Seeking and sharing health information on social media: A net valence model and cross-cultural comparison

TFS-18601; No of Pages 13 Technological Forecasting & Social Change xxx (2016) xxx–xxx

Contents lists available at ScienceDirect

Technological Forecasting & Social Change

Seeking and sharing health information on social media: A net valence model and cross-cultural comparison Yibai Li a, Xuequn Wang b,⁎, Xiaolin Lin c, Mohammad Hajli d a

Kania School of Management, University of Scranton, Scranton, PA 18505, USA School of Engineering and Information Technology, Murdoch University, 90 South St., Perth, Western Australia 6150, Australia Department of Decision Sciences and Economics, College of Business, Texas A&M University-Corpus Christi, Corpus Christi, TX 78412, USA d Institute for Interdisciplinary Cultural Studies, Allameh Tabataba'i University, Dehkadeh-ye-Olympic, Tehrān, Iran b c

a r t i c l e

i n f o

Article history: Received 30 July 2015 Received in revised form 8 July 2016 Accepted 12 July 2016 Available online xxxx Keywords: Health information Social network sites Social media Net valence model

a b s t r a c t In the past few years, social media has changed the way people seek and share health information. However, despite its significant advantages, social media still faces many challenges in user adoption and participation regarding health information. This study focuses on the factors that affect users' intentions to seek and share health information on social media. A net valence model was developed based on social support theory and prior e-service adoption research. Two studies, one in China and the other in Italy, were conducted to test the model. The results show that the proposed net valence model can effectively explain users' intentions to seek and share health information on social media. The results also show important cultural differences. An extensive literature review reveals that this study is among the first to investigate the non–healthcare professionals' intentions to seek and share health information in the context of social media using cross-culture samples. © 2016 Elsevier Inc. All rights reserved.

1. Introduction In the past few years, social media such as YouTube, Facebook, and Twitter has drastically changed the landscape of the health care industry. It has profoundly changed the ways in which health care providers deliver services. More than 1,500 hospitals in the U.S. now have an online presence on social media (Honigman, 2013). Nearly 5,000 U.S. health organizations (e.g., hospitals, drug companies, and health insurers) have accounts on social media platforms (Allied Health World, 2012). Social media serves as a type of “concierge” practice that can give quick answers to patients' questions, make appointments, or facilitate follow-up discussions. It greatly saves patients' time and improves the quality of health providers' services. In a prior survey (Allied Health World, 2012), 49% of polled users expected to hear from their doctors within a few hours, and 60% of doctors said social media had improved the quality of care delivered to patients. The Mayo Clinic's podcast audience rose by 76,000 after the Clinic started using social media. Social media has many advantages in providing health care information. Using social media, health care providers can post health information not only in text but also in more easily accessible forms, such as

⁎ Corresponding author. E-mail addresses: [email protected] (X. Wang), [email protected] (X. Lin), [email protected] (M. Hajli).

images and videos, which can be retrieved at any time of day. Social media empowers health care consumers, providing them with immediate access to an incredible amount of health information and a variety of perspectives on health topics. The information and knowledge that used to be exclusive to health care providers has now become available to all social media users (Ker et al., 2014; Lambert and Loiselle, 2007). In addition, compared to Web 1.0 technologies such as static websites, social media not only facilitates health information-seeking activities but also allows users to share health information. Users can share their health care knowledge, experiences, and symptoms, and they can post reviews about health products, medicines, and doctors. In fact, of those who sought health information on social media, 40% also shared their personal health experiences (Fox and Duggan, 2013). These reciprocal seeking and sharing activities bond patients who have similar health concerns, and they can form online support groups or self-help groups. Prior studies have shown that these health groups can provide patients with informational, emotional, and social support as well as help them cope with their illnesses (Maloney-Krichmar and Preece, 2005). However, despite its significant advantages in providing health care information, social media still faces many challenges in user adoption and participation regarding health information (Antheunis et al., 2013). In contrast to traditional health care services in which the value is primarily created by health care organizations, the health care services provided on social media rely on collaboration among users (Möller et al., 2008). The success of health care information services on social media requires users to actively participate in not only seeking

http://dx.doi.org/10.1016/j.techfore.2016.07.021 0040-1625/© 2016 Elsevier Inc. All rights reserved.

Please cite this article as: Li, Y., et al., Seeking and sharing health information on social media: A net valence model and cross-cultural comparison, Technol. Forecast. Soc. Change (2016), http://dx.doi.org/10.1016/j.techfore.2016.07.021

2

Y. Li et al. / Technological Forecasting & Social Change xxx (2016) xxx–xxx

health information but also sharing it. Therefore, what factors affect users' intentions to seek and share health information on social media becomes the focus of this study. Many prior studies have investigated the antecedents of informationseeking and sharing behaviors in a general business context. However, few studies have examined the antecedents of these two behaviors on social media in the health care context. The context of health care differs significantly from a general business context because health care consumers are more sensitive to potential risks associated with health-related decisions (Bansal et al., 2010). Before seeking or sharing health information on social media, health care consumers have to ensure that the benefits of the behavior outweigh the risks (Bansal et al., 2010). Therefore, the net valence model (NVM) is an appropriate model for this study. The NVM suggests that people will engage in an activity if the associated benefits outweigh the corresponding costs (Fishbein, 1967; Lewin et al., 1944). The NVM has proven to be a powerful model in explaining the adoption of eservices (Featherman and Pavlou, 2003; Featherman and Wells, 2004; Featherman et al., 2006, 2010), but few studies have used it to explain health information–seeking and sharing behaviors on social media. Therefore, our research question: how does net valence model explain users' intentions to seek and share health information on social media? The NVM is also consistent with the cost-benefit view of customer value. According to this view, customers assess the value based on what customers receives (e.g., benefits) and the costs what they have to give up (e.g., sacrifices) (Brodie et al., 2009; Woodruff, 1997). Thus, the NVM can assess how a trade-off is involved when people make their decision to seek and share health information on social media. To develop the NVM, we conducted an extensive literature review on related research areas. Drawing from social support theory (Cobb, 1976), we identified such benefit factors as emotional support and informational support. From prior e-service adoption research (Featherman and Pavlou, 2003; Hajli et al., 2015), we identified such benefit factors as credibility and perceived usefulness as well as risk factors such as mental intangibility, privacy risk, time risk, social risk, and psychological risk. Two studies were conducted to test the NVM. One was conducted in China, the other in Italy. According to Hofstede et al. (2010), Italians have a high level of uncertainty avoidance, whereas Chinese culture emphasizes low uncertainty avoidance. We chose these two countries to examine whether this cultural difference impacts users' intentions to seek and share health information on social media. This study has two significant contributions. First, this study is among the first to apply the NVM to investigate the non–healthcare professionals' intentions to seek and share health information in the context of social media. This study also shows how factors from the social support theory and prior e-service adoption models play a role, through perceived benefits and perceived risks, in influencing users' intentions to seek and share health information on social media. Second, this study is among the first cross-cultural studies to investigate users' intentions to seek and share health information on social media using the theories mentioned above. The rest of the paper is organized as follows: We first extensively review the literature of health information-seeking and sharing behavior. We then discuss the theoretical foundation on which the hypotheses are based, and present the method of the study. Finally, implications, limitations, and opportunities for future studies, are discussed. 2. Literature review Prior studies have discussed a wide range of social media health care applications (Boulos et al., 2007; Fox and Duggan, 2013; Vance et al.,

2009). Popular social media health care applications include wikis, blogs, podcasting, streaming video services, social bookmarking, collaborative tagging (folksonomies) and tag clouds, and many others (Boulos et al., 2007). These applications have been widely used for sharing and seeking health information by not only patients, health seekers, and lay users but also health care professionals. Based on the user types and their roles, four categories of behaviors are identified as shown in Fig. 1. Health care professionals include people such as doctors, nurses, social workers, psychologists, counselors, chaplains, volunteers, trainers, and trainees who have gone through health care training and have been socialized into the health care profession (Buckman, 1992). Social media has been widely used by health care professionals for a variety of reasons. Two-thirds of doctors use social media for professional purposes (Honigman, 2013), and 60% of physicians use social media to follow what their colleagues are sharing and discussing (Allied Health World, 2012). Social media is also used for training, exchanging health records, and facilitating speedy communication and collaboration among clinicians (Boulos et al., 2007). The other type of users, the non–healthcare professionals, includes patients, health seekers, and lay users. Lay users “are those who have not gone through the training or socialization into the particular profession (such as medicine, nursing, chiropractic) which we refer to as the index profession.” (Cifter and Dong, 2009, p. 2). Social media provides a wide variety of bidirectional communication tools that connect patients who have similar health concerns. Patients who have chronic diseases, disabilities, or cancers or are recovering from surgeries may find social media particularly useful. Health-related topics that people frequently discuss on social media include specific diseases or medical problems, certain medical treatments or procedures, weight control, health insurance, food safety or recalls, drug safety or recalls, advertised drugs, medical test results, aging, pregnancy, childbirth, and health care costs (Fox and Duggan, 2013). Through an extensive literature review, we find that the research of health care on social media has not kept up with its applications. Specifically, few studies have examined patients', health seekers', or lay users' (shown in Fig. 1) intentions to seek and share health information on social media. Here health information-seeking is the search for and receipt of messages that help “to reduce uncertainty regarding health status” and “construct a social and personal (cognitive) sense of health.” (Tardy and Hale, 1998, p. 338). The primary rationale for searching for health information is to alleviate uncertainty about health decisions (Lambert and Loiselle, 2007). Health information–seeking behavior is an important component of coping with illness and health-related uncertainty (Lambert and Loiselle, 2007). We then extend the literature search to related topics in other contexts, and find that health information-seeking behavior has been studied long before the Internet era. Before the Internet, physicians held almost exclusive access to health care information. Apart from health care providers, the public's sources for health information were limited to local experts and the mass media (Kassulke et al., 1993). In the late 1990s, however, the Internet quickly became a major source of health information (Cotten and Gupta, 2004). The Internet enjoys many advantages in providing health information. First, it is an immediate, convenient, and comprehensive source of health information. Second, it is anonymous. It allows health seekers to ask awkward, sensitive, or detailed questions without the risk of facing judgment, scrutiny, or stigma. Third, it decreases the inequalities associated with health care provision and decision making. In the early 2000s, the emergence of social media further changed how the public could seek health information, introducing new possibilities to this research. In addition to all the advantages of the Internet, social media provides a wide variety of bidirectional communication tools that connect health seekers who have similar health concerns.

