Search tips
Search criteria 


Logo of nihpaAbout Author manuscriptsSubmit a manuscriptHHS Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
Soc Sci Med. Author manuscript; available in PMC 2012 July 1.
Published in final edited form as:
PMCID: PMC3126911

Determinants and beliefs of health information mavens among a lower-socioeconomic position and minority population


People of lower-socioeconomic position (SEP) and most racial/ethnic minorities face significant communication challenges which may negatively impact their health. Previous research has shown that these groups rely heavily on interpersonal sources to share and receive health information; however, little is known about these lay sources. The purpose of this paper is to apply the concept of a market maven to the public health sector with the aims of identifying determinants of high health information mavenism among low-SEP and racial/ethnic minority groups and to assess the information they may be sharing based on their own health beliefs. Data for this study were drawn from the baseline survey (n=325) of a US randomized control intervention study aimed at eliciting an understanding of Internet-related challenges among lower-SEP and minority individuals. Regression models were estimated to distinguish significant determinants of health information mavenism among the sample. Similarly, bivariate and logistic multivariable models were estimated to determine the association between health information mavenism and accurate health beliefs relating to diet, physical activity and smoking. The data illustrate that having a larger social network, being female and being older were important factors associated with higher mavenism scores. Additionally being a moderate consumer of general media as well as fewer years in the US and lower language acculturation were significant predictors of higher mavenism scores. Mavens were more likely than non-mavens to maintain accurate beliefs regarding diet; however, there was no distinction between physical activity and smoking beliefs between mavens and non-mavens. These results offer a unique understanding of health information mavenism which could better leverage word-of-mouth health communication efforts among lower-SEP and minority groups in order to reduce communication inequalities. Moreover, the data indicate that health information mavens may serve as an ideal point of intervention in attempts to modify health beliefs with the goal of reducing health disparities among these populations.

Keywords: communication inequalities, health disparities, health information mavenism


The demands placed on the public as health consumers are ever increasing. The confluence of an information rich environment along with the multitude of health care decisions the public is asked to make, from healthy behavior choices to what health plan to subscribe to, create unique challenges for health communication (Viswanath, 2005). These challenges are further magnified for the most disadvantaged populations, such as members of lower-socioeconomic position (SEP) and most racial/ethnic minorities. These groups within the United States and other nations with significant barriers to health care access not only carry a disproportionate burden of poor health but they are also subject to significant communication inequalities- differences in the access, use and processing of health information, which perpetuate gaps in health knowledge and beliefs and ultimately exacerbate disparities in health outcomes (Viswanath, 2006). Factors such as limited literacy skills, and lack of access to popular sources of health information (e.g. the Internet) are some of the leading communication inequalities that these groups face (Cotten & Gupta, 2004; Gansler, et al., 2005; Kontos, Bennett, & Viswanath, 2007; Lorence & Park, 2007; Rudd, Moeykens, & Colton, 2000).

As a result of these barriers, studies within the US have shown that lower-SEP and minority groups, especially those for whom English is a second language, are more likely to rely on interpersonal sources for health information compared to their counterparts (Cheong, 2007; Risker, 1995; Vanderpool, Kornfeld, Finney Rutten, & Squiers, 2009). In addition, given that many of these individuals lack access to health care and thus regular contact with medical providers, communication about health can be limited to lay interpersonal sources such as family and friends within one s social network rather than professional sources such as medical providers (Bonds, Foley, Dugan, Hall, & Extrom, 2004; Smith, Dixon, Treena, Nutbeam, & McCaffery, 2009).

It follows then, that social marketers and other health communicators who wish to disseminate information to more disadvantaged populations should work to incorporate lay interpersonal communication channels and word-of-mouth (WOM) exchange into their campaign strategies. However, little is known about lay interpersonal sources of health information, in general, and specifically among lower-SEP and minority groups within the US. Most of the literature surrounding interpersonal communication and health focuses either on patient-provider communication (Brundage, Feldman-Stewart, & Tishelman, 2010; Kreps, Arora, & Nelson, 2003; Makoul, 2003; Teutsch, 2003) or identifying and training people to serve as opinion leaders, peer-leaders or community health workers to aid in the implementation of designated health promotion/disease prevention interventions (Fisher, 1975; Klepp, Halper, & Perry, 1986; Perry, Klepp, Halper, Hawkins, & Murray, 1986; Valente & Pumpuang, 2007). Minimal attention has been paid in the public health literature to identifying who within lower-SEP and racial/ethnic minority networks may be providing health information, and what type of information they may be sharing based on their own health knowledge and beliefs.

