton, 2015). Furthermore, Josefsson (2006) indicated that different modes of information seeking behavior can lead to

S. Jiang, C.E. Beaudoin / Computers in Human Behavior 58 (2016) 240e248

dissimilar health consequences, with active search, as compared to passive attention, exerting stronger impact on a health information seeker's health literacy. Considering the inconsistent results on the degree to which Internet use can influence health literacy, there is a need for additional research in this area. This need is further underscored by the lack of research on the social mechanisms that underlie this relationship. Given these two gaps in the literature, as well as continued problematic levels of health literacy in the United States, the present study investigates the process by which Internet use facilitates the development of health literacy. Inspired by Uses and Gratifications theory (Katz, Blumler, & Gurevitch, 1973) and, more specifically, in the Cognitive Mediation Model and derivative models (Beaudoin, 2008; Beaudoin & Thorson, 2004; Eveland, 2002; Eveland & Dunwoody, 2001; Eveland, Shah, & Kwak, 2003; Ho, Peh, & Soh, 2013; Jensen, 2011; Lo, Wei, & Su, 2013), this study postulates the staged roles of motivation for health-related Internet use, health-related Internet use and perceived health information overload in predicting health literacy. 2. Conceptual framework This study articulates a multi-step model for the development of health literacy (see Fig. 1). This conceptual model includes six integral paths: 1) motivation for health-related Internet use to health-related Internet use; 2) motivation for health-related Internet use to perceived health information overload; 3) healthrelated Internet use to perceived health information overload; 4) motivation for health-related Internet use to health literacy; 5) health-related Internet use to health literacy; and 6) perceived health information overload to health literacy. The model draws generally from the Cognitive Mediation Model (CMM) (Eveland & Dunwoody, 2001), which posits that various motivations drive people to pay attention to news media and proactively process news information, which, in turn, influences their knowledge development. However, motivation is not expected to influence knowledge acquired directly and, instead, has indirect effects as mediated by news attention and news elaborative processing. CMM has been tested in different contexts with different gratifications sought and with cross-sectional and panel data (Beaudoin & Thorson, 2004; Eveland, 2001, 2002; Eveland et al., 2003). Moreover, with a general basis in CMM, Beaudoin (2008) proposed a framework that shares even greater

241

commonality with the current model. That study tested a derivative of CMM that included four steps: 1) social resource motivation for Internet use; 2) Internet use; 3) perceived information overload; and 4) interpersonal trust. This model, thus, shares similarity with our current model in terms of steps 1, 2 and 3. In CMM, the influence of motivation can be understood under the umbrella of Uses and Gratifications theory (U&G). U&G assumes that people are actively involved in evaluating the potential benefits of media use. They are goal-oriented and purposefully pay attention to media content that can satisfy their needs. Therefore, people's motivation plays a key role in explaining their subsequent media use, with their choice of a medium based in their sought motives (Blumler, 1979). In CMM, Eveland (2001) expanded upon this linkage between motivation and media use to include subsequent information processing and the outcome of public affairs knowledge. Thus, CMM is different from prior U&G research in that it considers cognition and final effects on knowledge development. In addition, unlike U&G, CMM posits that motivation itself does not have a direct role in media effects and, instead, activates information processing. Thus, based on U&G and CMM, and particularly guided by Beaudoin's (2008) derivative model, this study proposes a conceptual framework in which health-related Internet use and perceived health information overload mediate the effects of Internet use motivation on health literacy. 2.1. Path 1: motivation for health-related internet use to healthrelated internet use U&G theory posits that audiences are active and goal-directed. Central to this approach is the concept of motivation (Blumler, 1979). Motivation can be conceived as an impetus of action (Deci & Ryan, 1985), or a point of common ground between needs, cognitions and emotions (Reeve, 1997). Previous studies have widely demonstrated that motivation for media use is positively associated with different types of media use, including via magazines (Payne, 1988), radio (Armstrong & Rubin, 1989), and TV (Rubin, 1983). With the advent of the Internet, motivation has also been significantly correlated with email use (Cho, De Zuniga, Rojas, & Shah, 2003), instant messaging (Leung, 2001), blog use (Chung & Kim, 2008), and social media use (Park, Kee, & Valenzuela, 2009). In regards to health-related Internet use, motivation to obtain health information has been considered as a strong predictor of some Internet

Fig. 1. Conceptual framework on development of health literacy.

242

S. Jiang, C.E. Beaudoin / Computers in Human Behavior 58 (2016) 240e248

activities, including adopting social networking sites to improve quality of life (Omar, Rashid, & Majid, 2014) and utilizing online care pages to gain social support (Anderson, 2011). Specific to CMM, Eveland and colleagues (2002; 2003) operationally defined motivation in terms of surveillance, which entails how individuals use news media to learn about their social environment. In a follow-up study, Beaudoin and Thorson (2004) added two other dimensions of motivation, namely anticipated interaction and guidance. More recently, in a CMM study of knowledge about the H1N1 pandemic, Ho et al. (2013) operationalized motivation in terms of surveillance and guidance gratifications, as well as need for cognition. CMM research has found strong empirical support for this linkage from motivation to media use, including from surveillance gratification to news attention (Eveland, 2002; Eveland & Dunwoody, 2001; Eveland et al., 2003; Ho et al., 2013), guidance gratification to news attention (Ho et al., 2013), and guidance, surveillance, and anticipated interaction gratifications to news reliance (Beaudoin & Thorson, 2004). In addition, in his derivative model, Beaudoin (2008) documented the significant path from social resource motivation to Internet use. The current study relies on a different approach to operationally defining motivation for media use. Instead of tapping a person's level of intention to use the media for a specific purpose such as surveillance, the current study asks about a person's preferred communication medium for seeking out health and medical information in a hypothetical situation of strong need. Consistent with some prior research on information gratifications (Stafford, Stafford, & Schkade, 2004), this measurement approach generally centers on surveillance motivation, entailing respondents' need for relevant information and content. Despite this difference in measurement, our expectation is consistent with prior research on U&G and CMM. H1: Motivation for health-related Internet use is positively associated with health-related Internet use.