Please cite this article as: Li, Y., et al., Seeking and sharing health information on social media: A net valence model and cross-cultural comparison, Technol. Forecast. Soc. Change (2016), http://dx.doi.org/10.1016/j.techfore.2016.07.021

Y. Li et al. / Technological Forecasting & Social Change xxx (2016) xxx–xxx

3

Fig. 1. Participants and roles of social media health care service.

The research of offline and Internet health information–seeking behavior has extended from the individual decision-making perspective to group dynamics and sociality perspective. The antecedents of the health information-seeking behavior that have been studied include individual factors and contextual factors. The individual factors include health information need (Lambert and Loiselle, 2007); sociodemographic characteristics; and individuals' personality traits, expectations, goals, beliefs, values, attitudes, emotions, moods, education, skills, and resources (Borgers et al., 1993). The contextual or situational factors include the characteristics of individuals' environments, sources of information, and informationseeking contexts; trust; and relationships (Allen, 1996). Theoretical discussions on the health seekers' health information–seeking behaviors have been summarized in Lenz (1984), Johnson (1997); Longo (2005); Lambert and Loiselle (2007). In contrast to the research of health information-seeking behavior, the research of health information sharing behavior was scarce before the year 2000. This is probably because before Web 2.0 and social media, there was hardly an option for non–healthcare professionals to share health information. After the year 2000, we find that a few studies begin to examine this topic, but theories and more empirical evidences are needed in this area. A summary of the empirical studies on this topic (i.e., the purposes of the study, theories, antecedents, focal factors, methods and findings) is provided in the Appendix A. Based on our review, we found that few studies have examined non-healthcare professionals' intentions to share health information in social media. Besides, our review also finds that few studies have examined health seekers' behaviors across different cultural contexts. According to Hofstede et al. (2010), there are five dimensions of national cultures: power distance, uncertainty avoidance, individualism/collectivism, masculinity/feminity, and long/short term orientation. Uncertainty avoidance refers to the degree to which “individuals feel threatened by, and try to avoid, ambiguous situations by establishing more formal rules and rejecting deviant ideas and behaviors” (McCoy et al., 2005, p. 212). Uncertainty avoidance is relevant in the context of online health services because online health services are uncertain in nature, and there is no guaranteed that the online health information is of good quality. People can experience various kinds of risk when dealing with online health information. Therefore, it is important to examine how people with high vs. low level of uncertainty avoidance make their decisions to seek/share health information online. To summarize, our review confirms that there have been few studies to examine non-healthcare professionals' intentions to seek

and share health information across different cultural contexts in the context of social media. The following section will explain more about theoretical underpinnings and describe the hypotheses development process. 3. Theoretical foundation and hypotheses development 3.1. Theoretical foundation Using technology can have various risks (Weeger et al., 2015a). In the context of online health services, people may receive false information and misleading advices, leading to negative impacts on their health. To consider both positive and negative factors, the NVM is proposed to understand people's decisions. The basic argument of the NVM is that people would engage in an activity if the associated benefits outweighed the corresponding costs (Fishbein, 1967; Lewin et al., 1944). Previous literature has used the NVM to examine the adoption of Internet banking (Lee, 2009), bring-yourown-device services (Weeger et al., 2015b), and social network sites (Li and Wang, in press). Following the NVM, we suggest that to use online health services, individuals need to perceive that the associated benefits outweigh the costs. Here, perceived benefits reflect the overall positive utility of the health information on social media. Building on previous literature on social support theory (Cobb, 1976) and e-service adoption (Hajli, 2014; Hajli et al., 2015), we propose that in the context of social media health care applications, the antecedents of perceived benefits include perceived usefulness and credibility of health information on social media as well as emotional and informational support from social media health communities. Our selections are mainly based on the context of our study: social media health communities. Other factors in previous literature, such as cost advantage in the context of outsourcing (Gewald and Dibbern, 2009), are not relevant for the context of this study. Previous literature has conceptualized cost through risk perception. Risk perception is the subjective expectation of a possible loss, and it can negatively affect people's intentions to engage in particular behaviors. Peter and Tarpey (1975) identify six dimensions of perceived risk regarding consumer decision strategies: financial, performance, psychological, physical, social, and time risk. Building on previous literature on e-service adoption (Featherman and Pavlou, 2003; Featherman and Wells, 2010), we propose that the antecedents of perceived risks associated with the health information on social media consist of mental

Please cite this article as: Li, Y., et al., Seeking and sharing health information on social media: A net valence model and cross-cultural comparison, Technol. Forecast. Soc. Change (2016), http://dx.doi.org/10.1016/j.techfore.2016.07.021

4

Y. Li et al. / Technological Forecasting & Social Change xxx (2016) xxx–xxx

intangibility, privacy, time, and social and psychological risk. Here mental intangibility is added to reflect the ambiguous nature of online environment; privacy risk is added to reflect people's concern over their private information. Financial and physical risks are not relevant for the context of this study and are thus dropped; performance risk is duplicated with perceived usefulness as one antecedent of perceived benefits and is therefore dropped as well. 3.2. Antecedents of perceived benefits In previous technology adoption literature, perceived usefulness—that is, the degree to which people perceive that using certain technology can help them increase job performance—is an important factor influencing people's intentions to adopt certain technologies (Venkatesh et al., 2003). In the context of social media health care applications, we define perceived usefulness of health information on social media as the degree to which people perceive that health information can help them better understand certain health issues. People use social media health care applications to gain better knowledge about health issues. When they perceive that health information is useful, they are more likely to develop a positive attitude toward social media health care applications and continue using them in the future. Therefore, we hypothesize the following: H1a. Perceived usefulness of health information positively relates to perceived benefits. The credibility of online health information is “the believability and trustworthiness of information” (Hajli et al., 2015, p. 239). Credibility of online health information is an important factor because people need to be confident in the quality of health information on social media (Rains and Karmikel, 2009). When people perceive that health information has a high level of credibility, they are more likely to trust social media health care applications and continue to use them. In the context of online health community, Hajli et al. (2015) find that credibility is important to attract people to search for health information online. In other words, when consumers feel online health information has a high level of credibility, they benefit from the online health community as a source of information. Therefore, we argue the following: H1b. Credibility of health information positively relates to perceived benefits. Social support is the perceived care, love, and support of members of a group (Cobb, 1976); it includes both emotional aid and informational support (Wellman et al., 1996). Social support is one of the main advantages of using social media health care applications (Maloney-Krichmar and Preece, 2005). Health communities on social media contain a lot of heath information, which can be valuable for people who seek to better understand health issues (Hajli et al., 2013). People can use this social platform to search for health information or receive emotional support (Hajli et al., 2015). Thus, health communities on social media can be suitable places for individuals who may need informational or emotional support (Maloney-Krichmar and Preece, 2005). When people benefit from social support, they are more likely to turn to social media as a source of information and support (Hajli et al., 2015). Therefore, we have the following hypotheses:

people interpret evaluation cues to “ensure a clear, mentally tangible representation of the object” (Laroche et al., 2004, p. 375). Therefore, mental intangibility is an important concept in an online environment where a service is imperceptible and impalpable (Featherman and Wells, 2010). For example, online health services, which are intangible and lack physical evaluation cues, can be quite challenging for people to evaluate and decide whether to adopt (Featherman and Wells, 2010). In such a context, the mental intangibility of social media health care applications can make people perceive that they are risky to use, therefore increasing social media health care applications' perceived risk (Featherman and Wells, 2010). Therefore, we state the following: H2a. Mental intangibility positively relates to perceived risk. Privacy risk is the “potential loss of control over personal information” (Featherman and Pavlou, 2003, p. 455). When using social media health care applications, people may need to discuss health issues in greater detail with doctors or other users to get more relevant health information and advices (Bansal et al., 2010). Therefore, it is possible that their personal information will be disclosed. For example, hackers can break into social media health care applications, steal information, and sell it. Online health communities may also reveal personal information without people's authorization to their employers, leading to discrimination (Rindfleisch, 1997). Because people probably do not wish their personal information to be disclosed, we have the following hypothesis: H2b. Privacy risk positively relates to perceived risk. Time risk is the degree to which people's “assessment of potential losses to convenience, time, and effort caused by wasting time researching, purchasing, setting up, switching to, and learning how to use the e-service” (Featherman and Wells, 2010, p. 114) Because online health communities are often unstructured, it is possible that people will spend time using them but not find relevant health information. Besides, even if people find the needed health information, it is still possible that the information is inaccurate, misleading, or doubtful. Therefore, we state the following: H2c. Time risk positively relates to perceived risk. Social risk is “potential loss of status in one's social group as a result of adopting a product or service, looking foolish or untrendy” (Featherman and Pavlou, 2003, p. 455). Therefore, when social risk is higher, people perceive their images to be lower in their social groups. In the context of online health services, people may believe they will look foolish if others know they are getting health information from social media instead of from doctors. Hence, we propose the following: H2d. Social risk positively relates to perceived risk.

