Social Media in Virtual Marketing

7/26/2019 Social Media in Virtual Marketing

1/18

Market ForcesCollege of Management Sciences

Vol. X, No. 1June 2015

15Research

Abstract

Social media usage in the world and especially in Pakistan has a high growth due to whichit (social media) has a poten al of becoming an e ff ec ve marke ng tool. Despite its compar-a vely low cost and signi cance, marketers are not e ff ec vely u lizing social media. Thus theaim of this study is to measure the in uence (e ff ect) of four social variables: social capital,trust, homophily and interpersonal in uence on electronic word of mouth (eWOM) communi-ca on. The sample size for the study is 300 and preselected enumerators collected the data

from the leading shopping malls of the city.

Although the scales and measures adopted for this study have been earlier validated in oth-er countries, however the same were re-ascertained on the present set of data. A er prelimi-nary analysis including normality and validity the overall model was tested through StructuralEqua on Model (SEM). This was carried out in two stages - ini ally CFA for all the constructswas ascertained which was followed by CFA of the overall model.

Developed conceptual framework was empirically tested on the present set of data in Pa-kistan which adequately explained consumer a tudinal behavior towards electronic word ofmouth (eWOM) communica on. Three hypotheses failed to be rejected and one was rejected.Trust was found to be the strongest predictor of electronic word of mouth (eWOM) communi-ca on, followed by homophily and social capital. Interpersonal in uence has no rela onshipwith electronic word of mouth (eWOM) communica on. The results were consistent to earlierliterature. Implica on for markers was drawn from the results.

Keywords: eWOM, social capital, trust, homophily and interpersonal in uence, social media

Social Media in Virtual Marketing: Antecedents to Electronic Word ofMouth Communication

Dr. Tariq Jalees 1, Huma Tariq, Syed Imran Zaman , S. Hasnain A. Kazmi Email: [email protected]

1 Corresponding author: Dr. Tariq Jalees is an Associate Professor and HOD Marke ng, College of Management Sciences, Karachi Ins tute ofEconomics and Technology


2/18



16 Research

1. Introduc onSocial marke ng networks due to their

popularity and a high growth trend have be-come an important communica on medium

(S. Li & Li, 2014). But s ll marketers worldover and in Pakistan are not u lizing themeffi ciently and e ff ec vely (Khan & Bha ,2012). Social media is not a subs tute totradi onal adver sing medium, as tradi on-al medium is s ll required to create interestand induce trial (Chu & Choi, 2011) (S. Li &Li, 2014).

Social media coupled with electronicword-of-mouth (eWOM) communica onis very e ff ec ve and e ffi cient in changingconsumers a tude and behavior towardsa product and/or brand (Zhang, Craciun, &Shin, 2010). As compared to word-of-mouthcommunica on(WOM) communica on,electronic word-of-mouth (eWOM) adver-

sing is faster, swi er and has a global reach(Gil de Ziga, Jung, & Valenzuela, 2012). Inview of its signi cance from marke ng per-spec ve, it is important to inves gate the

determinants that a ff ect electronic word-of-mouth (eWOM) communica on (Aiello etal., 2012) (M. Y. Cheung, Luo, Sia, & Chen,2009). Thus the aim of this study is to mea-sure the e ff ect of amophily, social capital, in-terpersonal in uence and trust on electron-ic word-of-mouth communica on (eWOM)by extending the conceptual frameworkdeveloped by Chu (2009) in a non-westernenvironment like Pakistan.

The rest of the paper is structured asfollows; ini ally electronic word-of-mouth(eWOM) is discussed followed by discussions

on the rela onships depicted in the concep-tual framework. Subsequently, methodolo-gy is discussed followed by results contain-ing SEM model and other requried output.

A er discussion and conclusion sec ons li-ma on, and implica ons for marketers arediscussed.

.2. Literature Review2.1. Electronic Word-of-mouth (eWOM)Communica on

Marketers and social scien sts pay spe-cial a en on to interpersonal communi-ca on as it signi cantly changes consumera tude and behavior (C. M. Cheung & Tha-dani, 2010). A bulk of literature is availableon the power of word-of-mouth (WOM)communica on and its e ff ects on brandimage, brand loyalty and purchase inten-

on (Bauernschuster, Falck, & Woessmann,2011) (Gil de Ziga et al., 2012). With theadvent and popularity of social media, ithas also become a medium for the word-of-mouth (WOM) communica on more com-monly known as electronic word-of-mouth(eWOM) (C. M. Cheung & Thadani, 2010).Electronic word-of-mouth (eWOM) commu-nica on refers to all the comments, opinionscommunicated by current, past or poten alusers through social media (Hennig-Thurau,Gwinner, Walsh, & Gremler, 2004).

