An exploratory investigation of the investment information search behavior of individual domestic investors

An exploratory investigation of the investment information search behavior of individual domestic investors

Telematics and Informatics 29 (2012) 187–203 Contents lists available at SciVerse ScienceDirect Telematics and Informatics journal homepage: www.els...

275KB Sizes 0 Downloads 2 Views

Telematics and Informatics 29 (2012) 187–203

Contents lists available at SciVerse ScienceDirect

Telematics and Informatics journal homepage: www.elsevier.com/locate/tele

An exploratory investigation of the investment information search behavior of individual domestic investors Wee-Kheng Tan ⇑, Yu-Jie Tan Department of Information & Electronic Commerce, Kainan University, No. 1, Kainan Road, Luchu, Taoyuan County 33857, Taiwan

a r t i c l e

i n f o

Article history: Received 18 March 2011 Received in revised form 23 August 2011 Accepted 14 September 2011 Available online 19 September 2011 Keywords: Investment information sources Social networks Social capital Technology readiness

a b s t r a c t This paper investigates the information search behavior of individual investors, particularly the roles played by online and offline social networks, and from the perspective of social capital and technology readiness. The paper uses rough sets analysis as the analytical tool. This study shows that despite the popularity of social network websites, the social capital of online community is still low when compared to the offline community. Hence, online communities play a less limited role in investment informational social support. Furthermore, investors with low social capital are often characterized by more intense reliance within narrow subgroups for investment information. Investors, especially those with high-investment risk profile, are truly hybrid information consumers. Investors who are innovative in technology and transact online are likely to be the users of online sources. Young investors are more likely to seek advice from online friends. The interaction effect among investment information sources is explicitly displayed via the rough sets decision rules. Ó 2011 Elsevier Ltd. All rights reserved.

1. Introduction There are many studies on information sources and they look from various perspectives such as credibility, satisfaction and decision-making (Arndt, 1967; Bansal and Voyer, 2000; Bickart and Schindler, 2002; Engel et al., 1995; Fodness and Murray, 1999; Punj and Staelin, 1983; Schwarzkopf, 2007; Westbrook, 1987). Consumers are likely to conduct extensive search if the purchased products are expensive or risky (Beatty and Smith, 1987; Moore and Lehmann, 1980). There are suggestions that consumers make purchase decisions after rational and unbiased processing of all the available market information. However, there are also counter-arguments that information processing is selective and is influenced by the social dynamics between groups (Oberlechner and Hocking, 2004; Scharfstein and Stein, 1990; Shiller, 1989). Investment is a risky activity where ‘‘investors have to predict the unknown realization of market outcomes at the time of purchase’’ (Lin and Lee, 2004, p. 320). Information is essential ‘‘in making investment decisions, a high-consequence decision task’’ (Loibl and Hira, 2009, p. 24) and wealth maximization (Nagy and Obenberger, 1994). Professional and institutional investors can depend on information sources such as industry newsletters, company’s financial statements, auditors’ reports, and financial analyst’s or brokerage firm’s reports for in-depth information and analysis. However, many individual domestic investors consider these investment information sources to be too detail, difficult to understand or expensive. Individual domestic investors often get information from mass media such as newspapers, television and the Internet. They can also rely on offline (real) social networks or online (virtual) communities for investment information.

⇑ Corresponding author. Tel.: +886 3 341 2500x6184; fax: +886 3 341 2373. E-mail address: [email protected] (W.-K. Tan). 0736-5853/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.tele.2011.09.002

188

W.-K. Tan, Y.-J. Tan / Telematics and Informatics 29 (2012) 187–203

Despite the above, ‘‘consumer information search for investment decisions have not received much attention’’ (Lin and Lee, 2004, p. 320) and ‘‘literature on the particular topic of external information search for investment information is scarce’’ (Loibl and Hira, 2009, p. 24). There are also limited studies on the relationship of information sources (both offline and online sources) and social networks in the context of investment information search behavior of individual domestic investors. This exploratory study strives to fill this research gap. In particular, this study hopes to shed some light on how these information consumers are influenced by offline and online social networks, and from the perspective of social capital (Lin, 1999, 2001) and technology readiness (Parasuraman, 2000; Parasuraman and Colby, 2001). The following investment information sources are considered in this study: online media source; online interpersonal source; offline media sources; offline interpersonal source; and professional sources. Besides the popular demographic variables and the investment risk preference variable, this study includes social capital variables, technology readiness variables and online application preference variable to characterize Taiwan individual domestic investors. Social capital variables are used to quantify the impact of social networks. Technology readiness variables are adopted since information sources also include online media and online friends. As for research approach, data are obtained from a survey of the Taiwanese individual investors, and rough sets analysis, a multi-criteria method, is used as the research method. In summary, given the research gap and the popularity of social networking websites, the interrelated research aims of this exploratory study are to: - Advance our understanding of the role of social capital in the search behavior of investment information consumers, - Provide evidence of the influence of online activities on the preferred investment information sources of information consumers. Thus, the research questions are: - Given the popularity of social networking websites, does the online social capital accumulated by the individual domestic investors influence their investment information search behavior in terms of choosing investment information sources? - Given the presence of online social capital, does offline social capital continue to play a role in deciding the investment information search behavior of individual domestic investors? In the remaining of the paper, we first review the existing literature. Section 3 discusses the key research method in this study. In Section 4, we present our results and the next section on the potential explanations of the results, research limitations and future research directions. Section 6 concludes.

2. Literature review Investment involves substantial amount of money and the potential for financial loss (Lin and Lee, 2004). Hence, investment can be a risky activity. Perceived risk is a kind of subjected expected loss (Peter and Ryan, 1976) and there are considerable studies that examine the impact of risk on traditional consumer decision (such as Lee, 2009; Lin, 2008). People also differ in their propensity to take risk (MacCrimmon and Wehrung, 1986) and propensity can affect investment decision making. Financial risk tolerance can be defined as ‘‘the amount of uncertainty or investment return volatility that an investor is willing to accept when making a financial decision’’ (Hallahan et al., 2003, p. 484). In the same study, Hallahan et al. (2003) found that gender, age, income and net wealth are important determinants of an individual’s attitude towards risk. There is also a negative but non-linear relationship between age and risk tolerance. According to (Grable, 2000), females have a lower preference for risk than males. Marital status and risk tolerance are not uniformly related (Hallahan et al., 2003). In other studies, individuals with higher education are believed to exhibit higher risk tolerance (Grable and Lytton, 1999). Investors who invest in New Zealand e-commerce companies are younger; more experience seeking and impulsive than investors who invest in more conventional companies (Hunter and Kemp, 2004). Information search is essential for making wise investment decisions (Guo, 2001). The theory of imperfect market information ‘‘implies that the benefits of searching include purchasing products with better appreciation potential that enable a higher potential return, reducing risk, increasing satisfaction with the decision, or accumulating investing experience that contributes to one’s stored knowledge’’ (Lin and Lee, 2004, p. 320). Previous literature has identified many factors that influence investment information consumer search behavior. Knowledgeable information consumers conduct more information search than less knowledgeable consumers (Brucks, 1985). According to (Loibl and Hira, 2009), higher-educated male investors with higher earnings are more likely to practice high-information search strategy. Internet-based information sources also dominate the high-information strategy. Except for reluctant investors, investors use a mixture of Internet-based mass media, interpersonal and workplace sources for investment information. Younger age is positively related to investment information search (Lin and Lee, 2004). Barniv and Cao (2009) suggested that increased uncertainty stimulates investor demand for analyst research. Studies have found that foreign investors are more informed and performed better than local investors (Grinblatt and Keloharju, 2000; Seasholes, 2000). However, there are also counter-arguments that it is not necessary so (Choe et al., 2000; Dvorak, 2005). Investment information search involves costs. Individuals also have limited information-processing abilities (Palomino et al., 2009). Hence, information consumers often engage in trade-off. Lesser