H1c. Emotional support positively relates to perceived benefits.

Psychological risk refers to using certain products or services that may negatively affect people's peace of mind or self-perceptions (Mitchell, 1992). In other words, people may feel frustrated or lose their selfesteem or self-perception (ego) due to using a certain product or service (Featherman and Wells, 2010). Because the quality of health information on social media cannot be guaranteed, people may perceive health information on social media as being invalid and doubt the quality of the information (Bernhardt et al., 2002). Thus, they may feel concerned about the potential negative consequences of applying the health information in their daily lives. Therefore, we argue the following:

H1d. Informational support positively relates to perceived benefits.

H2e. Psychological risk positively relates to perceived risk.

3.3. Antecedents of perceived risks

3.4. Perceived benefits and intention to seek/share health information

To account for the intangibility of information products on the Internet, Laroche et al. (2001) proposed the additional dimension of intangibility called mental intangibility. Mental intangibility is whether

We define perceived benefits as people's assessments of benefits associated with using social media health care applications. Previous literature has argued that people consider perceived benefits when

Please cite this article as: Li, Y., et al., Seeking and sharing health information on social media: A net valence model and cross-cultural comparison, Technol. Forecast. Soc. Change (2016), http://dx.doi.org/10.1016/j.techfore.2016.07.021

Y. Li et al. / Technological Forecasting & Social Change xxx (2016) xxx–xxx

deciding whether to adopt certain technologies (Kim and Olfman, 2011) or technology-related services (Lee, 2009). When the perceived benefits are substantial, people overcome barriers associated with adoption (Porter and Donthu, 2006). Therefore, when people believe that the perceived benefits of online health services are high, they are more likely to seek health information on social media. Besides, when people experience the various benefits of social media health care applications, they may become more likely to keep participating in online health communities and share health information especially when they have knowledge of a specific health topic (Hajli et al., 2015). To summarize, we have the following hypotheses: H3a. Perceived benefits positively relate to people's intentions to seek health information. H3b. Perceived benefits positively relate to people's intentions to share health information.

5

4. Study 1 methods In this section we present data collection procedures, participants, measurements, data analysis and results for our first study. 4.1. Data collection procedures and participants The data was collected via an online survey with snowball sampling. 184 Chinese participants were recruited. However, 28 participants neither shared nor seek health information before, which means that they are not familiar with the context of this study. Therefore, those participants were dropped and we got a total of 156 participants. Here over half (66.0%, n = 103) were from a junior level business class at a large public university of China. The rest (34.0%, n = 53) were referred by student participants. About 1% of their final course credit was rewarded to those students who participated or referred additional participants. Table 1 lists participants' demographic information.

3.5. Perceived risk and intention to seek/share health information

4.2. Measurement

Perceived risk is defined as “the amount that would be lost (i.e., that which is at stake) if the consequences of an act were not favorable, and the individual's subjective feeling of certainty that the consequences will be unfavorable” (Cunningham, 1967, p. 85). In this study, we define perceived risk as people's perception about potential negative outcomes caused by using social media health care applications. When people perceive that there is a high level of risk in using social media health care applications, they probably will not feel comfortable using the services and will feel concerned about the information in online health communities. Thus, they probably would not want to participate in online health communities and share health information as well. Therefore, we argue the following:

Items of our study were adapted from previous literature. Specifically, items for mental intangibility were adapted from Laroche et al. (2004); items for privacy risk, time risk, social risk, psychological risk, perceived risks were based on Featherman and Pavlou (2003); items for perceive usefulness of health related information were adapted from Thong et al. (2002); items for credibility of health related information were adapted from Flanagin and Metzger (2000); items for emotional and informational support were based on Hajli (2014); items for perceived benefits were adapted from Benamati and Rajkumar (2008); items for intention to share and seek health related information in social media were based on Lin and Lu (2011) and Liang et al. (2011). Each question was measured on a 7-point, Likerttype scale. Items for credibility were anchored on 1 = not at all to 7 extremely, while the rest were anchored on 1 = strongly disagree to 7 = strongly agree. The final items used are shown in the Appendix B.

H4a. Perceived risk negatively relates to people's intentions to seek health information. H4b. Perceived risk negatively relates to people's intentions to share health information. Our research model is presented in Fig. 2.

4.3. Analysis and results Our model was tested with partial least squares (PLS), which is appropriate for predictive models and theory building (Barclay et al., 1995; Chin, 2010). SmartPLS (Ringle et al., 2005) was used for the analysis, and the bootstrap re-sampling method (using 5000 samples) was used to determine the significance of the paths in the structural model. We first assessed common method variance (CMV) with two tests (Lindell and Whitney, 2001; Podsakoff et al., 2003). In the first test, an exploratory factor analysis of all items extracted 10 factors explaining 72.77% of the variance. The first factor accounts for 25.11% of the variance, and no single factor accounting for significant loading (at the p b 0.10 level) for any item. Second, the second-smallest positive correlation among the manifest variables was used as a conservative estimate for CMV (Lindell and Whitney, 2001). After adjustment, all significant correlations remained significant. Therefore, CMV is probably not a concern in our study. Data analysis was conducted in two stages: measurement model and structural model assessment. In the first stage, convergent validity was confirmed by meeting the following criteria (Gefen and Straub, 2005;

Table 1 Sample demographic information (Study 1). Category

Gender Fig. 2. Research model.

Age

Female Male

Student sample

“Snowball” sample

(n = 103)

(n = 53)

70 (68.0%) 33 (32.0%) 21.35 (SD 0.9)

25 (47.2%) 28 (52.8%) 32.6 (SD 10.1)

Please cite this article as: Li, Y., et al., Seeking and sharing health information on social media: A net valence model and cross-cultural comparison, Technol. Forecast. Soc. Change (2016), http://dx.doi.org/10.1016/j.techfore.2016.07.021

6

Y. Li et al. / Technological Forecasting & Social Change xxx (2016) xxx–xxx

Table 2 Items and descriptive statistics (Study 1). Items

Mean

S.D.

Loading

CR

AVE

PU1 PU2 PU3 PU4 Cred1 Cred2 Cred3 Cred5 ES1 ES2 ES3 IS1 IS2 IS3 PB1 PB2 MI2 MI3 PR1 PR2 PR3 TR1 TR2 TR3 SR1 SR2 SR3 PsyR1 PsyR2 RISK1 RISK2 SHA1 SHA2 SHA3 SHA4 SEE1 SEE2 SEE3 SEE4 SEE5 SEE6

5.42 5.13 5.13 4.86 4.44 4.12 4.29 3.87 4.94 5.29 5.02 5.22 5.29 5.26 4.55 4.72 4.26 3.40 3.90 4.36 5.02 3.60 4.01 3.29 2.83 2.51 2.83 2.63 3.52 3.54 3.37 4.56 4.15 4.80 5.07 5.25 4.85 5.33 5.53 5.47 5.07

1.00 1.05 0.99 1.03 1.02 0.98 1.03 1.16 0.98 0.94 0.95 1.00 0.90 0.90 1.09 1.04 1.24 1.29 1.66 1.62 1.31 1.42 1.47 1.34 1.18 1.08 1.31 1.18 1.53 1.47 1.33 1.15 1.27 1.21 1.10 0.93 1.08 0.89 0.82 0.95 1.00

0.77 0.87 0.85 0.74 0.90 0.79 0.93 0.73 0.82 0.88 0.82 0.86 0.92 0.92 0.95 0.94 0.72 0.95 0.89 0.92 0.80 0.89 0.83 0.90 0.85 0.90 0.88 0.82 0.85 0.92 0.90 0.92 0.88 0.85 0.69 0.85 0.81 0.87 0.70 0.73 0.69

0.88

0.66

0.91

0.71

0.88

0.70

0.93

0.81

0.94

0.89

0.83

0.71

0.90

0.76

0.91

0.76

0.91

0.77

0.82

0.70

0.91

0.83

0.90

0.70

0.90

0.61

Hulland, 1999): First, each item loaded significantly on their respective constructs and none of the loadings were below the 5 value of 0.50 (Table 2). Second, the composite reliabilities (CRs) of all constructs were above 0.70 (Table 2). Finally, the average variance extracted (AVE) from all constructs was above the threshold value of 0.50 (Tables 2). Discriminant validity was established by ensuring that the correlations between constructs were below 0.85 (Brown, 2006) and that the square root of AVE for each construct exceeded all correlations between that construct and any other construct (Gefen and Straub, 2005) (Table 3). Overall, our measures demonstrated good psychometric properties. In the second stage, we assessed the structural model with R2 measures and path coefficients. R2 can show the amount of variance in the dependent variable explained by independent variables (Gefen et al., 2000), and path coefficients indicate the significance of relationships between constructs. H1a, stating that perceived usefulness of health related information is positively associated with perceived benefits, was supported (b = 0.39, p b 0.001). H1b argues that creditability of health information is positively associated with perceived benefits. This hypothesis was not supported (b = 0.05, p N 0.05). H1c posits that emotional support is positively related to perceived benefits. This hypothesis was not supported (b = 0.11, p N 0.05). H1d states that informational support is positively associated with perceived benefits. This hypothesis was supported (b = 0.21, p b 0.05). H2a states that mental intangibility is positively associated with perceived risk. This hypothesis was not supported