2.2. Path 2: motivation for health-related internet use to perceived health information overload Previous research has postulated that motivation for media use is positively related with information elaboration. Related CMM research has documented positive relationships between assorted motivation dimensions (e.g., surveillance gratification, anticipated interaction, guidance) and information processing (Beaudoin, 2008; Eveland, 2001; Eveland et al., 2003; Jensen, 2011). For example, Eveland (2002; 2003) found that surveillance gratifications predicted news elaboration. Moreover, Ho et al. (2013) documented that motivation for seeking H1N1-related news was a strong predictor of elaboration. In terms of information processing, we focus on the concept of information overload, which represents “a state of affairs where an individual's efficiency in using information in their work is hampered by the amount of relevant, and potentially useful, information available to them” (Bawden, Holtham, & Courtney, 1999). If an idea is overly complicated, it is difficult for people to determine what information is relevant and useful and, thus, will be less likely to learn and adopt a recommended behavior. As Petty and Cacioppo (1986) found, motivation can be helpful in such situations in helping reduce information overload. When individuals are motivated, they engage in issue-relevant thinking. When this occurs, people access relevant images and experiences from memory and elaborate on the new information in conjunction with the memory. Through this process, motivation can help individuals identify relevant information and reduce information overload. In

his CMM derivative, Beaudoin (2008) found support for this logic, documenting that social resource motivation for Internet use reduced information overload. We hypothesize similarly in terms of the path from motivation for health-related Internet use to perceived health information overload. H2: Motivation for health-related Internet use is inversely associated with perceived health information overload.

2.3. Path 3: health-related internet use to perceived health information overload Information overload can result from people's continued efforts in information search. In the limited capacity model, Lang (2000) contended that, when exposed to excessive media content, message processors have limited capacity, in terms of recognition memory, resource allocation, orienting behavior, and reaction time. In other words, people have difficulty in processing too many media messages they receive because the processing of mediated content requires people to devote cognitive effort constantly, which is a finite capacity in everyone (Ji, Ha, & Sypher, 2014). The Internet provides individuals with easy access to large amounts of information. Due to their limited cognitive ability, Internet users may find health information functionally inaccessible, increasing their information overload. Internet users usually encounter an incredible amount of irrelevant information when using online search engines given that web search engines rank their search results by link popularity or other algorithms rather than by relevance, which can result in difficulty in finding the most desirable results (Eysenbach & Kohler, 2002). For example, a national survey indicated that 66% of U.S. adults perceived that there is too much information about the obesity issue and concerns on the Internet, and it is hard to know what weight one should maintain to be healthy (Chan & Huang, 2013). In addition, social media have recently become an important alternative for seeking health information. The ease of online communication results in an enormous augmentation of postings and tweets on social networking sites (Soucek & Moser, 2010). Faced with a large volume of social media content, people can conduct keyword searches of this context, but the results of social media searches, much like those of via web search engines, are filtered by posting popularity (e.g., like, retweet, comment), which may derive results that are not highly relevant and can, thus, increase information overload (De la Torre-Díez,
n-Rodríguez, 2012). Díaz-Pernas, & Anto In addition to the quantitative issue of information overload, online health information is often deficient in its quality. Health information found from the Internet is vast and often replete with medical jargon or technical wording, with some online health sources lacking credibility, all of which can prevent people from obtaining satisfactory answers to their health-related questions and can increase their perceptions of information overload (Cline & Haynes, 2001; Rice, 2006). Considering such barriers in online health information seeking, we hypothesize that Internet use for health purposes would be associated with increased perceived health information overload. H3: Health-related Internet use is positively associated with perceived health information overload.

2.4. Path 4: health-related internet use to health literacy Prior research has widely documented the significant relationship between media use and health literacy (Ghaddar, Valerio,

S. Jiang, C.E. Beaudoin / Computers in Human Behavior 58 (2016) 240e248

Garcia, & Hansen, 2012; Meppelink & Bol, 2015). The Internet offers many opportunities for people to consume information and build health and medical knowledge. Underscoring the Internet's provision of health information to people, about 60% of all U.S. adults in 2010 reported having searched for health-related topics on the Internet, with about 20% doing so regularly (Fox, 2011). Furthermore, in 2011 and 2013, 16.8% and 23.2% of U.S. adults, respectively, reported having read and shared health and medical information on social networking sites (National Cancer Institute, 2014). In addition, 52% of U.S. cell phone owners in 2012 reported having looked up health information on their phone, which marked a dramatic increase from 17% in 2010 (Fox & Duggan, 2012). Such Internet use can help improve people's health literacy, which is defined as a person's ability to obtain, interpret and understand basic health information and services, and use such information to improve health (USDHHS, 2000). For example, one study showed that patients who regularly browsed health information from an Internet-based patient portal had 1.4 times higher health literacy than non-users (Sarkar et al., 2010). In another randomized intervention that trained older people to use tablet computers, it was found that instruction on a broad range of practical applications involving Internet use led to significant improvements in episodic memory and processing speed (Chan, Haber, Drew, & Park, 2014). In CMM, the pertinent linkage is from news attention to knowledge gain. In particular, previous studies have shown that news attention and reliance can enhance learning (Beaudoin & Thorson, 2004; Eveland, 2001, 2002; Eveland et al., 2003; Ho et al., 2013; Jensen, 2011). By extension, we predict that Internet use for health purposes will help improve people's health literacy. H4: Health-related Internet use is positively associated with health literacy.