W.-K. Tan, Y.-J. Tan / Telematics and Informatics 29 (2012) 187–203

189

information search behavior takes place when the search costs are high (Loibl and Hira, 2009). Information consumers may also prefer highly visible, easy to process information (Palomino et al., 2009). Social network is closely related to the social support a member may get in time of needs (Lin, 2001; Bian and Ang, 1997; LaRue et al., 2010; Levy and Pescosolida, 2002). People rely on social networks for social support (Lee et al., 2005; Thoits, 1982; Vaux, 1988; Veiel and Baumann, 1992) and receive socioeconomic resources (Lin, 1999). Social support is also an important aspect of social interaction (Tanis, 2007). Social support is ‘‘information leading the subject to believe that he or she is cared for and loved, that he or she is esteemed and valued, and he or she belongs to a network of communication and mutual obligation’’ (Cobb, 1976, p. 379). It reduces uncertainty about the situation, enhances the perception of personal control in one’s life experience and contributes to mental and physical well-being (Albrecht and Adelman, 1987; Burleson et al., 1994; Uchino et al., 1996). The functional components of social support can appear as emotional support, instrumental support, informational support, affectionate support and social companionship (Vaux, 1988; Veiel and Baumann, 1992; House, 1986; House and Kahn, 1985; Leung and Lee, 2005; Sherbourne and Stewart, 1991). Informational support, which is the subject of this study, involves providing practical information and experiences (in the form of word-of-mouth) to increase the knowledge base of the information recipients (Tanis, 2007; Reeves, 2000), reduce uncertainty and make better decisions (Albrecht and Adelman, 1987; Wright, 2002). Information and advice on investment is a form of informational social support. Earlier studies show that social networks are not equal in the provision of social support. Social support is more forthcoming from neighborhood solidarities (Wellman, 1999). The latter comprises members who are tightly bounded. These members also enjoy strong relationship. Chinese societies are found on human relationships and they are useful resources (Hwang, 1987; Taylor et al., 2004). Online communities formed by families, friends, neighbors and colleagues are major resource for one to achieve its personal goals (Lee et al., 2005; Chuang and Chuang, 2008) and they are critical component when doing business (Hamilton, 1996). Kin ties tend to provide instrumental support while non-kin ties provide emotional support and social companionship (Freeman and Ruan, 1997; Wellman and Wortley, 1990). Friendship can be a source of support for single young adults (Bellotti, 2008). Social networking websites are now highly popular (Bausch and Han, 2006; Lipsman, 2007). Online social networks make use of the social affordances of Internet to produce ‘‘mediated social spaces in the digital environment that allow groups to form and be sustained primarily through ongoing communication processes’’ (Bagozzi and Dholakia, 2002, p. 3). Most of the online communities are organized around some distinct interest where membership is driven by volitional choice, members feel a ‘‘consciousness of kin’’, create and use shared conventions and language, and contribute Internet content actively (Bagozzi and Dholakia, 2002). Online community members are shown to share information and resources with other members (Walther, 1996), and have less constraints seeking information (Constant et al., 1996). Information sharing also takes place without geographical and time constraint. There are studies that suggested ‘‘online community members are reporting the kind of strong emotional and social bonds associated with local community, sharing the resources of stories and information, enjoying their time together online and working toward common goals’’ (Haythornthwaite, 2007, pp. 121). Besides social network, social capital and trust are the other dimensions of the same social phenomenon (Ferrary, 2003). Social capital can be the property of communities and nations (Fukuyama, 1995; Putnam, 1993). An umbrella term (Ferlander and Timms, 2001), social capital (Lin, 2001, 1982; Bourdieu, 1985; Van Der Gaag and Snijders, 2004) involves ‘‘social structure that facilitates certain actions of actors – whether personal or corporate actors – within the structure’’ (Coleman, 1988, p. 598). It is the sum of social resources accrued to an individual or a group, embedded within, available through, and derived from the network of relationships (Coleman, 1988; Bourdieu and Wacquant, 1992; Burt, 1992; Loury, 1977; Nahapiet and Ghoshal, 1998). It is a form of advantages and opportunities obtained by belonging to certain community. Relationships help to generate social capital (Lin, 1999). Social capital is widely felt in many facets of life such as job search (Granovetter, 1973) and income differences (Carroll and Teo, 1996). At the individual level, its importance lies in the benefit of social support (Requena, 2003). Use of Internet supplements social capital (Wellman et al., 2001). Similar to offline social networks, users can engage in online social activities, build and maintain social capital (Ellison et al., 2006; Ganley and Lampe, 2009). On the other hand, there are also studies that found that social networking websites support the formation and maintenance of weak ties rather than the creation of strong ties (Donath and Boyd, 2004), and Internet decreases social capital (Putnam, 2000). According to (Granovetter, 1973), online social networks are likely to comprise weak ties and bridging social capital instead of strong ties and bonding social capital. This disembedding process (Giddens, 1990), that is interaction over distance or with computers, can lower opportunities for trust building (Riegelsberger et al., 2007). There are some studies on the relationship between social capital and investment tools. The study by Guiso et al. (2004) found that households in areas with high level of social capital invest more in stock than in cash. Social relationship generates trust that glues community together and enforces norms of behavior (Anderson and Jack, 2002; Granovetter, 1985). Trust is an integral part of coordinated action among humans and an important lubricant of social activities (Arrow, 1974). It enables exchanges that may otherwise not take place (Lin, 2001). Trust involves the idea of risk (Deutsch, 1962; Green, 2007) and reduces moral hazard in dealing with another party in economic activity (Ferrary, 2003). A kind of reputation effect (Milgrom and Roberts, 1992), trust is the extent people find strangers trustworthy and is based on ‘‘a sense of confidence that others will respond as expected and will act in mutually supportive ways, or at least that others do not intend to do harm’’ (Onyx and Bullen, 2000, p. 24). Trust is also an important component of social capital (Fukuyama, 1995; Coleman, 1988). High level of social capital is associated with high level of trust (Banfield, 1958). For Paldam and Svendsen (2000), social capital is defined as the density of trust. Trust in the investment information providers is particularly needed because inaccurate information may cause investors to incur heavy losses (Green, 2007).

190

W.-K. Tan, Y.-J. Tan / Telematics and Informatics 29 (2012) 187–203

Using online information sources and engaging online friends involve whether users are mentally receptive to technology. Technology readiness can be used to measure this aspect. Technology readiness is a multi-dimensional psychographic construct and refers to the ‘‘propensity to embrace and use new technologies for accomplishing goals in home life and at work’’ (Parasuraman, 2000, p. 308). It is a state of mind and a result of ‘‘mental enablers and inhibitors that collectively determine a person’s predisposition to use new technologies’’ (p. 308). It is a concept that balances contributors (optimism and innovativeness) and inhibitors (discomfort and insecurity). Optimism refers to a positive view of technology and a belief that it offers people increased control, flexibility, and efficiency in their lives. Innovativeness describes a tendency to be a technology pioneer and thought leader. Discomfort refers to the perceived lack of control over technology and a feeling of being overwhelmed by it. Lastly, insecurity is a distrust of technology and skepticism about its ability to work properly. Technology readiness is used in a wide-range of consumer market and technology-related issues (Liljandera et al., 2006; Massey et al., 2007; Taylor et al., 2002; Tsikriktsis, 2004). 3. Rough sets analysis Rough sets analysis (Pawlak, 1982, 1984) is a knowledge discovery and data mining technique. It reveals the embedded data pattern of a dataset by presenting the results in the form of decision rules. According to (Pawlak, 2002), rough set theory overlaps with many other theories, such as fuzzy sets, evidence theory, and statistics. Rough sets analysis can also be viewed as a form of classification and pattern recognition technique. It can be an alternative or complement to traditional statistical approaches such as ANOVA/MANOVA, discriminant function analysis, logistic regression and cluster analysis (Conlon et al., 2004; Hair et al., 1998; Johnson and Dean, 2007). Rough sets analysis starts with an information system IS = (U,A) where the universe U = {x1, x2, . . ., xm} is a nonempty finite set. The set A is a nonempty finite set of attributes comprising a finite set of condition attributes and decision attributes. For every set of attributes B, an indiscernibility relation Ind(B) exists where two objects, xi and xj, are indiscernible by the set of attributes B in A, if b(xi) = b(xj) for every b. The equivalence class of Ind(B) is known as the elementary set in B. For any element xi of U, the equivalence class of xi in relation Ind(B) is represented as [xi]Ind(B). With X being the subset of elements in the universe U, lower approximation (BX) of X in B is the union of the elementary sets in X:

BX ¼ fxi 2 Uj½xi IndðBÞ  Xg

ð1Þ

and the upper approximation (BX) of X is the collection of elementary sets which have non-empty intersection with X:

BX ¼ fxi 2 Uj½xi IndðBÞ \ X–0g

ð2Þ

The analysis checks dependencies between attributes, reduces attributes that are not essential to characterize knowledge and analyzes the significance of attributes via reducts and core. Reducts are all the possible minimal subsets of attributes that lead to the same number of elementary sets as the whole set of attributes. They provide the same quality of classification as the set of all attributes. Core is the common part of all the reducts where the set of attributes are indispensable. Finally, text-like decision rules are generated (Komorowski et al., 1999; Pawlak, 1991, 1997; Tay and Shen, 2002). In the decision rule:

h)/ the left-hand and the right-hand terms are the condition and decision, respectively and the mathematical operator denotes propositional implication. The number of objects that satisfy the condition part of the rule is the strength of the rule. In summary, the general steps of rough set analysis are: (1) (2) (3) (4)

Determine the information system with the appropriate condition attributes and decision attribute(s). Calculate the upper and lower approximation. Find the core and reducts. Generate the ‘‘if-then’’ decision rules.

Rough sets analysis offers several advantages. It handles inexact, uncertain and vague datasets; copes with quantitative and qualitative attributes; generates minimal number of decision rules and offers straightforward interpretation of results (Imai et al., 2008; Shen and Loh, 2004; Shyng et al., 2007). Rough sets analysis is widely applied to many real-life issues. Reference (Goh and Law, 2003) used rough sets analysis to predict demand for tourism. When six factors, these being country of origin, Gross Domestic Product (GDP), relative consumer price index (CPI), population, volume of trade and foreign exchange rate, were used to determine tourism demand of the ten main tourist generating countries/region to Hong Kong, the analysis revealed that volume of trade and GDP are the most robust as they often appeared in the generated decision rules. Rough sets analysis is also applicable to finance and business-related issues such as business failure prediction (Beynon and Peel, 2001), insurance market (Shyng et al., 2007) and loan payment (Ruzgar and Ruzgar, 2008). Besides using rough sets analysis to determine behaviors that increase the risk of nonpayment or non-refunding of loan, (Ruzgar and Ruzgar, 2008) also compared rough sets approach with logistic regression. It was found that rough sets approach is better than logistic regression in classifying refund or non-refund of loan

W.-K. Tan, Y.-J. Tan / Telematics and Informatics 29 (2012) 187–203

191

payment. Rough sets analysis could also be used to analyze medical data which usually contain high degree of uncertainty. The study (Yang and Wu, 2009) used rough sets analysis on medical data of rhinology and throat logy to identify significant symptoms (such as fever, cough, sputum and nasal pain) that cause the diseases and generate strong decision rules to improve medical diagnosis. Rough sets analysis has also been employed to solve transportation-related matters such as hits and runs cases (Kim et al., 2008) and car ownership (Clark, 2009). The latter study utilized rough sets analysis on the Great Britain National Travel Survey dataset to generate a set of simple if-then decision rules that display combination of factors affecting the level of car ownership in a household. These rules could then be used to predict car ownership for other households. The paper further concluded with ‘‘the classification performance of the rough set method is on a par with standard statistical approaches to modelling data’’ (p. 392) until information on previous car ownership was included. Rough sets analysis was used to predict occurrence of debris flow to help to prevent loss of life and properties (Wan et al., 2008). Based on condition attributes such as effective watershed area, slope, shape factor and normalized difference vegetation index (NDVI), Wan et al. (2008) combined K-means analysis (a type of cluster analysis) and rough sets analysis to generate rational knowledge rules on likelihood of debris flow. It was found that three condition attributes, normalized difference vegetation index, slope and effective watershed area, are major factors influencing debris flow. The above literature review shows that gender, age, marital status, education and investment risk-taking profile of the information consumers can possibly affect their investment information search behavior. This study proposes that social networks via the offline and online social capital attributes may influence the search behavior of investment information consumers. This study further proposes that technology readiness attribute and online application attribute may affect the search behavior. To answer the research questions, this paper has to treat investment information search behavior of individual investors as a multi-attribute problem. Since rough sets analysis is a multi-attribute analytical tool, it is thus a possible research method candidate. Rough sets analysis can generate decision rules that display the different permutations of attributes under different circumstances. Hence, it goes beyond revealing which attributes are important. These decision rules reveal the different scenarios of how an attribute or more, single-handedly or in combination, contribute to a certain decision condition. It allows the interaction effects among information sources (Lee and Hogarth, 2000) to be shown explicitly via the decision rules. These permutations are more useful and richer in content than knowing that certain attribute is important. Since the rough sets analysis is an appropriate tool for this study, the method is adopted.