(b = −0.04, p N 0.05). H2b argues that privacy risk is positively associated with perceived risk. This hypothesis was supported (b = 0.15, p b 0.05). H2c states that time risk is positively associated with perceived risk. This hypothesis was supported (b = 0.15, p b 0.05). H2d states that social risk is positively associated with perceived risk. This hypothesis was not supported (b = −0.02, p N 0.05). H2e argues that psychological risk is positively associated with perceived risk. This hypothesis was supported (b = 0.61, p b 0.001). H3a, stating that perceived benefits are positively associated with intention to seek health information in social media, was supported (b = 0.49, p b 0.001). H3b argues that perceived benefits are positively related to intention to share health information in social media. This hypothesis was supported (b = 0.41, p b 0.001). H4a posits that perceived risk is negatively related to intention to seek health information in social media. This hypothesis was not supported (b = − 0.10, p N 0.05). H4b argues that perceived risk is negatively related to intention to share health information in social media. This hypothesis was supported (b = − 0.29, p b 0.001). We also examined variance inflation factors (VIF) and all values were below 2.2. Therefore, multicollinearity is not issue. These results are presented in Fig. 31. The Stone-Geisser (Q2) test was then conducted to assess the predictive quality of our model (Geisser, 1975; Stone, 1974). The model has estimation relevance with values of Q2 above 0. The Q2 for perceived usefulness was 0.66; the Q2 for credibility was 0.71; the Q2 for informational support was 0.81; the Q2 for emotional support was 0.70; the Q2 for perceived benefits was 0.89; the Q2 for mental intangibility was 0.71; the Q2 for privacy risk was 0.76; the Q2 for time risk was 0.76; the Q2 for social risk was 0.77; the Q2 for psychological risk was 0.69; the Q2 for perceived risk was 0.83; the Q2 for intention to seek health information was 0.61; and the Q2 for intention to share health information was 0.70. Therefore, our model has good predictive relevance.

4.4. Discussions The results show that perceived usefulness of health information and informational support are positively related to perceived benefits; privacy risk, time risk and psychological risk are positively related to perceived risk. Besides, perceived benefits are positively to intention to seek and share health information, while perceived risk is only negatively related to intention to share health information. Those results overall support our model. On the other hand, our first study can be limited because our sample is from China. Since people from China have a relatively low level of uncertainty avoidance (Hofstede et al., 2010), it is possible that Chinese in general are relatively tolerant toward the risks of online health services, and our results may not hold to other countries where people have a low level of uncertainty avoidance. To address this limitation, we conducted the second study. In this study, we collect our data in Italy where people have a high level of uncertainty avoidance (Hofstede et al., 2010). Since people with a high level of uncertainty avoidance try to avoid uncertainty whenever possible, we expect that Italian would probably feel more concerned about the risks of online health services. In other words, the relationship between relevant factors and perceived risk, and the relationship between perceived risk and intention to seek/share health information are probably stronger for Italian.

1 The purpose of snowball sampling in Study 1 is to recruit additional non-student participants. We performed Box's M test and the result was significant (p b .001). Therefore, the results should be interpreted cautiously. We also conducted the analysis with the student sample only, and most of the results were consistent, except that the relationship between information support and perceived benefits and that between privacy risk and perceived risk became non-significant.

Please cite this article as: Li, Y., et al., Seeking and sharing health information on social media: A net valence model and cross-cultural comparison, Technol. Forecast. Soc. Change (2016), http://dx.doi.org/10.1016/j.techfore.2016.07.021

Y. Li et al. / Technological Forecasting & Social Change xxx (2016) xxx–xxx

7

Table 3 Correlation between constructs and square-root of AVEs (on diagonal) (Study 1). Constructs

1

2

3

4

5

6

7

8

9

10

11

12

13

1 Perceived Usefulness 2 Credibility 3 Emotional Support 4 Informational Support 5 Perceived Benefits 6 Mental Intangibility 7 Privacy Risk 8 Time Risk 9 Social Risk 10 Psychological Risk 11 Perceived Risk 12 Intention to Seek Health Information 13 Intention to Share Health Information

0.81 0.31 0.40 0.42 0.54 0.01 −0.00 −0.21 −0.30 −0.23 −0.20 0.59 0.35

0.84 0.02 0.08 0.18 0.01 0.12 −0.06 −0.11 −0.02 −0.02 0.24 0.07

0.84 0.75 0.42 −0.08 −0.05 −0.06 −0.19 −0.16 −0.14 0.44 0.36

0.90 0.46 −0.01 0.048 −0.10 −0.19 −0.17 −0.04 0.43 0.28

0.94 0.06 −0.22 −0.31 −0.31 −0.25 −0.29 0.52 0.49

0.85 0.29 0.26 0.25 0.21 0.17 0.04 0.07

0.87 0.48 0.43 0.36 0.42 −0.07 −0.23

0.87 0.53 0.41 0.45 −0.26 −0.17

0.88 0.67 0.52 −0.22 −0.23

0.83 0.70 −0.23 −0.27

0.91 −0.24 −0.41

0.78 0.51

0.84

In our second study, participants were from Italy. The measurements are the same from the first study. Below we present data collection procedures, data analysis and results for our second study.

survey, 450 participants dropped out or failed the quality assurance questions. We also removed seven participants who originally came from other places, such as Asia and Africa. Thus, we got a total of 137 people from Italy who finished the survey successfully. The demographic data of the participants is shown in Table 4.

5.1. Data collection procedures and participants

5.2. Data analysis and results

A company named “SurveyCo” (pseudonym) was employed to recruit those Italian who seek health information through social media. SurveyCo employs various channels to recruit a variety of samples, and maintains a system to send out invitation. By filling up surveys, members can earn points, which can be converted to cash, a charitable donation, or gift cards. To eliminate fraudulent or duplicate answers, SurveyCo uses Survey Hub and Relevant ID with quality assurance questions that can detect careless respondents. Screening questions were created to filter out those participants who seldom used social media or those who seldom seek health information in social media. A total of 915 users were invited to the survey. Among those participants, 594 were qualified. While filling out the

SmartPLS was again used to analyze our data. Same tests were again conducted to assess CMV, and results showed that CMV is again not an issue. In first stage of measurement assessment, our measures again demonstrated good psychometric properties by meeting corresponding criteria (Tables 5 and 6). The results of hypotheses testing are shown in Fig. 4. Again, all values of VIF were below 2.2, and multicollinearity is not issue. We also summarize the results of hypotheses testing from two studies in Table 7. Finally, the Stone-Geisser (Q2) test was then conducted to assess the predictive quality of our model, and the Q2 values of all constructs were above 0, indicating good predictive relevance.

5. Study 2 methods

5.3. Cross-cultural comparison We also conducted cross-cultural comparison to examine if there is any difference for path coefficients from two studies. The comparison was conducted following Keil et al. (2000), and the results are shown in Table 8. Most of the relationships between relevant factors and perceived benefits/risk are non-significantly different between Chinese and Italian participants, except for credibility and psychological risk. Specifically, the relationship between credibility and perceived benefits is stronger for Chinese participants, while the relationship between psychological risk and perceived risk is stronger for Italian participants. Besides, the relationship between perceived benefits and intention to seek/share health information is stronger for Italian participants. Finally, the relationship between perceived risk and intention to share health information is stronger for Chinese participants.

Table 4 Sample demographic information (Study 2).

Gender Age

Fig. 3. Model Results with Chinese Sample (Study 1).

Category

Number of participants

Female Male 22 or below 23 – 28 29 - 35 36 - 45 Above 45

78 (56.9%) 59 (43.1%) 12 (8.8%) 15 (10.9%) 23 (16.8%) 32 (23.4%) 55 (40.1%)

Please cite this article as: Li, Y., et al., Seeking and sharing health information on social media: A net valence model and cross-cultural comparison, Technol. Forecast. Soc. Change (2016), http://dx.doi.org/10.1016/j.techfore.2016.07.021

8

Y. Li et al. / Technological Forecasting & Social Change xxx (2016) xxx–xxx

Table 5 Items and descriptive statistics (Study 2). Items

Mean

S.D.