2.5. Path 5: perceived health information overload to health literacy Information overload influences how people determine whether information processing should be continued or discontinued, as well as the subsequent impact it will have. When information supply exceeds one's information-processing capacity, the individual confronts problems in identifying relevant information, becomes overly selective and neglects a large amount of information, faces difficulties in understanding the association between details and the overall perspective, needs more time and effort to reach a decision, and makes inaccurate decisions (Eppler & Mengis, 2004). Similarly, when people are faced with overly complex health information, they may not perceive the information as being personally relevant or worthwhile and may not be able to surmise the necessary behavioral information (Davis, Williams, Branch, Green, & Whaley, 2000). Under such circumstances, people tend to develop negative attitudes toward the health information, resulting in an unwillingness to further seek and digest information and, finally, a reduction in health and medical knowledge and related skills. Thus, information overload is an important barrier to health literacy in terms of reducing people's understanding of health information, including diagnoses and treatment, and limiting their ability to evaluate and act upon health information (Kim, Lustria, Burke, & Kwon, 2007). In CMM, several studies have shown that elaboration has a significant impact on public affairs understanding, political affairs knowledge, and health information comprehension (Beaudoin & Thorson, 2004; Eveland, 2002; Jensen, 2011). For example, Ho et al. (2013) documented the significant effects of elaboration on H1N1 knowledge. Specific to information overload, Beaudoin

243

(2008) found evidence of an inverse correlation between perceived information overload and interpersonal trust. Based in these studies, we expect that people's perception of health information overload will be inversely associated with health literacy. H5: Perceived health information overload is inversely associated with health literacy.

2.6. Path 6: motivation for health-related internet use to health literacy Previous studies have shown that motivation for media use predicts knowledge development (Kwak, 1999; McLeod & McDonald, 1985; Vincent & Basil, 1997). In contrast to such direct effects, CMM contends that motivation does not exert direct effects on news learning and, in contrast, hypothesizes indirect effects as mediated by news attention and elaboration (Eveland, 2001). Support for this CMM path has been consistently found in some prior research (Eveland, 2002; Eveland et al., 2003). For instance, Beaudoin and Thorson (2004) documented non-significant paths from surveillance and guidance gratifications to political knowledge. Similarly, Ho et al. (2013) found that the effects of surveillance and other motivation dimensions were indirect on H1N1 knowledge, as mediated by news attention and elaboration. Finally, Jensen (2011) found that the effects of surveillance motivation on cancer news comprehension were fully mediated by news elaboration. Not all empirical testing is consistent here, however. Eveland (2001) found support in only two of three models. In his 1986 off-year election model and 1996 presidential election model, he documented a non-significant path from surveillance gratifications to knowledge gain, with significant indirect effects as mediated by news attention and elaboration. Counter to the theorizing, however, in the 1985 nonelection year model, the path from surveillance gratifications to public affairs knowledge was significant. Similarly, Lo et al. (2013) found evidence of partial mediation, with surveillance gratifications sought having a direct effect on increased knowledge about swine flu, as well as an indirect effect as mediated by news attention and news elaboration. Thus, this CMM path structure is found to be indicative of complete mediation by most empirical research. Moreover, Beaudoin (2008) provides additional support for complete mediation, with the effects of motivation on interpersonal trust mediated by both Internet use and perceived information overload. In line with the documentation of complete mediation in some of this prior research and the basic postulation of CMM, we predict that the impact of motivation for health-related Internet use on health literacy will be mediated. H6: The influence of motivation for health-related Internet use on health literacy is mediated by health-related Internet use and perceived health information overload. 3. Methods 3.1. Sample The six hypotheses were tested with secondary analysis of the US-based Health Information National Trends Survey (HINTS). The HINTS program is a nationally-representative survey that is periodically administered by the National Cancer Institute. With the aim of assessing the American public's health information use and health behavior, the HINTS mail survey includes varied topics such as how people find health information, what information sources

244

S. Jiang, C.E. Beaudoin / Computers in Human Behavior 58 (2016) 240e248

they use, and their feelings about the search process (http://hints. cancer.gov/). The current study analyzed HINTS 4 Cycle 3 dataset, which was collected from September 2013 to December 2013. Questionnaires were sent to participants by mail, with a $2 monetary incentive to encourage participation. The final HINTS 4 Cycle 3 sample consists of 3,185 respondents, with an overall response rate of 35.19%. Because the current study centers on online health activities and outcomes, we excluded respondents who reported having never used the Internet before. Also, missing values were replaced by grand mean to generate a complete dataset. Therefore, the final sample size was 2305. 3.2. Measurement Motivation for health-related Internet use was operationally defined as one's willingness to select the Internet based on sought motives for health information. Similar to some prior research on information gratifications (Stafford et al., 2004), the current study measured motivation for health-related Internet use by asking respondents to identify the communication channel they would use when they had a strong need for information about health or medical topics. For responses, 15 communication channels were listed (e.g., Internet, doctors, family, friends, books, brochures, newspapers, magazines, cancer organizations, library). If respondents selected the Internet as a means to satisfy their health information needs, it was coded as 1 (and 0 otherwise). Health-related Internet use was operationally defined as the adoption of sub-types of Internet use for health purposes. Drawn from prior research using HINTS dataset (Xiao, Sharman, Rao, & Upadhyaya, 2014), health-related Internet use was measured by asking respondents to identify, in the past 12 months, whether they had performed the following five health-related activities on the Internet: 1) sharing health information on social networking sites, 2) participating in online support groups, 3) watching healthrelated videos on YouTube, 4) looking for health or medical information on the Internet, and 5) looking for a health care provider on the Internet. An additive index of these dichotomous items (0 ¼ no, 1 ¼ yes) was then created to represent health-related Internet use. The higher score indicates more diverse Internet use for health purposes. Perceived health information overload was operationally defined in terms of whether one's past health information seeking experience was overly complex and time-consuming. Generally adapted from the Information Seeking Experience (ISEE) scale (Arora et al., 2008), perceived health information overload was assessed with three items that asked respondents to identify the degree to which they agree with the following statements: 1) It took a lot of effort to get the information you needed; 2) you felt frustrated during your search for the information; and 3) the information you found was hard to understand. Factor analysis (principal components analysis) identified one dimension to these items (eigenvalue ¼ 2.28, variance explained ¼ .76; Cronbach's a ¼ .84). Responses were scored on a 4-point scale (1 ¼ strongly disagree to 4 ¼ strongly agree) and then summed and averaged. Health literacy was operationally defined as one's knowledge and understanding of health-related issues. Basically adapted from the HINTS Health Literacy Screening Measure (Champlin & Mackert, 2015), five items were selected to measure health literacy. The first three items asked if respondents had ever heard of patient engagement in medical research, genetic tests, and the cervical cancer vaccine or HPV shot. The fourth item asked if respondents believed that the U.S. Food and Drug Administration (FDA) regulates tobacco products. Responses for these first four items were dichotomous (0 ¼ no, 1 ¼ yes). For the fifth item, respondents were first asked to read an ice cream food label provided