In contrast, this more casual word-of-mouth information exchange among family and friends has been the focus of numerous marketing studies and advertising campaigns. Marketers have historically tried to identify and target certain types of individuals in an effort to leverage word-of-mouth information exchange, which has been shown to play an important role in shaping consumers attitudes and behaviors (Arndt, 1967; Bayns, 1985). These individuals have been labeled by the marketing sector as “market mavens”. A market maven is an interpersonal source of communication whose influence and consumer product information is based on general knowledge and experience (Abratt, Nel, & Nezer, 1995; Feick & Price, 1987). Although the concept of a maven is similar to that of an early adopter or opinion leader which have been extensively examined in relation to public health promotion (Rogers, 2004; Valente & Pumpuang, 2007), mavens are distinct in that they possess general rather than specific information and influence (Feick & Price, 1987). For example, early adopters knowledge is limited to a domain that they themselves have experienced, such as information on breast cancer because they (or a close relation) have personally experienced breast cancer (Rogers, 2004). Opinion leaders, another similar concept, do not need to have personal experience of a given domain, but like adopters their knowledge and influence is typically restricted to a specific domain (e.g. diabetes), rather than more broadly across several areas of health. This is most likely due to the fact that in public health, opinion leaders are typically selected and trained for a given health intervention or program (Valente & Pumpuang, 2007). Market mavens, in contrast, are people who have general knowledge about health and, in fact, consider it important to share their knowledge with others in their social network. That is, those within the social network of market mavens are subject to incidental exposure to health information, an important source of knowledge and persuasion.

The purpose of this paper is to apply the concept of a market maven to the public health sector with the primary aim of using regression analyses to indentify characteristics of higher health information mavenism among a lower-SEP, racially and ethnically diverse population in the United States. The application of this marketing tool will aid in audience segmentation and help garner insight into the distinguishing social and media-related characteristics of health information mavens, which could then be utilized for future public health communication dissemination efforts. The second aim of our study is to use multivariable logistic regression to assess mavens health beliefs concerning physical activity, diet and smoking in comparison to the beliefs of non-mavens in order to better understand what type of opinions or information they may be sharing among their social networks. Health beliefs are critical in shaping health behaviors and thus in attempts to change population health. There is substantial evidence, based on theories of health behavior such as the Theory of Planned Behavior, that indicate beliefs are important factors in influencing individual’s intentions to engage in health behaviors and ultimately the behavior itself (Azjen, 1991; Fishbein & Azjen, 1975). If mavens in lower-SEP and racial/ethnic minority groups possess health beliefs that are more or less aligned with national recommendations and standards of good health, they could prove to be a vital source in attempts to decrease existing communication inequalities as well as influence positive behavior change among these populations (Viswanath, 2006). However, if mavens maintain beliefs about health that are not aligned with general recommendations and well-established medical evidence, they could play a significant role in exacerbating communication inequalities and observed health disparities in sharing opinions counter to public health messaging. This scenario presents a more serious and larger public health issue that would warrant attention. It is our hope that by segmenting health information mavens among a low-SEP and minority population and assessing their health beliefs this evaluation will help to shape future health communication and social marketing initiatives targeted at lower-SEP and racial/ethnic minority groups both within the United States as well as inform strategies for improved health communication in other nations with significant health disparities or overall poor health.


Data source

Data for this study were drawn from the baseline survey of “Click to Connect: Improving health literacy through computer literacy” (C2C), a randomized control intervention study funded through the US National Cancer Institute (grant #5R01CA122894). Human subjects approval for this investigation was granted by the Dana-Farber Cancer Institute Institutional Review Board and the baseline data were collected from September 2007 through November 2009. The aim of the study is to elicit a better understanding of computer and Internet related challenges and barriers among groups of people from lower-socioeconomic position. The goal of C2C is to assess how lower-SEP individuals who have limited experience with and access to the Internet use the Web for health information, and whether information seeking, knowledge, beliefs, and health behaviors change as participants learning, use, and navigation of the Internet grows over time. As a component of the C2C study, participants are asked to complete a 45 minute telephone survey both at baseline and at 12 months post enrollment. Only baseline data were examined for the purposes of this study. Participants were enrolled across three study waves, resulting in a baseline survey sample of 325.

Sample characteristics

C2C participants were recruited from adult education centers in the metro-Boston, Massachusetts area from 2007–2009, and were eligible to participate in the study if they met the following criteria: no home Internet access; enrollment in a General Educational Development (or GED) or pre-GED class (this program, when passed, certifies that the taker has American or Canadian high school-level academic skills; or enrollment in a high-level English for Speakers of another Language (ESOL) class; between the ages of 25 and 60, and had a working phone number. A majority of the sample (59%) was at or below the federal poverty level. There were more women than men in the sample, but there was a wide distribution of people at different ages across the sample (see Table 1). The sample was racially and ethnically diverse, with half of the participants identifying as non-Hispanic black and a quarter identifying as Hispanic (see Table 1).