Loading

CR

AVE

PU1 PU2 PU3 PU4 Cred1 Cred2 Cred3 Cred5 ES1 ES2 ES3 IS1 IS2 IS3 PB1 PB2 MI2 MI3 PR1 PR2 PR3 TR1 TR2 TR3 SR1 SR2 SR3 PsyR1 PsyR2 RISK1 RISK2 SHA1 SHA2 SHA3 SHA4 SEE1 SEE2 SEE3 SEE4 SEE5 SEE6

4.70 4.77 4.74 4.57 4.22 4.07 4.07 4.07 4.64 4.77 4.74 4.93 4.96 4.79 4.37 4.46 4.49 3.36 4.39 4.77 4.82 3.86 3.85 3.82 3.92 3.64 3.78 3.64 3.89 4.01 4.29 4.39 3.97 4.04 4.25 4.50 4.34 4.38 4.83 4.55 4.41

1.21 1.23 1.22 1.26 1.22 1.35 1.19 1.44 1.14 1.20 1.16 1.18 1.09 1.08 1.38 1.31 1.21 1.42 1.62 1.48 1.49 1.42 1.30 1.40 1.56 1.52 1.49 1.49 1.48 1.36 1.39 1.38 1.45 1.38 1.47 1.37 1.34 1.34 1.28 1.38 1.43

0.80 0.88 0.90 0.71 0.92 0.91 0.90 0.85 0.86 0.92 0.91 0.85 0.93 0.89 0.97 0.96 0.89 0.78 0.77 0.81 0.87 0.73 0.93 0.65 0.83 0.91 0.89 0.88 0.82 0.89 0.89 0.92 0.90 0.96 0.92 0.92 0.90 0.92 0.85 0.79 0.81

0.89

0.68

0.94

0.80

0.93

0.81

0.92

0.80

0.96

0.93

0.82

0.70

0.86

0.67

0.82

0.61

0.91

0.77

0.84

0.72

0.88

0.79

6.1. Implications for theory

0.96

0.86

0.95

0.75

This study has several important theoretical implications. First, this study is among the first to have applied the NVM to investigate non– healthcare professionals' intentions to seek and share health information in the context of social media. Prior studies on this topic focused more on health seekers' personal attributes, such as age, education, and individual health conditions. The proposed NVM clearly contributes to the theory of this research area. Second, this study provides empirical evidences that various factors, based upon the social support theory and prior e-service adoption research, can explain non-healthcare professionals' perceived benefit and risks to seek and share health information on social media. The results of this study reveal very interesting findings. Our study shows that most of the factors from these two theories work well to explain the behaviors of interest; however, two factors, mental intangibility and social risk, which prove useful in the e-service adoption, do not fit in the health care and social media context. This is probably because, 20 years after the introduction of all sorts of e-services on the Internet (e.g., online banking, online shopping), e-services are really an integral part of everyone's lives. Average consumers no longer find e-services hard to understand (intangibility) or against the social norm (social

Fig. 4. Model results with Italian sample (Study 2).

6. Discussions This study tries to examine what factors influence people's intentions to seek and share health information on social media. Using the social support theory and e-services adoption models, we developed a theoretical model to identify relevant positive and negative factors. We tested our model with two samples, one from China and one from Italy. The results show that both samples support our model, and there are significant cultural differences. Our study makes important theoretical and practical implications.

Table 6 Correlation between constructs and square-root of AVEs (on diagonal) (Study 2). Constructs

1

2

3

4

5

6

7

8

9

10

11

12

13

1 Perceived Usefulness 2 Credibility 3 Emotional Support 4 Informational Support 5 Perceived Benefits 6 Mental Intangibility 7 Privacy Risk 8 Time Risk 9 Social Risk 10 Psychological Risk 11 Perceived Risk 12 Intention to Seek Health Information 13 Intention to Share Health Information

0.83 0.37 0.43 0.52 0.64 0.04 0.13 −0.01 −0.01 −0.16 −0.06 0.60 0.52

0.89 0.38 0.25 0.50 0.11 −0.12 −0.14 −0.14 −0.20 −0.24 0.56 0.49

0.90 0.63 0.56 −0.01 0.05 −0.04 −0.02 −0.05 −0.07 0.65 0.58

0.89 0.59 −0.04 0.10 −0.05 −0.03 −0.06 −0.06 0.53 0−.46

0.96 0.08 0.00 −0.01 −0.06 −0.30 −0.19 0.72 0.77

0.84 0.24 0.38 0.15 0.13 0.11 0.08 0.10

0.82 0.55 0.45 0.33 0.38 −0.06 −0.08

0.78 0.52 0.33 0.40 −0.07 −0.01

0.88 0.61 0.50 −0.01 −0.09

0.85 0.55 −0.21 −0.24

0.89 −0.18 −0.14

0.87 0.77

0.93

Please cite this article as: Li, Y., et al., Seeking and sharing health information on social media: A net valence model and cross-cultural comparison, Technol. Forecast. Soc. Change (2016), http://dx.doi.org/10.1016/j.techfore.2016.07.021

Y. Li et al. / Technological Forecasting & Social Change xxx (2016) xxx–xxx Table 7 Summary of hypotheses testing. Hypothesis

Study 1 Study 2 supported? supported?

H1a: Perceived usefulness of health information is positively related to perceived benefits. H1b: Credibility of health information is positively related to perceived benefits. H1c: Emotional support is positively related to perceived benefits. H1d: Informational support is positively related to perceived benefits. H2a: Mental intangibility is positively related to perceived risk. H2b: Privacy risk is positively related to perceived risk. H2c: Time risk is positively related to perceived risk. H2d: Social risk is positively related to perceived risk. H2e: Psychological risk is positively related to perceived risk. H3a: Perceived benefits are positively related to people's intention to seek health information in social media. H3b: Perceived benefits are positively related to people's intention to share health information in social media. H4a: Perceived risk is negatively related to people's intention to seek health information in social media. H4b: Perceived risk is negatively related to people's intention to share health information in social media.

Yes

Yes

No

Yes

No

Yes

Yes

Yes

No

No

Yes Yes No Yes

No No No Yes

Yes

Yes

Yes

Yes

No

No

Yes

No

risk) by any means. Overall, the results of this study should allow researchers to better understand unique aspects of health information– seeking and sharing behaviors in the social media context. Third, this study is among the first cross-cultural studies to investigate users' intentions to seek and share health information on social media. The results of this study reveal very interesting cross-cultural findings. There are several major cultural differences between our Chinese and Italian samples in regard to people's intentions to seek and share health information on social media. Chinese people are in general less willing to either seek or share health information on social media. The effect size of the hypotheses H3a, H3b, H4a, and H4b is smaller in the Chinese sample than in the Italian sample. We also find that Italian people are more sensitive to concerns over the credibility of a social media health service, whereas Chinese people are more sensitive to psychological risk. This is probably because Chinese people may only look for general health information on social media. Therefore, the credibility of the health information on social media does not matter to them. When they have serious health concerns, they prefer to see a doctor in person. This also explains why Chinese people dislike seeking

9

and sharing health information: They think sharing their health concerns with a doctor in person is a better means of addressing their health concerns. The benefits of asking non–healthcare professionals health questions on social media do not outweigh the frustration they may get (psychological risk). Overall, this study opens a door to future cross-cultural studies on health information–seeking and sharing behavior. 6.2. Implications for practice This study also presents important practical implications. There are many health organizations (e.g., hospitals, drug companies and health insurers) with accounts on social media platforms. Evidence shows that social media improves the quality of services these organizations deliver to patients. The success and sustainability of these organizationally sponsored health-related social media platforms require that users actively participate in not only health information–seeking activities but also health information–sharing activities. For those social media platforms, this study gives a very valuable guideline yet a simple truth: that to encourage user participation, the benefits the users get from the services must outweigh the risks. To enlarge the benefits of the services, health organizations need to work on the usefulness and credibility of the health information, emotional support, and informational support. To improve usefulness as well as credibility and informational support, health organizations can work on the immediacy, convenience, comprehensiveness, and quality of the health information on social media. They can provide health information in more user-friendly formats, such as images, videos, and audio. For example, Househ et al. (2014) argue that video sharing is an effective way to deliver health information. To improve emotional support, health organizations should connect users who have similar health concerns as well as encourage users to make friends and build a culture of mutual respect and help in the social media health groups. In the meantime, health organizations need to work on minimizing the privacy, time, and psychological risks. A report shows that 63% of users are concerned that their health information will be shared in public and that 57% worry that their health information might get hacked (Allied Health World, 2012). To minimize users' privacy risk, social media platforms should give privacy trainings to every employee. They should also work on information technology security to ensure that users' health information does not get hacked. To minimize users' time and psychological risk, health organizations should work on the timesaving and convenience features of the social media platform. The health organizations may provide services on social media to allow patients to easily make appointments, to ask quick questions, and so forth.

Table 8 Test of differences between Chinese and Italian participants. Hypothesis

H1a H1b H1c H1d H2a H2b H2c H2d H2e H3a H3b H4a H4b

Construct

Perceived usefulness → Perceived benefits Credibility → Perceived benefits Emotional support → Perceived benefits Informational support → Perceived benefits Mental intangibility → Perceived risk Privacy risk → Perceived risk Time risk → Perceived risk Social risk → Perceived risk Psychological risk → Perceived risk Perceived benefits → Seek health information Perceived benefits → Share health information Perceived risk → Seek health information Perceived risk → Share health information

Chinese participants (n=156)

Italian participants (n=137)

standardized path coefficient

standardized path coefficient

0.394 0.045 0.107 0.209 −0.039 0.146 0.154 −0.020 0.611 0.494 0.405 −0.097 −0.289

0.348 0.251 0.173 0.232 −0.052 0.118 0.160 0.141 0.377 0.712 0.768 −0.045 0.004

T-value

Significant differences?

0.44 1.84⁎ 0.53 0.17 0.13 0.27 0.12 1.13 2.07⁎ 2.57⁎⁎ 4.17⁎⁎⁎

No Yes No No No No No No Yes Yes Yes No Yes

0.50 3.13⁎⁎⁎

⁎ p b 0.05. ⁎⁎ p b 0.01. ⁎⁎⁎ p b 0.001.