in the questionnaire. Based on the food label information, respondents were then required to calculate how many calories a person would eat if he/she ate an entire container of ice cream. Participants who answered correctly were scored 1, while others were scored 0. These five dichotomous items align with the multidimensional model of health literacy proposed by Zarcadoolas, Pleasant, and Greer (2005): scientific literacy (knowledge of medical research, genetic testing and HPV vaccine), civic literacy (knowledge of U.S. Food and Drug Administration), and fundamental literacy (calculation of calories). In the current study, responses to the five items were summed to create an index (0e5) on health literacy. Demographics and health-related variables were used as controls to reduce potential confounding effects. Demographics included age, gender (0 ¼ female, 1 ¼ male), education (ranging from 1 ¼ less than 8 years to 7 ¼ postgraduate), household income, and ethnicity (1 ¼ Non-Hispanic White, 0 ¼ Other). Health-related variables included health insurance coverage (1 ¼ yes, 0 ¼ no) and perceived health status (1 ¼ poor, 5 ¼ excellent). 3.3. Statistical analysis We performed structural equation modeling (SEM) to test our hypotheses, with the software Stata 13. In the covariance structure analysis, maximum likelihood of estimation was used. Good fit was signified by a non-significant p-value of the c2 statistic, a root mean square error of approximation (RMSEA) value of .06 or less, a comparative fit index (CFI) value of .95 or higher, and a root mean square residual (SRMR) value of below .05 (Hu & Bentler, 1999). In SEM, the control variables served as exogenous variables while endogenous variables included motivation for health-related Internet use, health-related Internet use, perceived health information overload, and health literacy. Paths were drawn from exogenous variables to all endogenous variables. Consistent with Fig. 1, paths were also drawn from motivation for health-related Internet use to the other three endogenous variables, from health-related Internet use to perceived health information overload and health literacy, and, lastly, from perceived health information overload to health literacy. (Exogenous control variables are not depicted in Fig. 1 for the sake of clarity of presentation.) Finally, to test closely for mediation, the Sobel test (1982) was used to yield z-score products with significance. 4. Results Descriptive statistics were shown in Table 1. The mean age was about 51, and 37.3% were male and 62.5% were White. The average education level was 5.11, which was “some college”, and the mean household income was 5.58, which was above $50,000. About 88% of respondents had health insurance coverage. Their average selfreported health condition was above good. The mean for motivation for health-related Internet use was about 43.7%. Health-related Internet use had the mean score of 1.75. The mean for perceived health information overload was about 2. The mean for health literacy was 2.56. The initial SEM was saturated, failing to provide a Chi-square value. Of the paths from endogenous variables, only one was non-significantdfrom motivation for health-related Internet use to health literacy. This path was, thus, pruned to maintain a more parsimonious model. Then the pruned model was run, resulting in a satisfactory model. This pruned model had a good fit, c2(1) ¼ 1.07, p ¼ .30; RMSEA ¼ .006 (90% confidence interval: .000 - .056); CFI ¼ 1.000; and SRMR ¼ .002. Fig. 2 shows this final SEM. The Bentler-Raykov squared multiple correlation coefficients indicated

S. Jiang, C.E. Beaudoin / Computers in Human Behavior 58 (2016) 240e248 Table 1 Descriptive statistics. Variable

Mean

SD

Age Gender (Male ¼ 1)a Education Household income Ethnicity (White ¼ 1)a Health insurancea Health status Motivation for health-related Internet usea Health-related Internet use Perceived health information overload Health literacy

51.16 37.27% 5.11 5.58 62.47% 88.03% 3.52 43.69% 1.75 2.01 2.56

15.10

a

1.47 1.99

.89 1.17 .77 1.13

Represents a frequency for dichotomous variable.

the following variance: motivation for health-related Internet use, 5.63%; health-related Internet use, 12.98%; health information overload, 4.95%, and health literacy, 15.09%. Table 2 shows the direct and indirect predictors of the model's endogenous variables. The depicted effects include those of exogenous control variables. For example, education has significant direct effects on motivation for health-related Internet use (b ¼ .14, p < .001), health-related Internet use (b ¼ .11, p < .001), and health literacy (b ¼ .16, p < .001). In addition, age had significant negative direct effects on