4. Survey and results A questionnaire-based survey was used. The questionnaire measurements were derived from literature review and modified according to the need of this study. The original items were translated from English to Chinese, and the draft of the questionnaire was subjected to pre-test and modified for better readability and clarity. The survey comprised four parts. First part dealt with the demographic profile of the investors. Demographic variables: gender, age group, marital status, educational level and employments, in the form of categorical variables, were included because previous literature showed that they affect the search behavior of investment information consumers. The second part of the survey asked investors on their view on technology. They are the technology readiness variables and online application preference variable. The questions on technology readiness and the list of online applications were adopted and finalized through relevant literature review and discussions with 20 Internet users. The degree of technology readiness was measured via 21 Likert-like statements that cover the different dimensions of technology readiness, with ‘1’ indicating ‘strongly disagree’ to ‘5’ indicating ‘strongly agree’. Investors were also asked to provide the extent they used 16 online applications through Likert-like statements with ‘0’ indicating ‘never’ to ‘4’ indicating ‘very often’. The 16 online applications are email, surf for information, transmit information, classify information, blog, multimedia, chit-chat, make friends, instant messaging, interactive online game, buying via e-tailers, bidding at e-auction platform, selling via Internet, online share buying/selling, online banking, and rotating credit association. The third part covered the measurement of social capital. Respondents were asked whether members of the offline world and online world were trustworthy, willing to help and took advantage if opportunities arose (Onyx and Bullen, 2000). Position generator method (Lin and Dumin, 1986; Lin et al., 2001) was used to measure individuals’ access to social capital through a sample of structural positions (occupations). A list of 24 occupations with occupational prestige varying from 22 (cleaner) to 85 (legislator) arranged in a randomized order in terms of prestige and followed the Standard International Occupational Prestige Scale (Ganzeboom and Treiman, 1996) was used in this study. The occupations comprised three tiers: Tier 1 (prestige ranging from 63 to 85): legislators, doctors, university professors, lawyers, CEOs of big enterprise and production managers; Tier 2 (prestige ranging from 40 to 60): middle school teachers, personnel managers, writers, reporters, nurses, computer programmers, administrative assistants of big enterprise, accountants, owners of small enterprises and policemen; and Tier 3 (prestige ranging from 22 to 38): farmers, receptionists, operators in a factory, hairdressers, taxi drivers, security guards and housemaids. Respondents were asked if they knew any of them in the offline and online world. If a person was known initially in the offline world and ongoing contact has now also extended to the online world, that person was categorized as belonging to the offline world. Four indexes were then constructed to measure the degree of presence of social capital: extensity of accessibility (number of different positions accessed), upper reachability of accessed social capital (highest prestige occupation accessed), lower reachability (lowest prestige occupation accessed) and range of accessibility to different hierarchical occupations (distance between the highest and lowest accessed occupations).

192

W.-K. Tan, Y.-J. Tan / Telematics and Informatics 29 (2012) 187–203

The fourth part dealt with investment information sources. Respondents were asked to rate their investment risk attitude from ‘‘very low risk’’ to ‘‘high risk’’. Through literature survey and discussion with investors, a list of 14 investment information sources was drawn up and the respondents were asked to choose the sources they actually used via a set of binary variables (‘1’ = used a particular information source, ‘0’ = do not use). Information sources included one online media source (information websites), one online interpersonal source (friends on the Internet), three offline mass media source (newspapers, magazines, radio and television), seven offline interpersonal source (family members, relatives, close friends, colleagues, classmates, normal friends, neighbors) and two professional sources (bankers, stockbrokers and insurance agents, paid financial service advisers). The survey was administrated through paper-based survey and a series of interpersonal interviews. Given the lengthy survey, this technique was used to obtain higher response rate and to improve the accuracy of this study. This technique was also used to ensure that the respondents were domestic individual investors, 20 year or above and had made investment decisions. College students and emerging adults (Arnett, 2000) were often the subject of many Internet and online community studies (Boyd, 2008; Steinfield et al., 2008). As most teenagers are receiving formal schooling and they depend on their parents for financial support, investment is not common among them. Hence, this study differs from these studies by focusing only on Taiwanese individual domestic investors who are 20 year old and above. It also justifies the use of technology readiness in this study to account for adults’ receptiveness to technology since this study deals with adults and they may not be as computer literate as teenagers. The sample was recruited through a convenience sampling method. A total of 96 valid returns were made (Table 1). The respondents were Taiwanese adults (20 year old and above), had made investment before, 57 (59.4%) were male, 55 (57.3%) were single and 77 (80.2%) received tertiary education. 4.1. Online application preference The Bartlett test of sphericity (758.58, q = 0.00) was conducted for the 16 online applications statements. The value of the KMO statistics was 0.79 that fell within the range of a good model. Thus, the test confirmed that factor analysis was appropriate. Principal axis factoring with varimax orthogonal rotation was then employed. The result suggested a four-factor solution (Table 2) and the factors extracted explained 67.75% of the total variance. These factors were named appropriately as information seeking, socializing, transaction (merchandize) and transaction (financial). Their reliability coefficients (Cronbach’s alpha) were acceptable or high reliability. ANOVA test showed that the mean of these four groups of online applications (Table 2) were significantly difference (F = 163.99, q < 0.05). Information seeking was the most popular group of online applications, followed by transaction (merchandise). A two-step clustering method that combined hierarchical and non-hierarchical clustering algorithms was conducted on the online applications of the dataset. Ward’s method with Squared Euclidean Distance, a hierarchical procedure, was used to generate an initial cluster solution. An initial two cluster solution was obtained and they were used as initial seeds for the kmeans cluster analysis, a non-hierarchical technique, to generate the final cluster solutions (Table 2). Cluster 1 and 2 were labeled descriptively as ‘‘information seeking’’ and ‘‘transaction (merchandize)’’, respectively. To check whether these clus-

Table 1 Profile of respondents. Number

Percent

Gender

Male Female

57 39

59.4 40.6

Marital Status

Married Single

41 55

42.7 57.3

Employment

Employed Students or waiting for further study Retired Unemployed

81 4 2 9

84.3 4.2 2.1 9.4

Age

20–29 30–39 40–49 50–59

27 45 12 12

28.1 46.9 12.5 12.5

Education

High School Junior college University (Bachelor) University (Master/PhD)

6 13 46 31

6.3 13.5 47.9 32.3

Investment Risk-taking Profile

Very low risk Low risk Medium risk High risk

15 33 36 12

15.62 34.38 37.50 12.50

193

W.-K. Tan, Y.-J. Tan / Telematics and Informatics 29 (2012) 187–203 Table 2 Online application preference (factor analysis, mean and final cluster centers). Items

Loading

(A) Factor analysis and mean Factor 1: Information seeking [Mean = 3.51; S.D. = 0.80] (Eigenvalue = 5.52; Variance = 32.82%; Cronbach Alpha = 0.82) Email Surf for information Transmit information Classify information Blog Multimedia (e.g. Youtube)

0.53 0.77 0.79 0.87 0.65 0.54

Factor 2: Socializing [Mean = 2.28; S.D. = 1.18] (Eigenvalue = 2.44; Variance = 15.27%; Cronbach Alpha = 0.80) Chit-chat Make friend Instant messaging Interactive online game

0.86 0.79 0.65 0.66

Factor 3: Transaction (merchandise) [Mean = 2.39; S.D. = 1.08] (Eigenvalue = 1.89; Variance = 11.82%; Cronbach Alpha = 0.84) Buying via e-tailers Bidding at e-auction platform Selling via Internet

0.88 0.91 0.55

Factor 4: Transaction (financial) [Mean = 2.19; S.D. = 1.24] (Eigenvalue = 1.26; Variance = 7.84%; Cronbach Alpha = 0.70) Online share buying/selling Online banking Rotating credit association

0.72 0.73 0.69

(B) Final Cluster Centers Information seeking Socializing Transaction (merchandize) Transaction (financial)

Cluster 1 (n = 57)

Cluster 2 (n = 39)

0.17 0.05 0.67 0.13

0.26 0.07 1.02 0.19

ters were appropriate, discriminate analysis was used with cluster groupings as dependent variable and the 4 extracted factors as predictor variables. Univariate ANOVA showed that cluster grouping differed significantly on each of the 4 predictor variables. The value of discriminant function was significantly different (chi-square = 113.258, q < 0.0005) and overall, it successfully predicted outcome for 100% of the cases. Hence, these clusters were appropriate. 4.2. Technology readiness Similarly, the Bartlett test of sphericity was conducted on the statements measuring technology readiness (Table 3). The KMO statistics was 0.79 with Bartlett test of sphericity being significant (930.47, q = 0.00). Five factors were extracted and they explained 67.86% of the total variance. Their reliability coefficients (Cronbach’s alpha) were acceptable or high reliability. The five factors were labeled as innovativeness, optimism, insecurity, discomfort (technology) and discomfort (social). ANOVA test showed that that the mean of these five factors (Table 3) were significantly different (F = 58.87, q < 0.05). Optimism has the highest mean, followed by insecurity. Two-step clustering procedure was also applied on technology readiness of the dataset. Final cluster 1 was labeled as ‘‘innovativeness’’ and cluster 2 as ‘‘optimism’’ (Table 3). To determine whether these clusters were appropriate, discriminate analysis was conducted with cluster grouping as the dependent variable and the factors extracted from factor analysis, that was, optimism, innovativeness, insecurity, discomfort (technology) and discomfort (social) as predictor variables. The value of discriminant function was significantly different (chi-square = 102.88, q < 0.0005). Overall, the function successfully predicted outcome for 100% of the cases. Hence, these clusters were appropriate. 4.3. Social capital Respondents felt that people in the offline world were more trustworthy than the online world (Table 4). In the offline world, 71 (73.95%) thought that about half or more could be trusted while in the online world, it was only 44 (45.83%). Eighty-five (88.55%) and 78 (81.26%) of respondents felt that more than half of the people in the offline world would extend a helping hand when in need and would not take advantage of them, respectively. The equivalent figures in online world were only 49 (51.04%) and 63 (65.62%). From the perspective of trust, social capital of the offline world was higher than the online world since the former was more trustworthy, more helpful and less prone to take advantage.