Table 1
Descriptive statistics of Click to Connect baseline survey sample (n=325)

Independent variables and covariates

Social environment characteristics

Participants social environment was assessed using the social network index from the National Institute on Aging s (NIA) Established Populations for the Epidemiologic Study of the Elderly (EPESE) study (Berkman, 1986; Berkman & Glass, 2000). The index includes such characteristics as marital status, having children, religious and community involvement, and the number of family and friends that one feels close to. The social network index range is 0–4 with those scoring a zero having no social network and those scoring a four having a very extensive social network (high score). This index was then dichotomized at the median into low and high, with respondents coded as low if they scored between one and two and high if they scored between three and four. No respondents scored a zero on the index.

Media environment characteristics

Health information seeking as well as both general media consumption and health-specific media exposure and attention were assessed in relation to health information mavenism. All media-related questions were based on those included in the 2003 and 2005 fielding of the National Cancer Institute s Health Information National Trends Survey (HINTS) (National Cancer Institute, 2010). Health information seeking, which captures individuals purposive rather than incidental interaction with health information, was measured with the question: “Did you ever try to find information on health?” those answering yes were coded as seekers and those responding no were considered nonseekers.

Weekly consumption of television and radio was measured by total hours per week. The weekly total was categorized into moderate exposure (sums at or below the national average, less than 22 hours per week for television and less than 15 hours per week for radio) and heavy exposure (above the national average, 22 hours or more per week for television and 15 hours or more per week for radio) (Bureau of Labor Statistics, 2009; The Nielsen Company, 2009). General news consumption was measured by how many days in the past week respondents: read a newspaper, watched the national news on television and watched the local news on television. Responses were dichotomized into most days of the week (≥4 days/wk) and less than most days (<4 days/wk) for each channel.

Health-specific media exposure was measured using the question format: “Some newspapers or general magazines publish a special section that focuses on health. In the past year, have you read health sections of the newspaper or general magazine?” and a follow-up question “About how often have you read such health sections in the past year?” with response options being once or more per week or less than once per week. This same format was used to capture exposure to health programs on local television news and incidental health information exposure on the Internet (“Some people notice information about health on the Internet, even when they are not trying to find out about a health concern. Have you read such health information on the Internet in the past year?”). Health-specific media exposure was coded into three categories: no exposure, less than once a week, once or more per week for all represented channels.

Health-specific media attention was measured with the question: “How much attention do you pay to information about health or medical topics that you happen to: see on TV, hear on the radio, read in newspaper, read on the Internet, get from a doctor or health care provider?” For each channel, participants had the following response options: a lot, some, a little, not at all, do not use, do not know. Answers were grouped based on previous uses of these variables into a lot/some and a little/not at all/do not use (Viswanath, et al., 2006). “Do not know” responses were coded as missing.

Socio-demographic characteristics

Several socio-demographic characteristics could potentially be influential in determining health information mavenism. We assessed the independent contribution of age (34yrs, 35–49yrs.,50yrs.); sex (male, female); race/ethnicity (non-Hispanic white, non-Hispanic black, Hispanic, other) and health status (excellent/very good/good, fair/poor). Because our survey sample had a high proportion of foreign-born individuals and individuals for whom English is a second language, we also included variables related to these communication-related demographics. Immigrant status was trichotomized into the following groups: born in the United States, in the US ten years or less, in the US more than ten years. Language was operationalized as language acculturation which was created from three variables identifying first or native language, language usually spoken at home and language respondents like to read in. The sum score of the three items was trichotomized into low, medium and high as done in previous research (Bennett, et al., 2008; Sorensen, et al., 2005). Respondents classified as high reported that English was either their first or primary language. Respondents with low language acculturation reported a foreign language as their first language and cited primarily speaking and reading in that language. Medium -level language acculturation individuals would fall in between the two extremes and represents individuals whose first language is not English but reported reading and speaking in English at least occasionally at home.

Dependent variables

Health information mavenism

To examine the association between social and media environment characteristics and health information mavenism, the outcome variable is a continuous health information mavenism score. The five item health information maven scale was adapted from the six item market maven scale (Feick & Price, 1987). The questions were: 1) I like introducing new health topics to my friends and family; 2) I like helping people by providing them with information about health; 3) People ask me for information about health; 4) If someone asked where to get the best information about a particular health topic, I could tell him or her where to go; 5) My friends think of me as a good source of information when it comes to new information about health. Response choices were scaled on a seven point Likert scale ranging from 1=strongly disagree to 7=strongly agree making the summed score range 5–35. The sixth item of the market maven scale was omitted from the health information maven scale because it specifically related to consumer goods and could not be adapted for a health context. The adapted five item scale had adequate internal consistency reliability with an overall Chronbach s alpha of 0.77.