245

health-related Internet use (b ¼ .11, p < .001) and health-related Internet use (b ¼ .27, p < .001). Hypothesis 1 predicted that motivation for health-related Internet use is positively associated with health-related Internet use. The results shown in Fig. 2 and Table 2 specify this positive relationship (b ¼ .13, p < .001). Thus, Hypothesis 1 is supported. Hypothesis 2 indicated that the greater the motivation for health-related Internet use, the lower perceived health information overload. As shown in Fig. 2 and Table 2, there is evidence of this negative relationship (b ¼ .11, p < .001), which supports the hypothesis. Hypothesis 3 posited that health-related Internet use is positively associated with perceived health information overload. The results depicted in Fig. 2 and Table 2 specify this relationship. Health-related Internet use was a positive predictor of perceived health information overload (b ¼ .05, p < .05), supporting Hypothesis 3. Hypothesis 4 posited that the greater health-related Internet use, the higher health literacy. As shown in Fig. 2 and Table 2, health-related Internet use was positively associated with health literacy (b ¼ .17, p < .001). Therefore, Hypothesis 4 is supported. Hypothesis 5 predicted that health information overload is inversely associated with health literacy. As illustrated in Fig. 2 and Table 2, this negative relationship was significant (b ¼ .08, p < .001). Thus, Hypothesis 5 is supported.

Fig. 2. Final SEM on development of health literacy.

Table 2 Predictors of endogenous variables in SEM.

Motivation for healthrelated Internet use Health-related Internet use Perceived health information overload Health literacy

Effect

Age

Education Gender Household Ethnicity Health Health Motivation for health- Health-related Perceived health (Male ¼ 1) income (White ¼ 1) status insurance related Internet use Internet use Information overload

Direct Indirect Direct Indirect Direct Indirect

.11*** NA .27*** .01*** .02 .01

.14*** NA .11*** .02*** .04 .01*

.01 NA .10*** .01 .02 .01

.08** NA .01 .01** .08** .01*

.04* NA .03 .01 .04 .01*

.03 NA .06** .01 .07** .01*

.07*** NA .01 .01** .07** .01**

.13*** NA .11*** .01***

.05* NA

Direct .01 .16*** Indirect .05*** .03***

.01 .02***

.11*** .01*

.17*** .01

.01 .01

.02 .01

P .03***

.17*** .01*

* p < .05. **p < .01. ***p < .001. Coefficients are standardized. NA ¼ not applicable; P ¼ pruned path.

.08*** NA

246

S. Jiang, C.E. Beaudoin / Computers in Human Behavior 58 (2016) 240e248

Hypothesis 6 posited that health-related Internet use and perceived health information overload mediate the effects of motivation for health-related Internet use on health literacy. As mentioned earlier, the direct path from motivation for healthrelated Internet use to health literacy was not significant and was, thus, pruned. That pattern, along with the direct effects from motivation for health-related Internet use and health-related Internet use to health literacy, provides a model consistent with this hypothesis (see Fig. 2). To examine these mediation effects more closely, the Sobel test was used to yield a z-value. The z-values for these two paths were both significant at the .001-level (i.e., 5.79 and 4.22). Both paths, thus, are indicative of mediation, which supports Hypothesis 6. 5. Discussion With a basis of the original CMM and its derivative research, the present study examines the roles of motivation for health-related Internet use, health-related Internet use and perceived health information overload in improving people's health literacy. It is notable that we expanded upon CMM and prior research by focusing on health literacy as the outcome variable. Our hypothesized model was supported in each case. As anticipated, motivation for health-related Internet use was associated with Internet use for health purposes, consistent with the traditional U&G theory and prior CMM research (Beaudoin, 2008; Beaudoin & Thorson, 2004; Blumler, 1979; Eveland, 2001). This result suggests that motivation for media use is a driving factor of the actual media use. The current study focuses on the health context, where learning health information is one of the most important needs for patients and others (Case, 2012). Thus, when people believe that the Internet can satisfy their health needs, they are more likely to use it for health activities. As hypothesized, motivation for health-related Internet use was negatively associated with perceived health information overload. This is not surprising given that previous studies supported the linkage between motivation and information elaboration (Beaudoin, 2008; Eveland, 2002; Eveland et al., 2003). Moreover, Beaudoin (2008) found evidence of an inverse association between social resource motivation and perceived information overload. The finding of the current study suggests that motivation to learn about health concerns and matters from the Internet is a strong driving force to reduce health information overload in the content-rich digital environment and help the information seeker find the most relevant and desirable information. It is, for this reason, important that designers and purveyors of online health information disseminate information in a form that is clear and unlikely to overwhelm online health information seekers. As expected, health-related Internet use had a positive effect on health literacy. This finding is consistent with previous studies that have documented the positive and direct influence of Internet use on health-related outcomes, such as self-management skills (Stinson, Wilson, Gill, Yamada, & Holt, 2009), medical knowledge (Lo et al., 2013) and health literacy (Gutierrez et al., 2014). This significant path demonstrates that, when people actively use the Internet for health activities, they naturally learn more about the health topic, increasing health literacy. Likewise, perceived health information overload was negatively associated with health literacy, which is consistent with prior research (Bawden & Robinson, 2009; Cline & Haynes, 2001). This inverse path indicates that, when people receive excessive health information or encounter difficulty in understanding health content, they may terminate or reduce information processing and, thus, learn less from the health information, reducing health literacy. These two findings are generally in line with prior CMM research that documented the