194

W.-K. Tan, Y.-J. Tan / Telematics and Informatics 29 (2012) 187–203

Table 3 Technology readiness (factor analysis, mean and final cluster centers). Items

Loading

(A) Factor analysis and mean Factor 1: Innovativeness [Mean = 3.26; S.D. = 1.00] (Eigenvalue = 2.91; Variance = 13.86; Cronbach Alpha = 0.90) Other people come to you for advice on new technologies In general, you are among the first in your circle of friends to acquire new technology when it appears You can usually figure out new high-tech products and services without help from others You keep up with the latest technological developments in your area of interest You enjoy the challenge of figuring out high-tech You have fewer problems than other people when using technology

0.79 0.85 0.76 0.82 0.80 0.71

Factor 2: Optimism [Mean = 3.90; S.D. = 0.86] (Eigenvalue = 5.86; Variance = 27.88; Cronbach Alpha = 0.88) You like the idea of doing business via computers because you are not limited to regular business hours You prefer to use the most advanced technology available You like computer programs that allow you to tailor things to fit your own needs Technology makes you more efficient You find technologies to be mentally stimulating Technology gives you more freedom Learning about technology is also rewarding

0.78 0.72 0.79 0.83 0.68 0.76 0.66

Factor 3: Insecurity [Mean = 3.72; S.D. = 0.95] (Eigenvalue = 2.31; Variance = 10.99%; Cronbach Alpha = 0.74) You feel that it is not safe to provide credit card number over a computer You feel that It is not safe to do any kind of financial business online You are concerned that information you send over the Internet will be seen by other people You want written confirmation for any online transaction you do online Factor 4: Discomfort (technology) [Mean = 2.97; S.D. = 0.97] (Eigenvalue = 1.46; Variance = 8.93%; Cronbach Alpha = 0.67) You feel that they do not explain things in terms you understand Sometimes, you think that technology are not designed for use by ordinary people

0.84 0.74

Factor 5: Discomfort (social) [Mean = 3.70; S.D. = 0.90] (Eigenvalue = 1.30; Variance = 6.20%; Cronbach Alpha = 0.63) Many new technologies have health or safety risks that are not discovered until after people have used them New technology makes it easy for government and companies to spy on people

0.79 0.80

(B) Final Cluster Centers Optimism Innovativeness Insecurity Discomfort (technology) Discomfort (social)

0.69 0.77 0.75 0.74

Cluster 1 (n = 81)

Cluster 2 (n = 15)

0.18 0.29 0.01 0.05 0.02

0.99 1.59 0.06 0.27 0.13

Position generator analysis also revealed that online network had less social capital than offline network (Table 4). Independent t-test showed that the four related indexes: extensity, range, upper and lower reachability were significantly different across offline and online network. Two-step clustering procedure was conducted separately on the social capital of offline and online network. Two final clusters, each for online and offline social capital, were produced (Table 4). Cluster 1 of offline social capital was named as ‘‘wide offline social capital’’ and cluster 2 as ‘‘narrow offline social capital’’. Cluster 1 and 2 of online network were labeled as ‘‘lacking online social capital’’ and ‘‘possess online social capital’’, respectively. For both types of social capital, the values of discriminant functions, with cluster grouping as dependent variable and factors extracted from factor analysis as predictor variables, were significantly different and predicted outcome for 100% of cases.

4.4. Investment information sources Half of the investors were conservative or moderate in investment risk-taking (Table 1). When asked which sources were actually used (they could made multiple choices), family members, close friends and colleagues were top in the information sources actually used (Table 5). For personal sources, all discussions took place in social groups. The study used the Rough Sets Data Explorer 2 (ROSE2) system (Predki et al., 1998; Predki and Wilk, 1999) for the four rough sets analysis. To answer the first research question, the following two rough set analyses were made: (1) Comparison 1: online media source (information websites) and offline mass media sources (newspapers, magazines, television and radio). (2) Comparison 2: online interpersonal sources (friends on the Internet) and offline interpersonal sources.

195

W.-K. Tan, Y.-J. Tan / Telematics and Informatics 29 (2012) 187–203 Table 4 social capital (trust, helpfulness and taking advantage, position generator analysis indexes, and final cluster centers). Offline World %

No.

(A) Trust, helpfulness and taking advantage Do you think people can be trusted? Almost all cannot be trusted 9 Most cannot be trusted 16 About half can be trusted 29 Most can be trusted 39 Almost all can be trusted 3

9.38 16.67 30.20 40.63 3.12

11 41 35 9 0

11.46 42.71 36.46 9.37 0

Do you think people will extend a helping hand? Almost all will not 2 Most will not 9 About half will 34 Most will 42 Almost all will 9

2.07 9.38 35.42 43.75 9.38

12 35 28 20 1

12.50 36.46 29.17 20.83 1.04

5 28 27 25 11

5.21 29.17 28.13 26.04 11.45

Do you think people will take advantage of you when opportunity arises? Almost all will 2 2.07 Most will 16 16.67 About half will not 26 27.08 Most will not 40 41.68 Almost all will not 12 12.50 Offline Network Mean

Online Network Mean

S.D.

5.52 15.20 11.31 9.71

1.81 9.10 29.87 20.77

2.97 16.91 33.81 24.61

Cluster 1 (n = 59)

Cluster 2 (n = 37)

Cluster 1 (n = 52)

Cluster 2 (n = 44)

11 51 79 29

5 24 60 36

0 0 0 0

4 20 65 45

Offline Network

(C) Final Cluster Centers Extensity Range Upper reachability Lower reachability

%

t-Value

S.D.

(B) Position Generator Analysis Indexes Extensity 8.83 Range 40.44 Upper reachability 71.70 Lower reachability 31.26

*

Online World

No.

10.97* 13.50* 11.49* 3.88*

Online Network

q < 0.05.

Table 5 Investment information sources actually used. No. Family members Offline close friends Colleagues Information websites (Internet) Magazines Bankers, stockbrokers & insurance agents Friends on the Internet Newspapers Classmates Radio and television Relatives Paid financial service advisers Normal friends Neighbors

69 47 43 28 26 17 16 14 9 8 7 4 2 1

Besides using the results of the above analyses to answer the second research question, this study performed two more rough sets analysis to obtain a more holistic picture on the role of offline social capital: (3) Comparison 3: offline interpersonal sources (family members, offline close friends and colleagues) and media.

196

W.-K. Tan, Y.-J. Tan / Telematics and Informatics 29 (2012) 187–203 Table 6 Rough sets condition attributes. Attributes

Category

A1. A2. A3. A4.

1 – Male; 2 – Female 1–2029; 2–3039; 3–4049; 4–5059 1 – Married; 2 – Single 1 – Illiterate; 2 – Self–educated; 3 – Primary school; 4 – Secondary school; 5 – Senior high school; 6 – Vocational school; 7 – Undergraduate; 8 – Postgraduate 1 – Employed; 2 – Part–timer; 3 – Student; 4 – Housewife; 5 – Retired; 6 – Unemployed 1 – Very low risk; 2 – Low risk; 3 – Medium risk; 4 – High risk 1 – Innovativeness; 2 – Optimism 1 – Information seeking; 2 – Transaction (merchandize) 1 – Lacking online social capital; 2 – Possesses online social capital 1 – Wide offline social capital; 2 – Narrow offline social capital

Gender Age group Marital status Educational level

A5. Employment A6. Investment risk profile A7. Technology readiness A8. Online application A9. Online social capital A10. Offline social capital

Table 7 Rough sets comparison 1 (approximation results and decision rules). Approximation results Class

No. of objects

Approximation

Accuracy

Quality of classification

0.96

Lower

Upper

1 14 2 17 3 14 Decision rules

14 16 13

14 18 15

1.00 0.89 0.87

Rule

A2

A3

A4

A7

A8

A9

A10

Strength (%)

D1 – Use online information media (8 rules obtained with 5 having strength P10%) R1a 1 – – 7 – – R1b – – – 7 – 2 R1c – 1 – 7 – 3 R1d – 2 – 8 – – R1e – 2 – 6 – –

A1

A5

A6

1 – – – –

– 2 – – –

– – 2 2 1

1 – – – –

21.43 21.43 21.43 14.29 14.29

D2 – Use offline information media (10 rules obtained with 5 having strength P10%) R2a – – – – – 2 R2b – 2 – 7 – – R2c 1 3 – – – – R2d – – – – – 1 R2e – 4 – – – 2

– – – – –

1 2 – – –

1 – – – –

1 1 – – –

23.53 11.76 11.76 11.76 11.76

– 2 – 2 –

1 – – – –

– – 1 – 1

21.43 14.29 14.29 14.29 14.29

D3 – Use offline and online information media (8 rules obtained with 5 having strength P10%) R3a – – – – – 4 – R3b 2 – – 8 – – – R3c – – 2 6 – – – R3d – – – 8 – 3 – R3e 1 1 – – 1 2 –

(4) Comparison 4: family members and offline close friends/colleagues. There were 10 condition attributes for the first three comparisons (Table 6). 4.4.1. Comparison 1 There were 3 decision attributes: D1 – Use Only Online Media Sources; D2 – Use Only Offline Mass Media Sources; D3 – Use Both Offline and Online Information Media. The quality of classification was 0.96 (Table 7). The core was {A1, A4, A6, A7, A8, A9, A10} and the reducts were {A1, A2, A4, A6, A7, A8, A9, A10} and {A1, A3, A4, A6, A7, A8, A9, A10}. Twenty-six decision rules were obtained. Eight (8) rules were obtained for those who used online information media (D1) of which 5 of them had strength which were more than 10%. Ten (10) rules were obtained for those who used offline information media (D2) of which 5 of them had strength which were more than 10%. Eight (8) rules were for those who used both offline and online information media (D3). Five (5) of them had strength which were more than 10% (Table 7).