Please cite this article as: Li, Y., et al., Seeking and sharing health information on social media: A net valence model and cross-cultural comparison, Technol. Forecast. Soc. Change (2016), http://dx.doi.org/10.1016/j.techfore.2016.07.021

10

Y. Li et al. / Technological Forecasting & Social Change xxx (2016) xxx–xxx

6.3. Limitations and opportunities for future studies

7. Conclusion

Our study also has several limitations. First, we selected social media in general as the context of our study. While our results provide valuable insights on how different factors influence people's seeking and sharing health related information, those results may not be able to generalize to all types of social media. Future studies are needed to examine users' participating behaviors in specific social media platforms (e.g., sponsored vs. non-sponsored health communities). Second, we collect data from two countries with different levels of uncertainty avoidance. However, our results can still be limited, and future studies are needed to test our model in additional countries. Third, while we use online surveys to collect data from both countries, the procedures are not fully consistent. Therefore, the results of this study should be interpreted cautiously. Forth, in our second study, a large portion of participants (40.1%) are 45 or older, which could limit the generalizability of our study. Notwithstanding these limitations, our findings open up a number of interesting opportunities for future research. One opportunity might be to further examine how other social factors can influence people's perceived benefits and risk in specific contexts. For example, some factor may play an important role in a specific context (e.g., guanxi in Chinese culture), and future studies are needed to examine those factors. In this study, we assume that people's seeking and sharing health related information have positive effects, especially from the perspective of health service provider. However, it is possible that health information online is misleading, and users may follow those health information in low quality. Future studies are needed to examine the consequences of people's participating behaviors and how to improve people's positive outcomes.

In the past few years, social media has drastically changed the landscape of the health care industry. It has profoundly affected the ways that health care providers deliver services. It has also changed the ways that people seek and share health information. However, despite its significant advantages, social media still faces many challenges in user adoption and participation regarding health information. The success of health care information services on social media requires users to actively participate in not only the health information–seeking activities but also the health information–sharing activities. This study focuses on the factors affecting users' intention to seek and share health information on social media. Based on social support theory and prior e-service adoption research, we develop the NVM model by integrating both benefit factors (including emotional support, informational support, credibility, and perceived usefulness) and risk factors (including mental intangibility, privacy risk, time risk, social risk, and psychological risk) to investigate users' intentions to seek and share health information in the social media context. Two studies were conducted to test the model. The results show that the proposed model can explain users' health information-seeking and sharing intention on social media effectively. There are also some very interesting cultural differences between the Chinese and Italian samples.

Acknowledgements This research is partially supported by a research fund from University of Scranton (fund number 840792) and the internal grant from the School of Engineering and Information Technology at Murdoch University.

Appendix A. Summary of Literature

References

Purpose

Borgers et al. To develop an educational (1993) communication between cancer outpatients and their specialists Holmes and To examine the importance of comprehensive discharge Lenz instruction (1997) Matthews et al. (2002)

To explore factors affecting medical information-seeking, treatment engagement, and emotional adjustment among African American cancer patients

Czaja et al. (2003)

To examine information-seeking behavior among cancer patients

Warner and To investigate the women's Procaccino health information-seeking behavior and their awareness of (2004) health information resources Szwajcer et al. (2005)

Chou et al. (2009)

Theories

Antecedents

Focal factors

Methods

Findings

Patients' needs, values and beliefs; unexpected situations; patients' skills; and specialists' and companions' behavior Anticipated self-care needs, Adult learning post-discharge information theory, model of information-seeking needs behavior n/a Limited knowledge, misinformation about cancer, mistrust, concerns about privacy, lack of insurance, religious beliefs, and fear and stigma associated with seeking emotional support n/a Contextual and Structural factors, predisposing factors, enabling reinforcing factors

Information-seeking behavior of cancer outpatients

Written questionnaires, focus group interviews Semi-structured interviews and content analysis

Identify factors influencing informationseeking behavior of cancer outpatients Show the importance of supplementing oral and written discharge instruction Identify factors affecting medical information-seeking, treatment engagement, and emotional adjustment

Health information–seeking behavior

Survey

Information Search Process

Health information needs, uncertainty

Health information–seeking behavior

Open-ended survey

Nutrition information sources

Nutrition information–seeking behavior

In-depth face-to-face interviews

Age, education, general health, distress, cancer experience, having health provider

Participating in an online support group, writing in a blog

Survey

n/a

To explore nutrition information- n/a seeking behaviors and motives for seeking nutrition information before and throughout the course of pregnancy n/a To identify the sociodemographic and health-related factors associated with social media users in the United States

Health information–seeking behavior

Focus group Medical information-seeking, interview treatment engagement, and emotional adjustment

Identify the determinants and consequences of information-seeking Women are active health information seekers and generally use the information located Pregnant women perceive pregnancy-specific nutrition information as important Younger age, poorer subjective health, and a personal cancer experience lead to blogging and support group participation

Please cite this article as: Li, Y., et al., Seeking and sharing health information on social media: A net valence model and cross-cultural comparison, Technol. Forecast. Soc. Change (2016), http://dx.doi.org/10.1016/j.techfore.2016.07.021

Y. Li et al. / Technological Forecasting & Social Change xxx (2016) xxx–xxx

11

Appendix A. (continued) (continued) References

Theories

Antecedents

Focal factors

Methods

Findings

To examine the extent to which educated Hispanic women seek out health information, sources used, and the factors that motivated health information-seeking behavior Genuis To investigate how women (2012) position themselves within their accounts of information-seeking, and the influence of positioning on interactions with health professionals Miller and To examine age differences in Bell (2012) the role of trust and ease of search in predicting whether individuals use the Internet to search for health information

Purpose

n/a

Attitudes, beliefs, susceptibility, subjective norms, behavioral control

Health information-seeking

Focus group interview

Educated Hispanic women were not involved in regular health information-seeking

Social positioning theory

Women's autonomous positioning, women's collaborative positioning, women's dependent positioning

Health information seeking

Interviews

Social positioning theory can be useful to examine patient–clinician interaction.

n/a

Age, information trustworthiness, and search challenges

Online health information seeking

Survey

Explores the gender differences in Internet health information-seeking, and the role of age and relationship status, and strategies to improve communication To study Korean Americans' health information-seeking behaviors

n/a

Gender, age, relationship status, strategies to improve communication

Online health information-seeking

Adopters trust Internet health information more than nonadopters; this difference increases in strength with age The gender differences are less than those suggested in previous literature

Structural influence model

Health Sociodemographic, health status, health care access, Cancer information-seeking experience, trust in health information

Suggs et al. (2010)

Hallyburton and Evarts (2014)

Oh et al. (2014)

To explore differences in elderly n/a patients' health disclosures by target (i.e. disclosing to a partner vs. another person) Rowley et al. To understand young people's n/a (2015) evaluation of health information

Checton and Greene (2015)

Taber et al. (2016)

To study the effects of self-affirming in the healthcare context

n/a

Target of the information-sharing, relationship quality and efficacy Authority, style, content, usefulness, brand, ease of use, recommendation, credibility, and verification Self-affirming

ES1

The health related information provided on social media enables me to acquire more useful information about health. Discussing health related information on social media improves my efficiency in obtaining helpful health information. The health related information provided on social media is useful for to understand health. The heath related information provided on social media enables me to make more informed and accurate decisions regarding health.

ES2

Credibility of health information (Flanagin and Metzger, 2000) To each degree do you rate the health related information provided on social media? 1 “not at all” to 7 ”extremely”

IS3

Cred1 Cred2 Cred3 Cred5

PB1

PU1

PU2 PU3 PU4

believability accuracy trustworthiness completeness

Immigration status influences Korean Americans' health information-seeking behaviors Survey Significant differences exist between the two groups in perceived support Questionnaire-based Develop the scale of survey the trust in online health information

Elderly patients' disclosing of heart-related conditions Trust judgement, online health information seeking Information-seeking, Survey perception of providers and health care communication, engagement in medical research

Spontaneous self-affirmation is associated with positive outcomes in health contexts

Emotional support (Hajli, 2014)

Appendix B. Measurement Perceive usefulness of health information (Thong et al., 2002)

Cross-sectional survey

ES3

When faced with difficulties, some people on social media are on my side with me. When faced with difficulties, some people on social media comforted and encouraged me. When faced with difficulties, some people on social media expressed interest and concern in my well-being.

Informational support (Hajli, 2014) IS1

On social media, some people would offer suggestions when I needed help. When I encounter a problem, some people on social media would give me information to help me overcome the problem. When faced with difficulties, some people on social media would help me discover the cause and provide me with suggestions.

IS2

Perceived benefits (Benamati and Rajkumar, 2008)

PB2

I think discussing health related information on social media offers me a lot of advantages. I consider discussing health related information on social media to be beneficial.

Please cite this article as: Li, Y., et al., Seeking and sharing health information on social media: A net valence model and cross-cultural comparison, Technol. Forecast. Soc. Change (2016), http://dx.doi.org/10.1016/j.techfore.2016.07.021

12

Y. Li et al. / Technological Forecasting & Social Change xxx (2016) xxx–xxx

Mental intangibility (Laroche et al., 2004) MI2 MI3

The health related information I obtain from social media is not the sort of item that is easy to picture. The health related information I obtain from social media is a difficult item for me to understand.

Privacy risk (Featherman and Pavlou, 2003) PR1

PR2 PR3

If I discuss health related information on social media, I would lose control over the privacy of my personal information. My personal information would be less confidential if I discuss health related information on social media. Discussing health related information on social media would lead to a loss of privacy for me because my personal information would be used without my knowledge.

Time risk (Featherman and Pavlou, 2003) TR1

TR2 TR3

Discussing health related information on social media would lead to a loss of convenience for me because I would waste a lot of time. The time I spend in discussing health related information on social media makes it risky. I would waste a lot of time in discussing health related information on social media.