Nutrition & physical activity beliefs

To assess the association between health information mavenism and general health beliefs we measured beliefs regarding nutrition, physical activity, and smoking using validated items from the national US Behavioral Risk Factor Surveillance Survey and the National Health Interview Survey (The Centers for Disease Control and Prevention, 2010a, 2010c). We assessed beliefs in relation to the national recommendation for servings of fruits and vegetables using the question: “In your opinion, how many servings of fruits and vegetables should a person eat each day for good health?”. Responses were coded being aligned with the national recommendations if they fell within the Centers for Disease Control and Prevention s (CDC) national recommendation of 5 or more servings per day (The Centers for Disease Control and Prevention, 2010b). Because this question format was open ended, we excluded outlier responses in our analyses (n=1). Beliefs regarding physical activity were measured with the question: “In your opinion, how many days a week of physical activity or exercise are recommended for the average adult to stay healthy?”. Responses were coded being as being aligned with national recommendations if they again fell within the CDC s recommendation of most days a week (4–7 days/wk) (The Centers for Disease Control and Prevention, 2010d). It is important to note that within the United States these daily recommendations are widely communicated to the public via large-scale multi-million dollar social marketing campaigns such as the 5-a-Day Campaign and state-based physical activity programs (The Centers for Disease Control and Prevention, 2010e).

Smoking, secondhand smoke & cancer beliefs

Beliefs regarding the association between cigarette use and cancer were assessed by responses to two questions adapted from the American Legacy Foundation s 2003 American Smoking and Health Survey (ASHES) (American Legacy Foundation, 2005). Respondents were asked to rate how much they agree or disagree (on a four point Likert scale) to the following questions: “Smoking cigarettes has not been proven to cause cancer” and “Inhaling someone else’s cigarette smoke can cause lung cancer in nonsmokers.” Respondents answering that they strongly or somewhat disagree to the first statement were coded as having beliefs concordant with the well established medical evidence regarding the link between smoking and cancer. Respondents answering that they somewhat or strongly agree to the second statement were coded as having beliefs again concordant with medical evidence regarding the link between secondhand smoke and cancer.


To investigate our primary research question regarding general characteristics of health information mavens, we first estimated the unadjusted association between listed demographic, social and media-related characteristics and health information mavenism score. Since the mavenism outcome variable was continuous (range: 5 to 35), we estimated the adjusted associations using backwards stepwise multivariable linear regression techniques. Though the distribution of mavenism was slightly left skewed (Hougaard s measure of skewness: −1.21), an OLS regression model was justified since our sample size was large enough to approximate a normal distribution. We conducted complete case analysis for each regression and therefore any subject with missing data for any of the variables of interest were excluded from analysis. Because the number of missing never exceeded n=21, which represented less than 7% of the data, no adjustments were made to account for missing data in order to reduce the introduction of bias into our estimates. All variables that were significant at the 0.25 alpha level in bivariate analyses were considered for inclusion in a multivariable model of mavenism. We then used backwards selection techniques and removed variables one at a time until all variables included in the final multivariable model were significant at the 0.10 alpha level. Because this was an exploratory analysis and there were no a priori assumptions of which independent variables needed to be included, backwards stepwise selection allowed for a more parsimonious model than if we had included all variables regardless of significance and thus reduced error in our final model.

To examine our second research question that focused on the general health beliefs of mavens, we again first estimated unadjusted bivariate associations between high and low maven status and our four health belief outcomes relating to diet, physical activity and smoking. Participants were coded as ranking high on the maven scale if they scored 30 or higher since this cut off point groups all of the participants that agreed or strongly agreed with all of the scale items. Participants were coded as ranking low on the maven scale if they scored less than 30, indicating disagreement with the scale items. Next, we estimated multivariable logistic regression models using backwards stepwise selection techniques as previously described. We retained the health information maven variable in each logistic model, regardless of significance, since it was our primary variable of interest.


Bivariate results

Characteristics of health information mavens

Nearly half (44%) of the Click to Connect participants identified as being a health information maven. The mean maven score was 27.69 (range 5–35 and a standard deviation of 5.77), and scores were approximately normally distributed.

Several distinguishing trends of health information mavenism emerged from the bivariate analyses. Having a larger social network, being female and being older were important social and demographic factors that were associated with higher mavenism scores (see Table 2). While race/ethnicity was not significantly associated with mavenism, nativity/years in the US as well as language acculturation were.