197

W.-K. Tan, Y.-J. Tan / Telematics and Informatics 29 (2012) 187–203 Table 8 Rough sets comparison 2 (approximation results and decision rules). Approximation results Class

No. of objects

Approximation

Accuracy

Quality of classification

3 76 13

1.00 1.00 1.00

1.00

A3

A4

Lower

Upper

1 3 2 76 3 13 Decision rules

3 76 13

Rule

A2

A1

A5

A6

A7

A8

A9

A10

Strength (%)

– 1 – – – – –

1 1 – – – – –

– – – – – – –

– 1 – – 1 – –

19.74 18.42 15.79 11.84 11.84 11.84 10.53

D3 – Family/Colleagues/Close friends/Online friends (10 rules obtained with 2 having strength P10%) R3a – 2 2 – – – – – R3b – 2 2 – – 3 – 2

– 2

1 –

23.08 15.38

D2 – Family/Colleagues/Close friends (14 rules obtained with 7 having strength P10%) R2a – – – 8 – – R2b – – – 7 – – R2c – 4 – – – – R2d – 2 – – – 2 R2e 2 – – 8 – – R2f 1 – 1 7 1 – R2 g – 3 2 – – –

According to the decision rules, older investors tend to use offline media but younger people have less such constraints. Investors with lower risk propensity tend to use offline media and those with high risk-taking profile are likely to use both types of media. Investors who are innovative in technology tend to use online media. Investors who transact online are likely to use both types of media. Investors who have online social capital are more prone to use online media but lack of online social capital does not prevent them from using both types of media. 4.4.2. Comparison 2 There were 3 decision attributes: D1 – Use Online Friends Source; D2 – Use Family/Colleagues/Close Friends Source; D3 – Use Both Family/Colleagues/Close Friends/Online Friend. The quality of classification was 1.00 (Table 8). The core was {A1, A2, A3, A4, A6, A7} and the reducts were {A1, A2, A3, A4, A6, A7, A8, A9}, {A1, A2, A3, A4, A6, A7, A8, A10} and {A1, A2, A3, A4, A6, A7, A9, A10}. Twenty-six decision rules were obtained. Two (2) rules were obtained for those who depended on online friends (D1). All of them had strength which was more than 10%. Fourteen (14) rules were obtained for those who approached Family/Colleagues/Close Friends (D2) of which 7 of them had strength which were more than 10%. Ten (10) rules were for those who approached Family/Colleagues/Close Friends/Online Friends (D3) and two of them had strength that was more than 10% (Table 8). Ignoring the rules under decision condition D1 (Online Friends) as it only has 3 objects, those who approached only family members, colleagues and closed friends for information tend to be investors with a wide range of age groups. However, those investors who include online friends tend to be younger. Investors who also include online friends tend to have higher risktaking profile. All of them have wide offline social capital. 4.4.3. Comparison 3 There were 3 decision attributes: D1 – use only offline interpersonal sources; D2 – use only mass media sources (online and offline), D3 – use both offline interpersonal sources and mass media sources. The quality of classification was 0.90 (Table 9). Core was {A1, A2, A3, A4, A5, A6, A8, A9, A10}. The reduct was {A1, A2, A3, A4, A5, A6, A8, A9, A10}. Thirty-nine decision rules were obtained. Seventeen (17) rules were obtained for those who used offline personal sources (D1) of which 7 of them had strength which were more than 10%. Three (3) rules were obtained for those who used media alone (D2) and all of them had strength which was more than 10%. Nineteen (19) rules were for those who used both offline personal sources and media (D3) of which 4 of them had strength which was equal to or more than 10% (Table 9). Ignoring the rules generated under decision condition D2 (Media only) because it only has five objects, females like to approach family members, close friends and colleagues. Male investors also like to do so but they also obtain advice from media. Those with higher risk investment profile use a combination of offline personal sources and media but those with lower risk profile tend to be less dependent on media. Those investors lacking in online social capital mostly approach family members, friends and colleagues. 4.4.4. Comparison 4 As this analysis did not consider online features, technology readiness and online application cluster were excluded. There were 3 decision attributes: D1 – Family; D2 – Offline Close Friends/Colleagues; D3 – Family and Offline Close

198

W.-K. Tan, Y.-J. Tan / Telematics and Informatics 29 (2012) 187–203

Table 9 Rough sets comparison 3 (approximation results and decision rules). Approximation results Class

No. of objects

Approximation

Accuracy

Quality of classification

53 6 44

0.83 0.67 0.84

0.90

A3

A4

Lower

Upper

1 49 2 5 3 40 Decision rules

44 4 37

Rule

A2

A1

A5

D1 – Offline personal sources (17 rules obtained with 7 having strength P10%) R1a – – – – – R1b – 2 2 – – R1c 2 2 – – 1 R1d 2 – – 8 – R1e 2 – – – 1 R1f – 4 1 – – R1g – – – 7 –

A6

A7

A8

A9

A10

Strength (%)

1 – – – – – 1

1 – 1 – 1 – –

– 2 2 1 2 – –

1 1 – 2 1 – –

– – – – – 1 –

14.29 12.24 12.24 12.24 12.24 10.20 10.20

1 – – –

– – – 2

– – – –

15.00 10.00 10.00 10.00

D3 – Use both offline personal sources and media (19 rules obtained with 4 having strength P10%) R3a – – 1 7 1 2 – R3b – – 2 – – 4 1 R3c 1 – – 6 – 3 – R3d 1 – – 7 – 2 –

Table 10 Rough sets comparison 4 (approximation results and decision rules). Approximation Results Class

No. of objects

Approximation

Accuracy

Quality of classification

21 22 55

0.71 0.77 0.87

0.90

A3

A4

Lower

Upper

1 18 2 19 3 52 Decision rules

15 17 48

Rule

A2

A1

D1 – Family members (8 rules obtained with 5 having strength P10%) R1a 1 – 2 – R1b 1 2 2 8 R1c 2 – 2 7 R1d – 2 2 – R1e 1 3 2 8

A5

A6

A9

A10

Strength (%)

– – – – –

2 3 – 1 –

2 – 1 – –

– – – 2 –

16.67 16.67 11.11 11.11 11.11

– 1 4 – 4

1 2 – 1 1

– – – – –

21.05 15.79 10.53 10.53 10.53

– 1 2 2

– – 1 1

25.00 17.31 15.38 11.54

D2 – Offline Close Friends/Colleagues (10 rules obtained with 5 having strength P10%) R2a 1 2 1 – – R2b – – 1 – – R2c 2 2 2 – – R2d – – – – 3 R2e – – – 8 –

D3 – Family and Offline Close Friends/Colleagues (16 rules obtained with 4 having strength P10%) R3a 2 – – – – 3 R3b 2 – 1 – 1 – R3c – – – 8 1 – R3d 2 – 2 – – –

Friends/Colleagues. Quality of classification was 0.90 (Table 10). Core was {A1, A2, A3, A4, A6, A9, A10} and the reduct was {A1, A2, A3, A4, A5, A6, A9, A10}. Thirty-four decision rules were obtained (Table 10). Eight (8) rules were obtained for those who approached Family (D1) of which 5 of them had strength which were more than 10%. Ten (10) rules were obtained for those who approached Offline Close Friends/Colleagues (D2) of which 5 of them had strength which were more than 10%. Sixteen (16) rules were for those who approached Family and Offline Close Friends/Colleagues (D3) and 4 of them had strength that was more than 10%. According to the decision rules, those investors who only approach family tend to be male investors but those who approach family members and close friends/colleagues simultaneously tend to be female investors. Investors who approach close friends and colleagues are likely to have higher risk-taking profile than those who only approach family members. Off-

W.-K. Tan, Y.-J. Tan / Telematics and Informatics 29 (2012) 187–203

199

line social capital of investors approaching only family members is more likely to be narrower than those who approach both sources. People who have online social capital are also likely to have wider offline social capital.