significant direct effects of news attention and information elaboration on public affairs, political, and health-related knowledge (Beaudoin & Thorson, 2004; Eveland, 2001, 2002; Eveland et al., 2003; Ho et al., 2013; Jensen, 2011). Also as predicted, health-related Internet use increased perceived health information overload, a finding consistent with prior research on online health information seeking (Cline & Haynes, 2001; Rice, 2006). When people are overwhelmed by health information found on the Internet and the quality of information is concerning, it will be difficult for them to fully understand health and medical information related to disease, symptoms, and treatment. Such information overload is especially strong when people are faced with the onslaught of information from various online media outlets (Kim et al., 2007). In this study, we measured health-related Internet use in terms of five different online health activities. Considering the fact that functional features of the five online health activities vary significantly and some are relatively new to the general public, such as health-related social media and online doctor-patient communication, it could be the case that people who partake in more such activities are intuitively expected to put forth more effort, encounter more difficulty, and feel greater frustration. While this study found support for the positive linkage between health-related Internet use and perceived health information overload, it was counter to Beaudoin's study (2008), which documented an inverse relationship. Explaining the different correlation between Internet use and perceived information overload could be a matter of measurement. Beaudoin (2008) measured Internet use in terms of its frequency and for general use, while we focused on the diversity of Internet use for health purposes. Thus, the conceptualization and measurement of Internet use matter in reducing information overload. Future studies should consider different ways to measure healthrelated Internet use, such as frequency and attention, and examine its differential impact on perceived health information overload. Another important finding pertains to our model's statistical mediation. Our results showed that motivation for health-related Internet use failed to influence health literacy directly. Instead, the impact of motivation was completely mediated by healthrelated Internet use and perceived health information overload. This finding is consistent with the original CMM and other derivative research, stating that motivation is only indirectly related to political learning (Wei & Lo, 2008), medical knowledge acquisition (Jensen, 2011) and interpersonal trust (Beaudoin, 2008). This adds to the growing body of research that supports complete mediation in CMM studies and suggests that media use and elaboration are consequent stages in improving health literacy and that motivation, by itself, is not enough to bring about change in one's health literacy. Five important limitations of this study should be noted. The first two limitations relate to measurement, which can often be a challenge in secondary data analysis. In this study, a single item was used to measure motivation. Future research should use multiple items to test this construct. Also, although the measurement of health literacy was based on one previous study (Champlin & Mackert, 2015), this new scale needs further validation. Second, the current study only investigated one type of motivation, which is the intention of Internet use to obtain health information and, thus, shares common ground with surveillance gratification. Research on CMM, of course, has informed us that a variety of motivations (e.g., anticipated interaction, guidance) are at play in different media use scenarios (Beaudoin & Thorson, 2004; Ho et al., 2013). Future studies should test whether other kinds of motivation can drive health-related Internet use and subsequent levels of health literacy. Third, this study only took into account one type of information

S. Jiang, C.E. Beaudoin / Computers in Human Behavior 58 (2016) 240e248

processing, perceived health information overload. As Ho et al. (2013) suggested, there may be merit to considering the more traditional CMM measure of elaboration processing to help explain health information processes. Fourth, it is notable the variance explained in health literacy was not very high. While our model employed the critical CMM variables, as well as assorted demographic and health control variables, future research should employ an assortment of other controls, including general literacy. Fifth, although SEM suggests a direction of influence, it cannot conclusively demonstrate causation. That said, prior research has documented the casual links of CMM using panel data (Eveland et al., 2003). Sixth, the HINTS response rate was somewhat low at 35%, which is below the average of 45% documented by a metaanalysis of mail surveys (Shih & Fan, 2008). This could weaken the overall representativeness of the survey and, as well as, the validity of the study. In conclusion, this study has made several significant theoretical contributions. First, our study extends previous CMM research and its derivative models by focusing on a different type of media usedhealth-related Internet usedand a different outcomedhealth literacydproviding additional empirical support for the applicability of CMM in the health context. Second, the indirect link between motivation and health literacy found in this study indicates that CMM effects occur in various contexts and across media platforms. Future studies should continue this stream of research to test different possible mediators, such as information elaboration, which would add significantly to the explanatory power of CMM. Third, the positive path from health-related Internet use to perceived health information overload was an important finding. This suggests a critical difference between measuring frequency of general Internet use, as Beaudoin (2008) did, and measuring diversity of health-related Internet use, as we did. To thoroughly understand media use and subsequent learning, close attention should be paid to the nature of the issue at hand (e.g., healthrelated vs. general), the media outlets (e.g., traditional media vs. new media), and the usage measurement (e.g., frequency vs. diversity). Without differentiating media use in this manner, important media effects may be obscured, with theoretical development waylaid. In terms of practical implications, our findings provide refined knowledge on health literacy development to health educators and health website designers. Simply motivating people to use the Internet for health purposes is not sufficient to improving their health literacy. Thus, future health interventions should incorporate strategies that can not only encourage people to seek out health information online, but also develop online technology skills that can help them use the Internet effectively and mitigate information overload. In addition, strategies such as inviting health care providers to communicate directly with patients via online forums, social networking sites, or patient portals might help improve people's understanding of health and medical information. In summary, understanding the mechanisms behind health-related Internet use and health literacy can help health practitioners and media designers better disseminate information, promote people's health literacy, and ultimately improve their health outcomes in this digital era.