Table 2
Social and media environment determinants of health information mavenism

Exposure and attention to health-specific media were also significant predictors of higher mavenism scores in the bivariate analyses. A linear trend was evidenced for both exposure and attention with participants reporting the highest levels of exposure and attention also reporting the highest mavenism scores while those participants with no exposure or paying little attention to health media reporting the lowest mavenism scores (see Table 2). This linear trend persisted across all three identified media sources (newspaper or magazine, local TV news and Internet). However, this trend was not evidenced for the only non-media related variable- information provided from a health care provider. Increased attention to medical providers was not a significant determinant of increased mavenism score. Another interesting relationship which emerged was that individuals who consumed a moderate level of general media, such as television and radio, reported higher mavenism scores compared to individuals who where heavy consumers of these media (see Table 2).

Health information mavenism & health beliefs

The second aim of our study was to ascertain health beliefs of mavens to determine whether or not mavens maintained more favorable health beliefs than non-mavens since health beliefs are influential in shaping health behaviors and because mavens are likely to be key sources of health information for others. We used national recommendations and established medical evidence about nutrition, physical activity and smoking as standards for comparison. Overall few participants maintained beliefs regarding nutrition and physical activity that were aligned with the CDC s national recommendations. Only 24% of C2C participants believed that individuals should eat 5–8 servings or more of fruit and vegetables daily to maintain good health. And, less than 60% of participants felt that individuals should engage in physical activity or exercise most days of the week to maintain good health (see Table 3). C2C participants beliefs regarding the effects of cigarette smoke on health were more aligned with well-established evidence in that slightly over 80% agreed that smoking has been proven to cause cancer and that secondhand smoke can cause lung cancer in non-smokers (see Table 4). In specific regards to our research question, we did find that health information mavens (30%) were more likely to have concordance with the CDC s recommendation for fruit and vegetable intake than non-mavens (20%). However, there were no other differences in health beliefs between mavens and non-mavens in bivariate analyses.

Table 3
Association between health information mavenism and physical activity and nutrition beliefs
Table 4
Association between health information mavenism and smoking/cancer beliefs

Multivariable results

Characteristics of health information mavens

Many of the significant social, media and demographic factors of health information mavenism in bivariate analyses remained significant in the adjusted multivariable model. Being highly networked within one s community as measured by scoring high on the social network scale was predictive of a 1.58 increase in mavenism compared to those who scored lower on the social network scale after controlling for other media and social demographic influences (see Table 2). On average, being female was predictive of a 1.42 unit increase in mavenism compared to being male. One of the most significant independent predictors of health information mavenism was having medium level language acculturation compared to those with higher levels of acculturation. After controlling for the influence of nativity and duration of living in the US along with other demographic and media-related factors, having medium-level language acculturation was associated with a 2.38 unit increase in mavenism score, respectively, compared to those having high-level language acculturation. (see Table 2).

Exposure to health segments of newspapers or magazines as well as incidental health information exposure via the Internet remained significant independent predictors of mavenism in the multivariable model; however exposure to health information on local TV news did not retain its significance in the model. On average, reading health segments of either newspapers or general magazines once or more per week was predictive of a 2 unit increase in mavenism score compared to never reading such segments. Similarly, any incidental exposure to health information on the Internet (once or less than once per week) was associated with a 2 unit increase in mavenism score compared to no such exposure, even after controlling for important demographic factors such as race/ethnicity and language. Interestingly, attention to health-specific media was not related to health information mavenism regardless of channel nor was health information seeking (see Table 2). In relation to general media consumption, moderate exposure to television remained a significant predictor of mavenism in the multivariable model and was associated with a 1.43 unit increase in mavenism score compared to heavy exposure.

Health information mavenism & health beliefs

In multivariable analyses, health information mavens were no more likely to maintain health beliefs that were aligned with national recommendations and established medical evidence than non-mavens for three of the four outcome variables of interest. Health information mavens had a 64% higher odds of believing that individuals should eat five or more servings of fruits and vegetables a day to maintain good health compared to non-mavens (see Table 3). However, mavens did not have higher odds of believing that individuals should exercise most days of the week for good health, that smoking has been proven to cause cancer or that secondhand smoke can cause lung cancer in non-smokers compared to non-mavens (see Tables 3 and and44).

Multivariable analyses also identified other media environment and socio-demographic characteristics that were associated with favorable health beliefs. Participants who were either born in the US or who ve been in the US for more than ten years along with participants who reported reading a newspaper on most days of the week (4 or more) had significantly higher odds of maintaining beliefs about physical activity that are aligned with national recommendations (see Table 3). US nativity and length of stay in the US were also associated with higher odds of agreeing that there is a link between cigarette smoking and risk of cancer; but, there were inconsistent trends between health-specific media exposure and beliefs about smoking and cancer (see Table 4). Meanwhile, higher language acculturation as well as being a former smoker were associated with higher odds of maintaining beliefs in concordance with the national recommendation for fruits and vegetables and agreeing that secondhand smoke can cause lung cancer in non-smokers (see Tables 3 and and44).