5. Discussion The results show that the online social capital accumulated by individual domestic investors has an influence on their investment information search behavior in terms of choosing investment information sources. However, the impact is less than anticipated even though social networking websites are highly popular. Hence, offline social capital continues to play a role in deciding the investment information search behavior of individual domestic investors. Online social capital having less than anticipated impact is due to the social capital of online community being lower than offline community. This is the case even for adults who access Internet and use its online applications. Getting advice from personal sources is a form of informational social support provided by social networks. As shown in this study, such personal sources weight heavily in investors’ information search. Extent of usefulness and trustworthiness of personal sources, and the potential to tap on them depend on whether their social networks are ‘‘rich’’ enough in members and the members are forthcoming with advice. Social capital is a relevant indicator. Higher social capital implies that one knows more people and they are probably from diversified background. Hence, investors have additional avenues to receive more and diversified information and advices. Tighter networks also generate trust in the advice given. From the perspective of trust, members of the online community are felt to be less trustworthy, less willing to extend a helping hand and more willing to take advantage when opportunities arise. From the perspective of position generator methodology, the number of people known (measured through extensity index) and its diversity (measured through range index) in online community are also much lower than offline community. People are also more likely to know people of higher occupational prestige in the offline world than in the online world. A key reason for such findings is computer-mediated communication allows users to remain anonymous if they want to do so (Bordia, 1997; Rice and Gattiker, 2001) and Internet users tend to use pseudo name to hide their true identity. Anonymity may make users more willing to express their true emotion, disclose more information and ask questions (Joinson, 2001; McKenna and Bargh, 2000). However, it also leads to de-individuation (Sproull and Kiesler, 1986), generates distrust and limits the usefulness of online community members in providing investment information. Another reason is people often use Internet to stay in contact with acquaintances already known in the offline world (which is counted as part of the offline community in this study). This is also prevalent in emerging adults (Subrahmanyam et al., 2008) where they use social networking sites to connect with people from their offline lives and do not look for strangers or add strangers to their online social network. Hence, the total social resources, best resources available and its differentiation are poorer in online community. However, the online media and social network cannot be discounted as sources of investment information. Being ‘‘hybrid’’ consumers (Wind et al., 2002), information consumers are willing to complement their existing offline information sources with online sources. Getting information and advice from more sources appear to be a way to mitigate and lower risk of bad investment. Hence, those with high risk-taking profile are likely to use a portfolio of offline and online media. They also rely heavily on personal sources, especially offline personal sources, whenever available and feasible. Using online information websites and engaging online friends require one to face and cope with technology. Results confirm earlier studies that younger investors and those investors who are at ease with technology are in a better position to capitalize on the advantages offered by the online media and social networks. They are also likely to have online social capital and use a combination of offline and online media while older investors tend to use offline media. Younger investors are more likely to seek advice from online friends. Offline social networks continue to play an important role in influencing the investment information search behavior because they are a useful form of traditional word-of-mouth. Like other studies on offline word-of-mouth (such as Bansal and Voyer, 2000; Bausch and Han, 2006), investors are found to place high reliance on advice from family members, close friends and colleagues when seeking investment advice. Clearly, investors lacking online social capital will rely more on offline networks for information. Results also show that social capital of offline network is higher than online one, implying offline networks tend to play a more important role in providing investment information. Those with narrow offline social capital approach family members but those with wide offline social capital can also approach close friends and colleagues for advice, or if they want to, in combination with family members. It does confirm the observations made by Fukuyama (1995), Guiso et al. (2004), and Banfield (1958) that low social capital are often characterized by more intense reliance on transactions within narrow subgroups. Therefore, looking from the perspective of social network measured by the social capital index, social network definitely has an influence on investment information search behavior of individual domestic investors. Firstly, investors who are socially more inactive in the offline world (with narrower offline social capital) are more likely to be inward-looking in terms of investment information sources. They are more dependent on family members for investment information. Secondly, investors who have reached out to more people in the online world are also likely to be socially active in the offline world. Investors who have online social capital are also likely to have wider offline social capital. It also demonstrates the intertwining of the online and offline environments for the socially-active investors. The impact of the two networks interacts with one another if the investors are active in both circles. Hence looking at the impact of offline or online social network separately may

200

W.-K. Tan, Y.-J. Tan / Telematics and Informatics 29 (2012) 187–203

not be adequate especially for this group of investors. Thirdly, investors are ready to tap on or exploit their available social networks whenever in need. Hence, those investors with high risk-propensity in investment tap on all kind of information sources available via their offline and online social networks (online friends). Investors who lack in online social capital but with wide offline social capital rely on family, friends and colleagues while those with narrow offline social capital will fall back on their families for advices. Fourth, online social capital influences the use of online friends as investment information source but does not prevent investors lacking online social capital to use non-personal online media as information source. Hence, investors with low online social capital use both offline and online media, in line with the spirit that investors who are in need will use whatever available information sources. A practical implication of this study is the prevalence of interaction effect of investment information sources. Commercial investment information providers and investment tool providers should probe further into this aspect to design the best mix of channels for the provision of investment information to their customers. Since young investors are likely to depend on online information sources (such as rules R1d and R1e of comparison 1, and rules R3a and R3b of comparison 2), online information providers should understand the investment profile and pattern of young people to cater to their needs. Awareness of information websites is high. Operators of information websites can further leverage on the popularity by including more online word-of-mouth to create the buzz and compete with offline personal sources. Operators can also design their websites in such a way that gives readers the impression that the information on the websites originated from many and diversified external sources. While leveraging on Internet as an information source, commercial investment information providers and investment tool providers should not under-estimate the role of traditional mass media since it is an important investment source, especially for older investors (rule R2e of comparison 1). Older investors are also more likely to rely on the physical word-of-mouth of family members, closed friends and colleagues in the offline world (rule R2c of comparison 2 and rule R1f of comparison 3). High-risk investment tool providers targeting at investors with high-risk propensity in investment should be all-rounded in providing investment information since these investors are prone to reach outward and use all available information sources (such as rule R3a of comparison 1, rule R3b of comparison 3 and rule R2e of comparison 4) which they can lay their hands on. These providers should also be careful of the online word-of-mouth spreading in the online world. The results obtained should be read with some caution. The nature of this study was exploratory and intended as a basis for future research. Future studies should strive to address a number of limitations in the present study. The current research has selected only a set of critical criteria for consideration in this study. However, other factors may also need to be considered. Further study can consider other criteria and augment them to this study. The sample size was relatively small. Despite this limitation, this study still provides useful ideas on the complex investment information-search process of investment information consumers and a basis for future research. Furthermore, this study was also conducted in Taiwan and a question arises as to whether the findings can be generalized to other cultures. Given these limitations, it is suggested that future research can utilize the results of this exploratory study by using larger and more diversified sample that includes several countries. The past 2 years have witnessed the explosive growth of social network sites, and users are getting increasingly more involved in them. Many internet users are now using these online sites to connect to others and become members of online virtual communities. In the process, the usage level will not just increase in breadth, it will also deepen. It is likely they will get acquainted with more friends unknown in the offline world. Hence, a longitudinal analysis can be performed in the future to investigate whether the online social capital has increased over time, and to confirm or modify the decision rules obtained in this study.

6. Conclusion This study shows that despite the popularity of social network websites, the social capital of online community is still low when compared to the offline community. Hence, online communities at present play a less limited role in investment informational social support. The interaction effect among investment information sources is explicitly displayed via the rough sets decision rules. The decision rules have presented several insights on the search behavior of investment information consumers as well as practical implications to investment tools providers and investment tool providers. Investors are hybrid investment information consumers and will use or exploits any investment information sources which they can lay their hands on. Investors are ready to tap on or exploit their social networks whenever in need. Older investors, young investors, and investors with highrisk propensity in investment are likely to use a different mix of investment information sources. Investors with low social capital (such as older investors) are often characterized by more intense reliance within narrow subgroups for investment information. Investors with narrow offline social capital approach family members while those with wide offline social capital also approach close friends and colleagues. Investors with online social capital (such as young investors) may use the additional avenue of online friends to tap for advices. Investors who are innovative in technology and transact online are likely to be the users of online sources. Young investors are more likely to seek advice from online friends. Investors, especially those with high-investment risk profile, are indeed true true-bred hybrid information consumers. They tap on all kind of information sources available through their offline and online social networks (online friends). Hence, investment tool providers and commercial investment information providers should probe further into this aspect to design the best mix of channels for the provision of information to their customers.

W.-K. Tan, Y.-J. Tan / Telematics and Informatics 29 (2012) 187–203

201

This is possibly one of the few studies to provide evidences on how the combination of offline and online social capital influences the search behaviors of investment information consumers. This study also includes and combines the concept of technology readiness with social capital. It contributes to the theoretical understanding of the role of social capital in the search behavior of investment information consumers. This is an area which requires more research attention.