References Anderson, I. K. (2011). The uses and gratifications of online care pages: a study of CaringBridge. Health Communication, 26, 546e559. Armstrong, C. B., & Rubin, A. M. (1989). Talk radio as interpersonal communication. Journal of Communication, 39, 84e94. Arora, N. K., Hesse, B. W., Rimer, B. K., Viswanath, K., Clayman, M. L., & Croyle, R. T. (2008). Frustrated and confused: the American public rates its cancer-related information-seeking experiences. Journal of General Internal Medicine, 23,

247

223e228. Bawden, D., Holtham, C., & Courtney, N. (1999). Perspectives on information overload. Aslib Proceedings, 51, 249e255. Bawden, D., & Robinson, L. (2009). The dark side of information overload, anxiety and other paradoxes and pathologies. Journal of Information Science, 35, 180e191. Beaudoin, C. E. (2008). Explaining the relationship between internet use and interpersonal trust: taking into account motivation and information overload. Journal of Computer-Mediated Communication, 13, 550e568. Beaudoin, C. E., & Thorson, E. (2004). Testing the cognitive mediation model: the roles of news reliance and three gratifications sought. Communication Research, 31, 446e471. Blumler, J. G. (1979). The role of theory in uses-and-gratifications research. Communication Research, 6, 9e36. Case, D. O. (2012). Looking for information: A survey of research on information seeking, needs and behavior. Bingley, UK: Emerald Group Publishing. Champlin, S., & Mackert, M. (2015). Creating a screening measure of health literacy for the health information national trends survey. Advance online publication American Journal of Health Promotion. http://dx.doi.org/10.4278/ajhp.140604ARB-253. Chan, M. Y., Haber, S., Drew, L. M., & Park, D. C. (2014). Training older adults to use tablet computers: does it enhance cognitive function?. Advance online publication The Gerontologist. http://dx.doi.org/10.1093/geront/gnu057. Chan, Y. M., & Huang, H. (2013). Weight management information overload challenges in 2007 HINTS: socioeconomic, health status and behaviors correlates. Journal of Consumer Health on the Internet, 17, 151e167. Cho, J., De Zuniga, H. G., Rojas, H., & Shah, D. V. (2003). Beyond access: the digital divide and internet uses and gratifications. IT & Society, 1, 46e72. Chung, D. S., & Kim, S. (2008). Blogging activity among cancer patients and their companions: uses, gratifications, and predictors of outcomes. Journal of the American Society for Information Science and Technology, 59, 297e306. Cline, R. J., & Haynes, K. M. (2001). Consumer health information seeking on the internet: the state of the art. Health Education Research, 16, 671e692. Davis, T. C., Williams, M. V., Branch, W., Green, K. W., & Whaley, B. (2000). Explaining illness to patients with limited literacy. In B. B. Whaley (Ed.), Explaining illness: Research, theory, and strategies (pp. 123e146). Mahwah, NJ: Lawrence Erlbaum Associates. Deci, E. L., & Ryan, R. M. (1985). Intrinsic motivation and self-determination in human behavior. New York: Plenum.
n-Rodríguez, M. (2012). A content De la Torre-Díez, I., Díaz-Pernas, F. J., & Anto analysis of chronic diseases social groups on Facebook and Twitter. Telemedicine and e-Health, 18, 404e408. Eppler, M. J., & Mengis, J. (2004). The concept of information overload: a review of literature from organization science, accounting, marketing, MIS, and related disciplines. The Information Society, 20, 325e344. Eveland, W. P., Jr. (2001). The cognitive mediation model of learning from the news: evidence from nonelection, off-year election, and presidential election contexts. Communication Research, 28, 571e601. Eveland, W. P., Jr. (2002). News information processing as mediator of the relationship between motivations and political knowledge. Journalism & Mass Communication Quarterly, 79, 26e40. Eveland, W. P., Jr., & Dunwoody, S. (2001). User control and structural isomorphism or disorientation and cognitive load? Communication Research, 28, 48e78. Eveland, W. P., Jr., Shah, D. V., & Kwak, N. (2003). Accessing causality in the cognitive mediation model: panel study of motivations, information processing, and learning during campaign 2000. Communication Research, 30, 359e386. Eysenbach, G., & Kohler, C. (2002). How do consumers search for and appraise health information on the world wide web? qualitative study using focus groups, usability tests, and in-depth interviews. British Medical Journal, 324, 573e577. Fox, S. (2011). The social life of health information 2011. Washington, DC: Pew Internet & American Life Project. Fox, S., & Duggan, M. (2012). Mobile health 2012. Washington, DC: Pew Internet & American Life Project. Ghaddar, S. F., Valerio, M. A., Garcia, C. M., & Hansen, L. (2012). Adolescent health literacy: the importance of credible sources for online health information. Journal of School Health, 82, 28e36. Guitton, M. J. (2015). Online maritime health information: an overview of the situation. International Maritime Health, 66, 139e144. Gutierrez, N., Kindratt, T. B., Pagels, P., Foster, B., & Gimpel, N. E. (2014). Health literacy, health information seeking behaviors and internet use among patients attending a private and public clinic in the same geographic area. Journal of Community Health, 39, 83e89. Ho, S. S., Peh, X., & Soh, V. W. (2013). The cognitive mediation model: factors influencing public knowledge of the H1N1 pandemic and intention to take precautionary behaviors. Journal of Health Communication, 18, 773e794. Hu, L. T., & Bentler, P. M. (1999). Cutoff criteria for fit indexes in covariance structure analysis: conventional criteria versus new alternatives. Structural Equation Modeling, 6, 1e55. Jensen, J. D. (2011). Knowledge acquisition following exposure to cancer news articles: a test of the cognitive mediation model. Journal of Communication, 61, 514e534. Ji, Q., Ha, L., & Sypher, U. (2014). The role of news media use and demographic characteristics in the possibility of information overload prediction. International Journal of Communication, 8, 699e714.