Both social and media environments are influential in determining health information mavenism. As would be expected based on the noted link between communication and social capital, one s social network is an important attribute of health information mavenism (Ackerson & Viswanath, 2009; Viswanath, 2008). Our data suggest that mavens are highly networked within their communities, reporting larger social networks and more civic engagement than those scoring lower on the maven scale. This integration within their communities allows mavens the opportunity for quick and easy dissemination of health information across members of their networks.

Another indication that social environment or context may play an important role in health information mavenism is that lower language acculturation was a significant predictor of higher mavenism scores. There is a growing body of research that suggests a high percentage of urban immigrants with limited English proficiency reside within what are considered to be “immigrant enclaves” or rather communities/neighborhoods with a high ethnic density (Dubowitz, Subramanian, Acevedo-Garcia, Osypuk, & Peterson, 2008; Osypuk, Bates, & Acevedo-Garcia, 2010; Osypuk, Roux, Hadley, & Kandula, 2009). Enclaves have been shown to have many protective benefits for their residents including easing communication barriers and challenges (Fernandez Kelly & Schauffler, 1996). By extension, it is likely that our respondents with low to medium language acculturation may potentially reside in such communities and therefore face fewer barriers to receiving and sharing health information with others in their community and social network. Individuals with medium level language acculturation may have the added benefit of being able to effectively communicate in both English and their native language, making them conduits of information, including health information, for their community. On the other hand, C2C participants with higher language acculturation may not reside in such protective communities and therefore face greater communication challenges due to their limited education, literacy skills (relative to their neighbors in non-immigrant communities) and lack of community support. Future research exploring health information mavenism should explicitly measure social contextual factors such as residing in an immigrant enclave and available neighborhood resources in effort to better understand this relationship. Additionally, the relationship among social networks, acculturation and residence and its implication for health information flow warrants greater attention.

In terms of distinguishing media characteristics of health information mavens, our data show that selective exposure to health-related content is more significant than overall media consumption. Increased exposure to television, radio, newspapers or even TV news were shown to have no influence on mavenism scores. In fact, higher than average consumption of television was actually negatively associated with mavenism. On the other hand, exposure to text-based health-specific media sources such as reading health segments of newspapers, general magazines or exposure to health information via the Internet were all positively associated with mavenism scores. These data, along with the fact that health information seeking and attention to health information were not significant determinants of mavenism, indicate that mavens may accumulate their health information somewhat incidentally though their routine use of health-specific sources as opposed to purposively seeking or scanning for health-related content. It is also interesting to note that among a sample with lower education levels and limited literacy skills that reliance on text-based sources including the Internet remain key factors in mavenism. Based on this finding, we argue that social marketers and health communication practitioners should not neglect these channels when attempting to target such audiences, but rather focus attention on developing appropriate low-literacy and accessible information for dissemination through these vehicles.

However, the most compelling results of our study are that health information mavens are no more likely than non-mavens to maintain general health beliefs that are concordant with national recommendations on many of our key outcome variables. Mavens were more likely than non-mavens to maintain accurate beliefs regarding recommended fruit and vegetable intake; however, the overall frequency among mavens is still considerably low with only 30% of mavens maintaining these beliefs, the lowest percentage across all four outcomes. Moreover, mavens were no more likely than non-mavens to maintain beliefs aligned with established medical evidence regarding physical activity and the relationships between smoking, secondhand-smoke and cancer. While it is important to note that these measures may not represent the totality of participants understanding about healthy behaviors they do, however, offer a glimpse of their beliefs and understanding. As such, these results are troubling for several reasons. First, nearly half of the study sample identified as being health information mavens and, in addition, mavenism was associated with being more networked. These results imply that there are a substantial number of individuals who maintain misaligned health beliefs engaging in interpersonal information exchange regarding health matters within their larger social networks. Moreover, our sample represents lower-SEP and racial/ethnic minorities who already carry a disproportionate disease burden compared to the general population. Research shows that beliefs are integral in shaping behaviors in that they greatly influence one s intentions to engage in a particular health behavior (Azjen, 1991; Fishbein & Azjen, 1975). Misaligned beliefs regarding disease prevention and health promotion behaviors may not only contribute but also exacerbate existing health disparities among these vulnerable groups. Further investigation into what type of health information mavens are sharing within their networks based on their health beliefs and attempts to clear misconceptions should be addressed by public health researchers and social marketers alike. Individuals with medium-level language acculturation, or who are bilingual, could be potential targets for efforts to bridge existing divides in that they are more likely to consider themselves health information mavens than those with higher language acculturation and they are more likely to maintain health beliefs in line with national recommendations and supported by medical evidence compared to those with lower language acculturation.