References Albrecht, T.L., Adelman, M.B., 1987. Communicating social support: a theoretical perspective. In: Albrecht, T.L., Adelman, M.B. (Eds.), Communicating Social Support. Sage, Newbury Park, CA, pp. 18–39. Anderson, A., Jack, S., 2002. The articulation of social capital in entrepreneurial networks: a glue or a lubricant? Entrepreneurship & Regional Development 14, 193–210. Arndt, J., 1967. Role of product-related conversations in the diffusion of a new product. Journal of Marketing Research 4, 291–295. Arnett, J.J., 2000. Emerging adulthood: a theory of development from the late teens through the twenties. American Psychologist 55, 469–480. Arrow, K., 1974. The Limits of Organization. Norton, New York. Bagozzi, R.P., Dholakia, U.M., 2002. Intentional social action in virtual communities. Journal of Interactive Marketing 16, 2–21. Banfield, E.G., 1958. The Moral Basis of a Backward Society. Free Press, New York. Bansal, H.S., Voyer, P.A., 2000. Word-of-mouth processes within a services purchase decision context. Journal of Service Research 3, 166–177. Barniv, R.R., Cao, J., 2009. Does information uncertainty affect investors’ responses to analysts’ forecast revisions? An investigation of accounting restatements. Journal of Accounting and Public Policy 28, 328–348. Bausch, S., Han, L., 2006. Social networking sites grow 47 percent, year over year, reaching 45 percent of web users. Available from: (accessed 12.12.2008). Beatty, S.E., Smith, S.M., 1987. External search effort: an investigation across several product categories. Journal of Consumer Research 14, 83–95. Bellotti, E., 2008. What are friends for? Elective communities of single people. Social Networks 30, 318–329. Beynon, M.J., Peel, M.J., 2001. Variable precision rough set theory and data discretisation: an application to corporate failure prediction. Omega 29, 561–576. Bian, Y., Ang, S., 1997. Guanxi networks and job mobility in China and Singapore. Social Forces 75, 981–1006. Bickart, B., Schindler, R.M., 2002. Expanding the scope of word of mouth: consumer-to-consumer information on the Internet. Advances in Consumer Research 28, 428–430. Bordia, P., 1997. Face-to-face versus computer-mediated communication: a synthesis of the experimental literature. Journal of Business Communication 34, 99–120. Bourdieu, P., 1985. The forms of capital. In: Richardson, J.G. (Ed.), Handbook of Theory and Research for the Sociology of Education. Greenwood, New York, pp. 241–258. Bourdieu, P., Wacquant, L., 1992. An Invitation to Reflexive Sociology. University of Chicago Press, Chicago, IL. Boyd, D., 2008. Why youth (heart) social network sites: the role of networked publics in teenage social life. In: Buckingham, D. (Ed.), Youth, Identity and Digital Media. The MIT Press, Cambridge, MA. Brucks, M., 1985. The effect of product class knowledge on information search behavior. Journal of Consumer Research 12, 1–16. Burleson, B.R., Albrecht, T.L., Sarason, I., 1994. Communication of Social Support: Messages, Interactions, Relationships and Community. Sage, Newbury Park, CA. Burt, R.S., 1992. Structural Holes: The Social structure of Competition. Harvard University Press, Cambridge, MA. Carroll, G.R., Teo, A.C., 1996. On the social networks of managers. Academy of Management Journal 39, 421–440. Choe, H., Kho, B.C., Stulz, R.M., 2000. Do domestic investors have much more valuable information about individual stocks than foreign investors? Seoul National University, Seoul. Chuang, Y.C., Chuang, K.Y., 2008. Gender differences in relationships between social capital and individual smoking and drinking behavior in Taiwan. Social Science & Medicine 67, 1321–1330. Clark, S.D., 2009. Characterising and predicting car ownership using rough sets. Transportation Research Part C 17, 381–393. Cobb, S., 1976. Social support as a moderator of life stress. Psychosomatic Medicine 38, 301–314. Coleman, J., 1988. Social capital in the creation of human capital. American Journal of Sociology 94, 95–120. Conlon, D.E., Porter, C.O.L.H., McLean Parks, J., 2004. The fairness of decision rules. Journal of Management 30 (3), 329–349. Constant, D., Kiesler, S.B., Sproull, L.S., 1996. The kindness of strangers: the usefulness of electronic weak ties for technical advice. Organization Science 7, 119–135. Deutsch, M., 1962. Cooperation and trust: some theoretical notes. Nebraska Symposium on Motivation 10, 275–318. Donath, J., Boyd, D., 2004. Public displays of connection. BT Technology Journal 22, 71–82. Dvorak, T., 2005. Do domestic investors have an information advantage? Evidence from Indonesia. Journal of Finance 60, 817–839. Ellison, N., Steinfeld, C., Lampe, C., 2006. Spatially bounded online social networks and social capital: the role of Facebook. In: Proceedings of the Annual Conference of the International Communication Association, 2006. Engel, J., Blackwell, R., Miniard, P., 1995. Consumer Behavior, eight ed. Dryden, Fort Worth. Ferlander, S., Timms, D., 2001. Local nets and social capital. Telematics and Informatics 18, 51–65. Ferrary, M., 2003. Trust and social capital in the regulation of lending activities. Journal of Socio-Economics 31, 673–699. Fodness, D., Murray, B., 1999. A model of tourist information search behaviour. Journal of Travel Research 37, 220–230. Freeman, L.C., Ruan, D., 1997. An international comparative study of interpersonal behavior and role relationships. L’Ann´ee Sociologique 47, 89–115. Fukuyama, F., 1995. Trust: The Social Virtues and the Creation of Prosperity. Free Press, New York. Ganley, D., Lampe, C., 2009. The ties that bind: social network principles in online communities. Decision Support Systems 47, 266–274. Ganzeboom, H.B.G., Treiman, D.J., 1996. Internationally comparable measures of occupational status for the 1988 international standard classification of occupations. Social Science Research 25, 201–239. Giddens, A., 1990. The Consequences of Modernity. Stanford University Press, CA. Goh, C., Law, R., 2003. Incorporating the rough sets theory into travel demand analysis. Tourism Management 24, 511–517. Grable, J.E., 2000. Financial risk tolerance and additional factors that affect risk taking in everyday money matters. Journal of Business and Psychology 14, 625–630. Grable, J.E., Lytton, R.H., 1999. Assessing financial risk tolerance: do demographic, socioeconomic and attitudinal factors work? Family Relations and Human Development/Family Economics and Resource Management Biennial, 80–88. Granovetter, M., 1973. The strength of weak ties. American Journal of Sociology 78, 1360–1380. Granovetter, M., 1985. Economic action and social structure: the problem of embeddedness. American Journal of Sociology 91, 481–510. Green, M.C., 2007. Trust and social interaction on the Internet. In: Joinson, A., McKenna, K., Postmes, T., Reips, U.-D. (Eds.), The Oxford Handbook of Internet Psychology. Oxford University Press, New York, pp. 43–51. Grinblatt, M., Keloharju, M., 2000. The investment behavior and performance of various investor types: a study of Finland’s unique data set. Journal of Financial Economics 55, 43–67. Guiso, L., Sapienza, P., Zingales, L., 2004. The role of social capital in financial development. American Economic Review 94, 526–556. Guo, C., 2001. A review on consumer external search: amount and determinants. Journal of Business and Psychology 15, 505–519.

202

W.-K. Tan, Y.-J. Tan / Telematics and Informatics 29 (2012) 187–203

Hair, J.F., Anderson, R.E., Tatham, R.L., Black, W.C., 1998. Multivariate Data Analysis. Prentice Hall, New Jersey. Hallahan, T., Faff, R., McKenzie, M., 2003. An exploratory investigation of the relation between risk tolerance scores and demographic characteristics. Journal of Multinational Financial Management 13, 483–502. Hamilton, G., 1996. The theoretical significance of Asian business networks. In: Hamilton, G. (Ed.), Asian business networks. Walter de Gruyter, New York, pp. 283–298. Haythornthwaite, C., 2007. Social networks and online community. In: Joinson, A., McKenna, K., Postmes, T., Reips, U.-D. (Eds.), The Oxford Handbook of Internet Psychology. Oxford University Press, New York, pp. 121–137. House, J.S., 1986. Social support and the quality and quantity of life. In: Andrew, F.M. (Ed.), Research on the Quality of Life, Survey Research Center, Institute for Social Research. University of Michigan, Ann Arbor, MI. House, J.S., Kahn, R.L., 1985. Measures and concepts of social support. In: Cohen, S., Syme, S.L. (Eds.), Social Support and Health. Academic Press, Orlando, FL, pp. 83–108. Hunter, K., Kemp, S., 2004. The personality of e-commerce investors. Journal of Economic Psychology 25, 529–537. Hwang, K., 1987. Face and favor: the Chinese power game. American Journal of Sociology 92, 944–974. Imai, S., Lin, C.W., Watada, J., Tzeng, G.H., 2008. Rough sets approach to human resource development of information technology corporations. International Journal of Simulation 9, 31–42. Johnson, A.R., Dean, W.W., 2007. Applied Multivariate Statistical Analysis. Prentice Hall, Pearson. Joinson, A.N., 2001. Self-disclosure in computer-mediated communication: the role of self-awareness and visual anonymity. European Journal of Social Psychology 31, 177–192. Kim, K., Pant, P., Yamashita, E.Y., 2008. Hit and run crashes: using rough set analysis with logistic regression to capture critical attributes and determinants, TRB 2008 Annual Meeting. Komorowski, J., Pawlak, Z., Polkowski, L., Skowron, A., 1999. Rough sets: a tutorial. In: Pal, S.K., Skowron, A. (Eds.), Rough Fuzzy Hybridization – A New Trend in Decision-Making. Springer Verlag, Singapore, pp. 3–98. LaRue, E.M., Mitchell, A.M., Terhorst, L., Karimi, H.A., 2010. Assessing mobile phone communication utility preferences in a social support network. Telematics and Informatics 27, 363–369. Lee, M., 2009. Predicting and explaining the adoption of online trading: an empirical study in Taiwan. Decision Support Systems 47, 133–142. Lee, J., Hogarth, J.M., 2000. Relationships among information search activities when shopping for a credit card. Journal of Consumer Affairs 34, 330–360. Lee, R.P.L., Ruan, D., Lai, G., 2005. Social structure and support networks in Beijing and Hong Kong. Social Networks 27, 249–274. Leung, L., Lee, P.S.N., 2005. Multiple determinants of life quality: the roles of Internet activities, use of new media, social support, and leisure activities. Telematics and Informatics 22, 161–180. Levy, J.A., Pescosolida, B.A., 2002. Social Networks and Mental Health. Elsevier, Oxford. Liljandera, V., Gillbergb, F., Gummerusc, J., van Rield, A., 2006. Technology readiness and the evaluation and adoption of self-service technologies. Journal of Retailing and Consumer Services 13, 177–191. Lin, N., 1982. Social resources and instrumental action. In: Marsden, P.V., Lin, N. (Eds.), Social Structure and Network Analysis. Sage, Beverly Hills, pp. 131– 145. Lin, N., 1999. Building a network theory of social capital. Connections 22, 28–51. Lin, N., 2001. Social Capital: A Theory of Social Structure and Action. Cambridge University Press, Cambridge. Lin, W.B., 2008. Investigation on the model of consumers’ perceived risk – integrated viewpoint. Expert Systems with Applications 34, 977–988. Lin, N., Dumin, M., 1986. Access to occupations through social ties. Social Networks 8, 365–385. Lin, Q.C., Lee, J., 2004. Consumer information search when making investment decisions. Financial Services Review 13, 319–332. Lin, N., Fu, Y., Hsung, R., 2001. The position generator: measurement techniques for social capital. In: Lin, N., Cook, K., Burt, R.S. (Eds.), Social Capital: Theory and Research. Aldine De Gruyter, New York, pp. 57–84. Lipsman, A., 2077. Social networking goes global. Available from: (accessed 10.09. 2009). Loibl, C., Hira, T.K., 2009. Investor information search. Journal of Economic Psychology 30, 24–41. Loury, G.C., 1977. A dynamic theory of racial income differences. In: Wallace, P.A., LaMonde, A.M. (Eds.), Women, Minorities and Employment Discrimination. Lexington Books, Lexington, MA, pp. 151–186. MacCrimmon, K.R., Wehrung, D.A., 1986. Taking Risks: The Management of Uncertainty. Free Press, New York. Massey, A.P., Khatri, V., Montoya-Weiss, M.M., 2007. Usability of online services: the role of technology readiness and context. Decision Sciences 38, 277– 308. McKenna, K.Y.A., Bargh, J.A., 2000. Plan 9 from cyberspace. the implications of the Internet for personality and social psychology. Personality and Social Psychology Review 4, 57–75. Milgrom, P., Roberts, J., 1992. Economics, Organization and Management. Prentice Hall, New Jersey. Moore, W.L., Lehmann, D.R., 1980. Individual differences in search behavior for a nondurable. Journal of Consumer Research 7, 296–307. Nagy, R.A., Obenberger, R.W., 1994. Factors influencing individual investor behavior. Financial Analysts Journal 7, 201–233. Nahapiet, J., Ghoshal, S., 1998. Social capital, intellectual capital, and the organizational advantage. The Academy of Management Review 23, 242–266. Oberlechner, T., Hocking, S., 2004. Information sources News and rumors in financial markets: insights into the foreign exchange market. Journal of Economic Psychology 25, 407–424. Onyx, J., Bullen, P., 2000. Measuring social capital in five communities. Journal of Applied Behavioral Science 36, 22–42. Paldam, M., Svendsen, G.T., 2000. An essay on social capital: looking for the fire behind the smoke. European Journal of Political Economy 16, 339–366. Palomino, F., Renneboog, L., Zhang, C., 2009. Information salience investor sentiment, and stock returns: the case of British soccer betting. Journal of Corporate Finance 15, 368–387. Parasuraman, A., 2000. Technology readiness index (TRI): a multiple-item scale to measure readiness to embrace new technologies. Journal of Service Research 2, 307–320. Parasuraman, A., Colby, C., 2001. Techno-ready Marketing: How and Why your Customers Adopt Technology. The Free Press, New York. Pawlak, Z., 1982. Rough sets. International Journal of Computer and Information Science 11, 341–356. Pawlak, Z., 1984. Rough classification. International Journal of Man–Machine Studies 20, 469–483. Pawlak, Z., 1991. Rough Sets, Theoretical Aspects of Reasoning about Data. Kluwer Academic Publishers, Dordrecht/Boston/London. Pawlak, Z., 1997. Rough sets. In: Lin, T.Y., Cercone, N. (Eds.), Rough Sets and Data Mining. Kluwer Academic Publisher, Dordrecht, pp. 3–8. Pawlak, Z., 2002. A primer on rough sets: a new approach to drawing conclusions from data. Cardozo Law Review 22, 1407–1415. Peter, J.P., Ryan, M.J., 1976. An investigation of perceived risk at the brand level. Journal of Marketing Research 13, 184–188. Predki, B., Wilk, S., 1999. Rough set based data exploration using ROSE system. In: Ras, Z.W., Skowron, A. (Eds.), Foundations of Intelligent Systems, Lecture Notes in Artificial Intelligence, 1609. Springer-Verlag, Berlin, pp. 172–180. Predki, B., Slowinski, R., Stefanowski, J., Susmaga, R., Wilk, S., 1998. ROSE – software implementation of the rough set theory. In: Polkowski, L., Skowron, A. (Eds.), Rough Sets and Current Trends in Computing, Lecture Notes in Artificial Intelligence, 1424. Springer-Verlag, Berlin, pp. 605–608. Punj, G.N., Staelin, R., 1983. A model of consumer information search behavior for new automobiles. Journal of Consumer Research 9, 366–380. Putnam, R.D., 1993. Making Democracy Work Civic Traditions in Modern Italy. Princeton University Press, Princeton. Putnam, R.D., 2000. Bowling Alone. Simon and Schuster, New York. Reeves, P.M., 2000. Coping in cyberspace: the impact of Internet use on the ability of HIV-positive individuals to deal with their illness. Journal of Health Communication 5, 47–59. Requena, F., 2003. Social capital, satisfaction and quality of life in the workplace. Social Indicators Research 61, 331–360.