248

S. Jiang, C.E. Beaudoin / Computers in Human Behavior 58 (2016) 240e248

Josefsson, U. (2006). Patients' online information seeking behavior. In M. Murero, & R. E. Rice (Eds.), The internet and health care: Theory, research, and practice (pp. 127e147). Mahwah, NJ: Lawrence Erlbaum Associates. Katz, E., Blumler, J. G., & Gurevitch, M. (1973). Uses and gratifications research. Public Opinion Quarterly, 37, 509e523. Kim, K., Lustria, M. L. A., Burke, D., & Kwon, N. (2007). Predictors of cancer information overload: findings from a national survey. Information Research, 12(4), 28. Kwak, N. (1999). Revisiting the knowledge gap hypothesis education, motivation, and media use. Communication Research, 26, 385e413. Lang, A. (2000). The limited capacity model of mediated message processing. Journal of Communication, 50, 46e70. Leung, L. (2001). College student motives for chatting on ICQ. New Media & Society, 3, 483e500. Lomanowska, A. M., & Guitton, M. J. (2014). My avatar is pregnant! representation of pregnancy, birth, and maternity in a virtual world. Computers in Human Behavior, 31, 322e331. Lo, V.-H., Wei, R., & Su, H. (2013). Self-efficacy, information-processing strategies, and acquisition of health knowledge. Asian Journal of Communication, 23, 54e67. Mano, R. S. (2014). Social media and online health services: a health empowerment perspective to online health information. Computers in Human Behavior, 39, 404e412. McLeod, J. M., & McDonald, D. G. (1985). Beyond simple exposure: media orientations and their impact on political processes. Communication Research, 12, 3e33. Meppelink, C. S., & Bol, N. (2015). Exploring the role of health literacy on attention to and recall of text-illustrated health information: an eye-tracking study. Computers in Human Behavior, 48, 87e93. Mitchell, S. E., Sadikova, E., Jack, B. W., & Paasche-Orlow, M. K. (2012). Health literacy and 30-day postdischarge hospital utilization. Journal of Health Communication, 17(3), 325e338. Murero, M., & Rice, R. E. (2006). E-health research. In M. Murero, & R. E. Rice (Eds.), The internet and health care: Theory, research, and practice (pp. 3e26). Mahwah, NJ: Lawrence Erlbaum Associates. National Cancer Institute. (2014). Health information national trends survey (HINTS). Retrieved from http://hints.cancer.gov/. Omar, A. S., Rashid, W. E. W., & Majid, A. A. (2014). Motivations using social networking sites on quality work life. Procedia-Social and Behavioral Sciences, 130, 524e531. Park, N., Kee, K. F., & Valenzuela, S. (2009). Being immersed in social networking environment: Facebook groups, uses and gratifications, and social outcomes. CyberPsychology & Behavior, 12, 729e733. Payne, G. A. (1988). Uses and gratifications motives as indicators of magazine readership. Journalism Quarterly, 65, 909e913. Petty, R. E., & Cacioppo, J. T. (1986). The elaboration likelihood model of persuasion.

New York: Springer. Reeve, J. (1997). Understanding motivation and emotion (2 ed.). Fort Worth, TX: Harcourt Brace College Publishers. Rice, R. E. (2006). Influences, usage, and outcomes of internet health information searching: multivariate results from the pew surveys. International Journal of Medical Informatics, 75, 8e28. Rubin, A. M. (1983). Television uses and gratifications: the interactions of viewing patterns and motivations. Journal of Broadcasting & Electronic Media, 27, 37e51.
pez, A., et al. (2010). The Sarkar, U., Karter, A. J., Liu, J. Y., Adler, N. E., Nguyen, R., Lo literacy divide: health literacy and the use of an internet-based patient portal in an integrated health system-results from the diabetes study of northern California (DISTANCE). Journal of Health Communication, 15, 183e196. Schulz, P. J., & Nakamoto, K. (2011). “Bad” literacy, the internet, and the limits of patient empowerment. In Paper presented at the AAAI spring symposium series, Stanford, CA, USA. Shih, T.-H., & Fan, X. (2008). Comparing response rates from web and mail surveys: a meta-analysis. Field Methods, 20(3), 249e271. Sobel, M. E. (1982). Asymptotic confidence intervals for indirect effects in structural equation models. Sociological Methodology, 13, 290e312. €thlin, F., Ganahl, K., Slonska, Z., Doyle, G., et al. (2015). Sørensen, K., Pelikan, J. M., Ro Health literacy in Europe: comparative results of the European health literacy survey (HLS-EU). Advance online publication The European Journal of Public Health, 25(6), 1053e1058. http://dx.doi.org/10.1093/eurpub/ckv043. Soucek, R., & Moser, K. (2010). Coping with information overload in email communication: evaluation of a training intervention. Computers in Human Behavior, 26, 1458e1466. Stafford, T. F., Stafford, M. R., & Schkade, L. L. (2004). Determining uses and gratifications for the internet. Decision Sciences, 35, 259e288. Stinson, J., Wilson, R., Gill, N., Yamada, J., & Holt, J. (2009). A systematic review of internet-based self-management interventions for youth with health conditions. Journal of Pediatric Psychology, 34, 495e510. USDHHS. (2000). Healthy people 2010: Understanding and improving health (2nd ed.). Washington, DC: U.S. Department of Health and Human Services, U.S. Government Printing Office. Vincent, R. C., & Basil, M. D. (1997). College students' news gratifications, media use, and current events knowledge. Journal of Broadcasting & Electronic Media, 41, 380e392. Wei, R., & Lo, V.-H. (2008). News media use and knowledge about the 2006 US midterm elections: why exposure matters in voter learning. International Journal of Public Opinion Research, 20, 347e362. Xiao, N., Sharman, R., Rao, H. R., & Upadhyaya, S. (2014). Factors influencing online health information search: an empirical analysis of a national cancer-related survey. Decision Support Systems, 57, 417e427. Zarcadoolas, C., Pleasant, A., & Greer, D. S. (2005). Understanding health literacy: an expanded model. Health Promotion International, 20, 195e203.