There are some limitations to our study that should be considered when evaluating the results. The cross-sectional nature of the data and analyses limits the understanding of the directionality of the relationships under study; however, neither of our research questions attempted to establish causality, but merely to highlight associations between variables of interest. Our study results also have limited external validity given the defined parameters of our participant sample. Additional research on health information mavenism should attempt to include a population-based representative sample in order to better understand attributes of mavens across levels of socio-economic position, race/ethnicity, place of residence and other social determinants.


Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Contributor Information

Emily Z. Kontos, Harvard School of Public Health, Department of Society, Human Development and Health, 401 Park Drive. Room 403F, Boston, MA 02215, Phone: 617.384.8724.

Karen M. Emmons, Harvard School of Public Health, Department of Society, Human Development and Health, Dana-Farber Cancer Institute, Medical Oncology.

Elaine Puleo, University of Massachusetts, Amherst, School of Public Health and Health Sciences.

K. Viswanath, Harvard School of Public Health, Department of Society, Human Development and Health, Dana-Farber Cancer Institute, Medical Oncology.


  • Abratt R, Nel D, Nezer C. Role of the market maven in retailing: a general marketplace influencer. Journal of Business and Psychology. 1995;10(1):31–56.
  • Ackerson L, Viswanath K. The social context of interpersonal communication and health. Journal of Health Communication. 2009;14(Suppl 1):5–17. [PubMed]
  • American Legacy Foundation. American Smoking and Health Survey (ASHES) Washington, D.C: 2005.
  • Arndt J. Role of product-related converstations in the diffusion of a new product. Journal of Marketing Research. 1967;4:291–295.
  • Azjen I. The Theory of planned behavior. Organizational Behavior and Human Decision Proceses. 1991;50:179–211.
  • Bayns B. Word of mouth: the indirect effects of marketing efforts. Journal of Advertising Research. 1985;4:291–295.
  • Bennett G, Wolin K, Okechukwu C, Arthur C, Askew S, Sorensen G, et al. Nativity and cigarette smoking among lower income blacks: results from the Healthy Directions Study. Journal of Immigrant and Minority Health. 2008;10(4):305–311. [PMC free article] [PubMed]
  • Berkman L. Social networks, support, and health: taking the next step forward. Am J Epidemiol. 1986;123(4):559–562. [PubMed]
  • Berkman L, Glass T. Social integration, social networks, social support, and health. In: Berkman L, Kawachi I, editors. Social Epidemiology. New York: Oxford University Press; 2000.
  • Bonds D, Foley K, Dugan E, Hall M, Extrom P. An exploration of patients’ trust in physicians in training. Journal of Health Care for the poor and Underserved. 2004;15(2):294–306. [PubMed]
  • Brundage MD, Feldman-Stewart D, Tishelman C. How do interventions designed to improve provider-patient communication work? Illustrative applications of a framework for communication. Acta Oncol. 2010;49(2):136–143. [PubMed]
  • Bureau of Labor Statistics. American Time Use Survey. Washington, DC: United States Department of Labor; 2009.
  • Cheong PH. Health communication resources for uninsured and insured Hispanics. Health Communication. 2007;21(2):153–163. [PubMed]
  • Cotten S, Gupta S. Characteristics of online and offline health information seekers and factors that discriminate between them. Social Science and Medicine. 2004;59:1795–1806. [PubMed]
  • Dubowitz T, Subramanian SV, Acevedo-Garcia D, Osypuk TL, Peterson KE. Individual and neighborhood differences in diet among low-income foreign and U.S.-born women. Womens Health Issues. 2008;18(3):181–190. [PMC free article] [PubMed]
  • Feick L, Price L. The market maven: a diffuser of marketplace information. Journal of Marketing. 1987;51:83–97.
  • Fernandez Kelly M, Schauffler R. Divided fates: immigrant children and the new assimilation. New York: Russell Sage Foundation; 1996.
  • Fishbein M, Azjen I. Belief, Attitude, Intention, and Behavior: An Introduction to Theory and Research. Reading, MA: Addison-Wesley; 1975.
  • Fisher AA. The measurement and utility of the opinion leadership concept for family planning programs. Health Educ Monogr. 1975;3(2):168–180. [PubMed]
  • Gansler T, Henley SJ, Stein K, Nehl EJ, Smigal C, Slaughter E. Sociodemographic determinants of cancer treatment health literacy. Cancer. 2005;104(3):653–660. [PubMed]