W.-K. Tan, Y.-J. Tan / Telematics and Informatics 29 (2012) 187–203

203

Rice, R.E., Gattiker, U.E., 2001. New media and organizational structuring. In: Jablin, F.M., Putman, L.L. (Eds.), The New Handbook of Organizational Communication. Sage, Thousand Oaks, CA, pp. 544–581. Riegelsberger, J., Sasse, M.A., McCarthy, J.D., 2007. Trust in mediated interactions. In: Joinson, A., McKenna, K., Postmes, T., Reips, U.-D. (Eds.), The Oxford Handbook of Internet Psychology. Oxford University Press, New York, pp. 53–69. Ruzgar, B., Ruzgar, N.S., 2008. Rough sets and logistic regression analysis for loan payment. International Journal of Mathematical Models and Methods in Applied Sciences 2, 65–73. Scharfstein, D.S., Stein, J.C., 1990. Herd behavior and investment. American Economic Review 80, 465–479. Schwarzkopf, D.L., 2007. Investors’ attitudes toward source credibility. Managerial Auditing Journal 22, 18–33. Seasholes, M., 2000. Smart Foreign Traders in Emerging Markets. Harvard Business School, Cambridge, MA. Shen, L., Loh, H.T., 2004. Applying rough sets to market timing decisions. Decision Support Systems 37, 583–597. Sherbourne, C.D., Stewart, A., 1991. The MOS social support survey. Social Science & Medicine 32, 705–714. Shiller, R.J., 1989. Market Volatility. MIT Press, Cambridge, MA. Shyng, J.Y., Wang, F.K., Tzeng, G.H., Wu, K.S., 2007. Rough set theory in analyzing the attributes of combination values for the insurance market. Expert Systems with Applications 32, 56–64. Sproull, L., Kiesler, S., 1986. Reducing social context cues: electronic mail in organizational communication. Management Science 32, 1492–1512. Steinfield, C., Ellison, N.B., Lampe, C., 2008. Social capital, self-esteem, and use of online social network sites: a longitudinal analysis. Journal of Applied Developmental Psychology 29, 434–445. Subrahmanyam, K., Stephanie, M.R., Waechter, N., Espinoza, G., 2008. Online and offline social networks: use of social networking sites by emerging adults. Journal of Applied Developmental Psychology 29, 420–433. Tanis, M., 2007. Online social support groups. In: Joinson, A., McKenna, K., Postmes, T., Reips, U.-D. (Eds.), The Oxford Handbook of Internet Psychology. Oxford University Press, New York, pp. 139–153. Tay, F.E.H., Shen, L., 2002. Economic and financial prediction using rough sets model. European Journal of Operational Research 141, 641–659. Taylor, S.A., Celuch, K., Goodwin, S., 2002. Technology readiness in the e-insurance industry: an exploratory investigation and development of an agent technology e-consumption model. Journal of Insurance Issues 25, 142–165. Taylor, S., Sherman, D., Kim, H., Jarcho, J., Takagi, K., Dunagan, M., 2004. Culture and social support: who seeks it and why? Journal of Personality and Social Psychology 87, 354–362. Thoits, P.A., 1982. Conceptual methodological and theoretical problems in studying social support as a buffer against life stress. Journal of Health and Social Behavior 23, 145–159. Tsikriktsis, N., 2004. A technology readiness-based taxonomy of customers – a replication and extension. Journal of Service Research 7, 42–52. Uchino, B.N., Cacioppo, J.T., Kiecolt-Glaser, J.K., 1996. The relationship between social support and psychological processes: a review with emphasis on underlying mechanisms and implications for health. Psychological Bulletin 119, 488–531. Van Der Gaag, M.P.J., Snijders, T.A.B., 2004. Proposals for the measurement of individual social capital. In: Flap, H., Volker, B. (Eds.), Creation and Returns of Social Capital. Routledge, London, pp. 199–218. Vaux, A., 1988. Social Support: Theory, Research and Intervention. Praeger, New York. Veiel, H.O.F., Baumann, U., 1992. The Meaning and Measurement of Social Support. Hemisphere Publishing Co., New York. Walther, J.B., 1996. Computer-mediated communication: impersonal, interpersonal and hyperpersonal interaction. Communication Research 23, 3–43. Wan, S., Lei, T.C., Huang, P.C., Chou, T.Y., 2008. The knowledge rules of debris flow event: a case study for investigation Chen Yu Lan River. Taiwan, Engineering Geology 98, 102–114. Wellman, B., 1999. The network community: an introduction. In: Wellman, B. (Ed.), Networks in the Global Village: Life in Contemporary Communities. Westview Press, CO, pp. 1–48. Wellman, B., Wortley, S., 1990. Different strokes from different folks: community ties and social support. American Journal of Sociology 96, 558–588. Wellman, B., Quan-Haase, A., Witte, J., Hampton, K., 2001. Does the Internet increase, decrease, or supplement social capital? Social networks, participation, and community commitment. American Behavioral Scientist 45, 437–456. Westbrook, R.A., 1987. Product/consumption-based affective responses and postpurchase processes. Journal of Marketing Research 24, 258–270. Wind, Y., Mahajan, R., Gunther, R., 2002. Convergence Marketing. Prentice Hall, Upper Saddle River, NJ, R.. Wright, K.B., 2002. Social support within an on-line cancer community: an assessment of emotional support, Perceptions of advantages and disadvantages and motives for using the community from a communication perspective. Journal of Applied Communication Research 30, 195–209. Yang, H.H., Wu, C.L., 2009. Rough sets to help medical diagnosis – evidence from a Taiwan’s clinic. Expert Systems with Applications 36, 9293–9298.