Social risk (Featherman and Pavlou, 2003) SR1

SR2

SR3

People who are important to me would think I'm foolish to discuss health related information on social media. Discussing health related information on social media would lead to a loss of status for me because my friends and relatives would think less highly of me. Discussing health related information on social media would harm the way others think of me.

Psychological risk (Featherman and Pavlou, 2003) PsyR1 PsyR2

I would probably get frustrated from discussing health related information on social media and feel foolish. In comparison to seeing a doctor, I would lose my peace of mind by discussing health information on social media.

Perceived risks (Featherman and Pavlou, 2003) RISK1

PISK2

I would expect that discussing health related information on social media confronts me with problems that I just don't need. There is a high level of risk that the expected benefits of discussing health related information on social media would not materialize.

Intention to share health information in social media (Liang et al., 2011; Lin and Lu, 2011) SHA1 SHA2 SHA3 SHA4

I intend to keep sharing health related information on social media in the future. I intend to share health related information on social media frequently in the future. I am willing to share health related information on social media. I am willing to share my health experience on social media.

Intention to seek health information in social media (Liang et al., 2011; Lin and Lu, 2011) SEE1 SEE2 SEE3 SEE4 SEE5 SEE6

I intend to keep seeking health related information on social media in the future. I intend to seek health related information on social media frequently in the future. I am willing to seek health related information on social media. I will seek related health information on social media when I need to. I will consider others' health experience on social media before I make a decision regarding health. I will ask others on social media to provide me with their suggestions before I make a decision regarding health.

References Allen, B., 1996. Information needs: A person-in-situation approach. In: Vakkari, P., Savolainen, R., Dervin, B. (Eds.), Information Seeking in Context. Taylor Graham, London, pp. 111–122. Allied Health World, 2012. A tweet a day keeps the doctor awayAvailable at http://www. alliedhealthworld.com/visuals/tweet-day-keeps-doctors-away.html (Accessed 6–29, 2015). Antheunis, M.L., Tates, K., Nieboer, T.E., 2013. Patients' and health professionals' use of social media in health care: motives, barriers and expectations. Patient Educ. Couns. 92 (3), 426–431. Bansal, G., Zahedi, F., Gefen, D., 2010. The impact of personal dispositions on information sensitivity, privacy concern and trust in disclosing health information online. Decis. Support. Syst. 49 (2), 138–150. Barclay, D., Higgins, C., Thompson, R., 1995. The partial least squares approach to causal modeling, personal computing adoption and use as an illustration. Technol. Stud. 2 (2), 285–309. Benamati, J.H., Rajkumar, T.M., 2008. An outsourcing acceptance model: an application of TAM to application development outsourcing decisions. Inf. Resour. Manag. J. 21 (2), 80–102. Bernhardt, J.M., Lariscy, R.A., Parrott, R.L., Silk, K.J., Felter, E.M., 2002. Perceived barriers to Internet-based health communication on human genetics. J. Health Commun. 7 (4), 325–340. Borgers, R., Mullen, P.D., Meertens, R., Rijken, M., Eussen, G., Plagge, I., Visser, A.P., Blijham, G.H., 1993. The information-seeking behavior of cancer outpatients: a description of the situation. Patient Educ. Couns. 22 (1), 35–46. Boulos, M.N.K., Hetherington, L., Wheeler, S., 2007. Second life: an overview of the potential of 3-D virtual worlds in medical and health education. Health Inf. Libr. J. 24 (4), 233–245. Brodie, R.J., Whittome, J.R., Brush, G.J., 2009. Investigating the service brand: a customer value perspective. J. Bus. Res. 62 (3), 345–355. Brown, T.A., 2006. Confirmatory Factor Analysis for Applied Research. The Guilford Press, New York, NY. Buckman, R., 1992. How to Break Bad News: A Guide for Health Care Professionals. JHU Press. Checton, M.G., Greene, K., 2015. Elderly patients' heart-related conditions: disclosing health information differs by target. Psychol. Health Med. 20 (5), 594–604. Chin, W.W., 2010. How to write up and report PLS analyses. In: Vinzi, V.E., Chin, W.W., Henseler, J., Wang, H. (Eds.), Handbook of Partial Least Squares: Concepts, Methods and Applications. Springer Verlag, New York, NY, pp. 655–690. Chou, W.S., Hunt, Y.M., Beckjord, E.B., Moser, R.P., Hesse, B.W., 2009. Social media use in the United States: implications for health communication. J. Med. Internet Res. 11 (4), e48. Cifter, A.S., Dong, H., 2009. User Characteristics: Professional vs. Lay Users. Royal College of Art, London. Cobb, S., 1976. Presidential address-1976. Social support as a moderator of life stress. Psychosom. Med. 38 (5), 300–314. Cotten, S.R., Gupta, S.S., 2004. Characteristics of online and offline health information seekers and factors that discriminate between them. Soc. Sci. Med. (1982) 59 (9), 1795–1806. Cunningham, S.M., 1967. The major dimensions of perceived risk. In: Cox, D.F. (Ed.), Risk Taking and Information Handling in Consumer Behavior. Harvard University Press, Cambridge, MA, pp. 21–33. Czaja, R., Manfredi, C., Price, J., 2003. The determinants and consequences of information seeking among cancer patients. J. Health Commun. 8 (6), 529–562. Featherman, M.S., Pavlou, P.A., 2003. Predicting e-services adoption: a perceived risk facets perspective. Int. J. Hum. Comput. Stud. 59 (4), 451–474. Featherman, M.S., Wells, J.D., 2004. The intangibility of E-services: effects on artificiality, perceived risk, and adoption. System Sciences, 2004. Proceedings of the 37th Annual Hawaii International Conference on. IEEE, pp. 177–187. Featherman, M.S., Wells, J.D., 2010. The intangibility of e-services: effects on perceived risk and acceptance. The DATA BASE for Advances in Information Systems. 41(2), pp. 110–131. Featherman, M.S., Miyazaki, A.D., Sprott, D.E., 2010. Reducing online privacy risk to facilitate e-service adoption: the influence of perceived ease of use and corporate credibility. J. Serv. Mark. 24 (3), 219–229.

Please cite this article as: Li, Y., et al., Seeking and sharing health information on social media: A net valence model and cross-cultural comparison, Technol. Forecast. Soc. Change (2016), http://dx.doi.org/10.1016/j.techfore.2016.07.021