  • Klepp KI, Halper A, Perry CL. The efficacy of peer leaders in drug abuse prevention. J Sch Health. 1986;56(9):407–411. [PubMed]
  • Kontos EZ, Bennett GG, Viswanath K. Barriers and facilitators to home computer and Internet use among urban novice computer users of low socioeconomic position. Journal of Medical Internet Research. 2007;9(4) [PMC free article] [PubMed]
  • Kreps GL, Arora NK, Nelson DE. Consumer/provider communication research: directions for development. Patient Educ Couns. 2003;50(1):3–4. [PubMed]
  • Lorence D, Park H. Study of education disparities and health information seeking behavior. CyberPsychology and Behavior. 2007;10(1):149–151. [PubMed]
  • Makoul G. The interplay between education and research about patient-provider communication. Patient Educ Couns. 2003;50(1):79–84. [PubMed]
  • National Cancer Institute. HINTS Questions. 2010 Retrieved April, 2010, from
  • Osypuk TL, Bates LM, Acevedo-Garcia D. Another Mexican birthweight paradox? The role of residential enclaves and neighborhood poverty in the birthweight of Mexican-origin infants. Soc Sci Med. 2010;70(4):550–560. [PMC free article] [PubMed]
  • Osypuk TL, Roux AV, Hadley C, Kandula NR. Are immigrant enclaves healthy places to live? The Multi-ethnic Study of Atherosclerosis. Soc Sci Med. 2009;69(1):110–120. [PMC free article] [PubMed]
  • Perry CL, Klepp KI, Halper A, Hawkins KG, Murray DM. A process evaluation study of peer leaders in health education. J Sch Health. 1986;56(2):62–67. [PubMed]
  • Risker DC. A segmentation analysis of consumer uses of health information. Health Marketing Quarterly. 1995;12(4):39–48. [PubMed]
  • Rogers EM. A prospective and retrospective look at the diffusion model. J Health Commun. 2004;9(Suppl):13–19. [PubMed]
  • Rudd R, Moeykens BA, Colton TC. Health and Literacy: A Review of Medical and Public Health Literature. In: Comings J, Garners B, Smith C, editors. Annual Review of Adult Learning and Literacy. New York: Jossey-Bass; 2000.
  • Smith S, Dixon A, Treena L, Nutbeam D, McCaffery K. Exploring patient involvement in healthcare decision making across different education and functional health literacy groups. Social Science and Medicine. 2009;69:1805–1812. [PubMed]
  • Sorensen G, Barbeau E, Stoddard A, Hunt M, Kaphingst K, Wallace L. Promoting behavior change among working-class, multiethnic workers: results of the healthy directions--small business study. American Journal of Public Health. 2005;95(8):1389–1395. [PubMed]
  • Teutsch C. Patient-doctor communication. Med Clin North Am. 2003;87(5):1115–1145. [PubMed]
  • The Centers for Disease Control and Prevention. Behavioral Risk Factor Surveillance System. Office of Surveillance, Epidemiology, and Laboratory Services; 2010a.
  • The Centers for Disease Control and Prevention. Fruits & Veggies-More Matters TM. 2010b Retrieved February, 2010, from
  • The Centers for Disease Control and Prevention. The National Health Interview Survey. 2010c
  • The Centers for Disease Control and Prevention. Physical Activity for Everyone: guidelines for adults. 2010d Retrieved February, 2010, from
  • The Centers for Disease Control and Prevention. Division of Nutrition, Physical Activity and Obesity; 2010e. Programs and Campaigns. from
  • The Nielsen Company. How US Adults Use Radio and Other Forms of Audio: Results from the Council for Research Excellence Video Consumer Mapping Study. New York: 2009.
  • Valente TW, Pumpuang P. Identifying opinion leaders to promote behavior change. Health Educ Behav. 2007;34(6):881–896. [PubMed]
  • Vanderpool RC, Kornfeld J, Finney Rutten L, Squiers L. Cancer information-seeking experiences: the implications of Hispanic ethnicity and Spanish language. Journal of Cancer Education. 2009;24:141–147. [PMC free article] [PubMed]
  • Viswanath K. The communications revolution and cancer control. Nature reviews cancer. 2005;5(10):828–835. [PubMed]
  • Viswanath K. Public communications and its role in reducing and eliminating health disparities. In: Thomson GE, Mitchell F, Williams MB, editors. Examining the health disparities research plan of the National Institutes of Health: Unfinished business. Washington D.C: Institute of Medicine; 2006. pp. 215–253.
  • Viswanath K. Social capital and health communications. In: Kawachi I, Subramanian S, Daniel K, editors. Social Capital and Health. New York: Springer; 2008. pp. 259–271.
  • Viswanath K, Breen N, Meissner H, Moser RP, Hesse B, Steele WR, et al. Cancer knowledge and disparties in the information age. Journal of Health Communication. 2006;11(Suppl 1):1–17. [PubMed]