Y. Li et al. / Technological Forecasting & Social Change xxx (2016) xxx–xxx Featherman, M.S., Valacich, J.S., Wells, J.D., 2006. Is that authentic or artificial? Understanding consumer perceptions of risk in e-service encounters. Inf. Syst. J. 16 (2), 107–134. Fishbein, M., 1967. Readings in Attitude Theory and Measurement. Wiley, New York, NY. Flanagin, A.J., Metzger, M.J., 2000. Perceptions of Internet information credibility. J. Mass Commun. Q. 77 (3), 515–540. Fox, S., Duggan, M., 2013. Health Online 2013. Pew Internet & American Life Project, Washington, DC. Gefen, D., Straub, D., 2005. A practical guide to factorial validity using pls-graph: Tutorial and annotated example. Commun. Assoc. Inf. Syst. 16 (5), 91–109. Gefen, D., Straub, D., Boudreau, M., 2000. Structural equation modeling and regression: guidelines for research practice. Commun. Assoc. Inf. Syst. 4, 1–77. Geisser, S., 1975. The predictive sample reuse method with applications. J. Am. Stat. Assoc. 70 (350), 320–328. Genuis, S.K., 2012. Social positioning theory as a lens for exploring health information seeking and decision making. Qual. Health Res. 23 (4), 555–567. Gewald, H., Dibbern, J., 2009. Risks and benefits of business process outsourcing: a study of transaction services in the German banking industry. Inf. Manag. 46 (4), 249–257. Hajli, M.N., 2014. The role of social support on relationship quality and social commerce. Technol. Forecast. Soc. Chang. 87, 17–27. Hajli, M., Bugshan, H., Lin, X., Featherman, M., 2013. From e-learning to social learning–a health care study. Eur. J. Train. Dev. 37 (9), 851–863. Hajli, M.N., Sims, J., Featherman, M., Love, P.E.D., 2015. Credibility of information in online communities. J. Strateg. Mark. 23 (3), 238–253. Hallyburton, A., Evarts, L.A., 2014. Gender and online health information seeking: a five survey meta-analysis. J. Consum. Health Internet 18 (2), 128–142. Hofstede, G., Hofstede, G.J., Minkov, M., 2010. Cultures and Organizations: Software of the Mind, Revised and Expanded 3rd Edition. McGraw-Hill, New York, NY. Holmes, K.L., Lenz, E.R., 1997. Perceived self-care information needs and informationseeking behaviors before and after elective spinal procedures. J. Neurosci. Nurs. 29 (2), 79–85. Honigman, B., 2013. 24 outstanding statistics & figures on how social media has impacted the health care industry. Available at https://getreferralmd.com/2013/09/healthcaresocial-media-statistics/ (Accessed 6–29, 2015). Househ, M., Borycki, E., Kushniruk, A., 2014. Empowering patients through social media: the benefits and challenges. Health Informatics J. 20 (1), 50–58. Hulland, J., 1999. Use of partial least squares (PLS) in strategic management research: a review of four recent studies. Strateg. Manag. J. 20, 195–204. Johnson, J.D., 1997. Cancer-related Information Seeking. Hampton Press. Kassulke, D., Stenner-Day, K., Coory, M., Ring, I., 1993. Information-seeking behaviour and sources of health information: associations with risk factor status in an analysis of three Queensland electorates. Aust. J. Public Health 17 (1), 51–57. Keil, M., Tan, B.C., Wei, K., Saarinen, T., Tuunainen, V., Wassenaar, A., 2000. A cross-cultural study on escalation of commitment behavior in software projects. MIS Q. 24 (2), 299–325. Ker, J., Wang, Y., Hajli, M.N., Song, J., Ker, C.W., 2014. Deploying lean in healthcare: evaluating information technology effectiveness in US hospital pharmacies. Int. J. Inf. Manag. 34 (4), 556–560. Kim, D., Olfman, L., 2011. Determinants of corporate web services adoption: a survey of companies in Korea. Commun. Assoc. Inf. Syst. 29, 1–24. Lambert, S.D., Loiselle, C.G., 2007. Health information—seeking behavior. Qual. Health Res. 17 (8), 1006–1019. Laroche, M., Bergeron, J., Goutaland, C., 2001. A three-dimensional scale of intangibility. J. Serv. Res. 4 (1), 26–38. Laroche, M., McDougall, G., Bergeron, J., Yang, Z., 2004. Exploring how intangibility affects perceived risk. J. Serv. Res. 6 (4), 373–390. Lee, M., 2009. Factors influencing the adoption of Internet banking: an integration of TAM and TPB with perceived risk and perceived benefit. Electron. Commer. Res. Appl. 8 (3), 130–141. Lenz, E.R., 1984. Information seeking: a component of client decisions and health behavior. Adv. Nurs. Sci. 6 (3), 59–72. Lewin, K., Dembo, T., Festinger, L., Sears, P.S., 1944. Level of aspiration. In: Hunt, J.M. (Ed.), Personality and the Behavior Disorders. Ronald Press, Oxford, pp. 333–378. Li, Y., Wang, X., 2016. Online social networking sites continuance intention: a model comparison approach. J. Comput. Inf. Syst. (in press). Liang, T.-P., Ho, Y.-T., Li, Y.-W., Turban, E., 2011. What drives social commerce: the role of social support and relationship quality. Int. J. Electron. Commer. 16 (2), 69–90. Lin, K.-Y., Lu, H.-P., 2011. Why people use social networking sites: an empirical study integrating network externalities and motivation theory. Comput. Hum. Behav. 27, 1152–1161. Lindell, M.K., Whitney, D.J., 2001. Accounting for common method variance in crosssectional research designs. J. Appl. Psychol. 86 (1), 114–121. Longo, D.R., 2005. Understanding health information, communication, and information seeking of patients and consumers: a comprehensive and integrated model. Health Expect. 8 (3), 189–194. Maloney-Krichmar, D., Preece, J., 2005. A multilevel analysis of sociability, usability, and community dynamics in an online health community. ACM Trans. Comput. Hum. Inter. 12 (2), 201–232. Matthews, A.K., Sellergren, S.A., Manfredi, C., Williams, M., 2002. Factors influencing medical information seeking among African American cancer patients. J. Health Commun. 7 (3), 205–219. McCoy, S., Galletta, D.F., King, W.R., 2005. Integrating national culture into is research: the need for current individual-level measures. Commun. Assoc. Inf. Syst. 15, 211–224. Miller, L.M.S., Bell, R.A., 2012. Online health information seeking the influence of age, information trustworthiness, and search challenges. J. Aging Health 24 (3), 525–541.

13

Mitchell, V.-W., 1992. Understanding consumers' behavior: can perceived risk theory help? Manag. Decis. 30 (2), 26–31. Möller, K., Rajala, R., Westerlund, M., 2008. Service innovation myopia? A new recipe for client-provider value creation. Calif. Manag. Rev. 50 (3), 31–48. Oh, K.M., Zhou, Q., Kreps, G., Kim, W., 2014. The influences of immigration on health information seeking behaviors among Korean Americans and native Koreans. Health Educ. Behav. 41 (2), 173–185. Peter, J.P., Tarpey Sr., L.X., 1975. A Comparative analysis of three consumer decision strategies. J. Consum. Res. 2 (1), 29–37. Podsakoff, P.M., MacKenzie, S., Lee, J., 2003. Common method biases in behavioral research: a critical review of the literature and recommended remedies. J. Appl. Psychol. 88 (5), 879–903. Porter, C.E., Donthu, N., 2006. Using the technology acceptance model to explain how attitudes determine internet usage: the role of perceived access barriers and demographics. J. Bus. Res. 59 (9), 999–1007. Rains, S.A., Karmikel, C.D., 2009. Health information-seeking and perceptions of website credibility: examining Web-use orientation, message characteristics, and structural features of websites. Comput. Hum. Behav. 25, 544–553. Rindfleisch, T.C., 1997. Privacy, information technology, and healthcare. Commun. ACM 40 (8), 93–100. Ringle, C.M., Wende, S., Will, S., 2005. SmartPLS 2.0 (M3) Beta, Hamburg. http://www. smartpls.de. Rowley, J., Johnson, F., Sbaffi, L., 2015. Students' trust judgments in online health information seeking. Health Informatics J. 21 (4), 316–327. Stone, M., 1974. Cross-validatory choice and assessment of statistical predictions. J. R. Stat. Soc. Ser. B 36 (2), 111–147. Suggs, L.S., Cowdery, J.E., Noll, D., 2010. Health information seeking behavior of young educated Hispanic women. Am. J. Health Stud. 25 (4), 186–195. Szwajcer, E.M., Hiddink, G.J., Koelen, M.A., van Woerkum, C.M.J., 2005. Nutrition-related information-seeking behaviours before and throughout the course of pregnancy: consequences for nutrition communication. Eur. J. Clin. Nutr. 59 (Suppl. 1), S57–S65. Taber, M.J., Howell, J.L., Emanuel, A.S., Klein, W.M.P., Ferrer, R.A., Harris, P.R., 2016. Associations of spontaneous self-affirmation with health care experiences and health information seeking in a national survey of US adults. Psychol. Health 31 (3), 292–309. Tardy, R.W., Hale, C., 1998. Getting “plugged in”: a network analysis of health-information seeking among ‘stay-at-home moms’. Commun. Monogr. 65 (4), 336–357. Thong, J.Y.L., Hong, W., Tam, K.Y., 2002. Understanding user acceptance of digital libraries: what are the roles of interface characteristics, organizational context, and individual differences. Int. J. Hum. Comput. Stud. 57 (3), 215–242. Vance, K., Howe, W., Dellavalle, R.P., 2009. Social internet sites as a source of public health information. Dermatol. Clin. 27 (2), 133–136. Venkatesh, V., Morris, M., Davis, G., Davis, F., 2003. User acceptance of information technology: toward a unified view. MIS Q. 27 (3), 452–478. Warner, D., Procaccino, J.D., 2004. Toward wellness: women seeking health information. J. Am. Soc. Inf. Sci. Technol. 55 (8), 709–730. Weeger, A., Wang, X., Gewald, H., 2015a. IT consumerization: BYOD-program acceptance and its impact on employer attractiveness. J. Comput. Inf. Syst. 56 (1), 1–10. Weeger, A., Wang, X., Gewald, H., Sanchez, O., Raisinghani, M., Grant, G., Pittayachawan, S., 2015b. Determinants of intention to participate in corporate BYOD-programs: the case of digital natives. Proceedings of the Academy of Management Meeting Annual Meeting, Vancouver, Canada. Wellman, B., Salaff, J., Dimitrova, D., Garton, L., Gulia, M., Haythornthwaite, C., 1996. Computer networks as social networks: collaborative work, telework, and virtual community. Annu. Rev. Sociol. 22, 213–238. Woodruff, R.B., 1997. Customer value: the next source for competitive advantage. J. Acad. Mark. Sci. 25 (2), 139–153. Yibai Li is an assistant professor in the Operation & Information Management Department, Kania School of Management, The University of Scranton. Li obtained his Ph.D. in Information Systems in Washington State University. Li obtained his B.S. in Computer Science and B.A. in Business Administration from Jilin University in China, and obtained a master of science in Management Information Systems and a data mining certificate from Oklahoma State University. His research interests include Social Computing, Discrete-event Simulation and Business Intelligence. His teaching areas include Management Information Systems, Business Intelligence Systems, SAP/ERP and Data Mining. Xuequn (Alex) Wang is a Lecturer in Murdoch University. He received his Ph.D. in Information Systems from Washington State University. His research interests include knowledge management, online communities, and idea generation. His research has appeared (or is forthcoming) in Communications of the Association for Information Systems, Journal of Organizational Computing and Electronic Commerce, Behaviour& Information Technology, Journal of Computer Information Systems, and Journal of Knowledge Management. Xiaolin Lin is a visiting assistant professor in the Department of Decision Sciences and Economics, College of Business, Texas A&M University-Corpus Christi. His research focuses on social commerce, healthcare IT, information security, and gender differences in IT behavioral research. His research has appeared in the Journal of Business Ethics, International Journal of Market Research, European Journal of Training and Development as well as conferences such as Americas Conference on Information Systems, HICSS, and DSI. Mohammad Hajli's research is on social media stream, and has published on social media, social commerce, and e-health research stream.

Please cite this article as: Li, Y., et al., Seeking and sharing health information on social media: A net valence model and cross-cultural comparison, Technol. Forecast. Soc. Change (2016), http://dx.doi.org/10.1016/j.techfore.2016.07.021