Tradi onal word-of-mouth (WOM) andelectronic word-of-mouth (eWOM) com-munica on although have some common

a ributes, but they di ff er signi cantly inseveral aspects (C. M. Cheung & Thadani,2010). The communica on process in elec-tronic-word-of-mouth (eWOM) communica-


3/18



17Research

on is swi er and e ff ec ve than tradi on-al word-of-mouth (WOM) communica on.Addi onally, the interac on in tradi onalword-of-mouth(WOM) communica on is

restricted to a small group, whereas in elec-tronic word-of-mouth (eWOM) the audi-ence is large and global (Ste ff es & Burgee,2009). The impact of electronic word-of-mouth (eWOM) communica on is strongerdue to its accessibility. Addi onally, the textbased communica ons remains on the in-ternet archives for a longer period (Hung &Li, 2007) (Sen & Lerman, 2007). Another im-portant aspect of electronic word-of-mouth

(eWOM) communica on is that it can be eas-ily measured and documented (Cha erjee,2001). In case of tradi onal word-of-mouth(WOM) communica on, the creditabilityof the senders can be established whereasno such provision is available in electronicword-of-mouth (eWOM) communica on (C.M. Cheung & Thadani, 2010).

2.2. Conceptual FrameworkPrevious sec on contains a compara ve

discussion on word-of-mouth (WOM) andelectronic word-of-mouth (eWOM) commu-nica ons. In the following sec ons a por onof the conceptual framework developedby Chu (2009) has been used in Pakistansscenario. (Refer to Figure 2.1). The rela on-ships of social capital, trust, homophily andinterpersonal rela onships with electronicword-of-mouth (eWOM) communica onsand derived hypotheses are discussed in the

following sec on.

Figure 2.1 Conceptual Framework

2.2.1. Social Capital and Electronic Word-of-mouth (eWOM)

Social capital refers to social rela onshipsof all the individuals who access social socialmedia sites (Gil de Ziga et al., 2012). It isinclusive of bondage and linkage (Chu & Kim,2011). Higher intensity of bondages and link-ages (social capital) has a stronger in uenceon electronic word of (eWOM) communica-

on (Chu & Kim, 2011) (Stephen & Lehmann,2008). Social media in essence is a socialcommunity mainly used for enhancing busi-

ness, personal and social life (Oh, Labianca,& Chung, 2006) (Putnam, 1993). The sharednorms, exchange of ideas by friends (socialcapital) through social media a ff ects elec-tronic word-of-mouth (eWOM) communica-

on (Bauernschuster et al., 2011) (Bearden& Etzel, 1982a) (Coleman, 2000). While mea-suring the e ff ect of social capital on elec-tronic word-of-mouth (eWOM) communi-ca on, it was found that the rela onship ofof senders and receivers signi cantly a ff ectsthe consumer a tude and behavior in gen-eral and par cularly towards a brand or/andproduct (M. Y. Cheung et al., 2009) (Kiecker &

Homophily

Social Capital

Trust

Inter PersonalInfluence

E Word OfMouth

H1

H3

H2

H4


4/18


18 Research

Cowles, 2002) (Park & Kim, 2009). Addi on-ally, this rela onship also a ff ects consum-ers pre and post evalua on of products andbrands (Goldenberg, Libai, & Muller, 2001)

(Litvin, Goldsmith, & Pan, 2008) (Price & Fe-ick, 1984). Others while elabora ng on theeff ect of social capital on electronic world ofmouth (eWOM) communica on observedthat social media helps their users to ful lltheir needs such as valida ng informa on,building and maintaining social rela onships(Chu & Kim, 2011) (Stephen & Lehmann,2008). Thus it has been hypothesized:

H1: Social capital posi vely eff ects elec-tronic word-of-mouth communica on(eWOM).

2.2.2. Trust and Electronic Word-of-mouth(eWOM) Communica on

Trust is an another cri cal variable thatpromotes electronic word-of-mouth (eWOM)communica on on social media sites (Chu,2009a). In the context of trust, users expectthat social media sites will provide an hon-est, creditable and coopera ve interac on(P. P. Li, 2007) (Rahn & Transue, 1998).

Several studies while exploring the ef-fect of trust on electronic word-of-mouth(eWOM) communica on suggested that ahigher level of trust between consumers andsocial media leads to more interac ve com-munica on (Chu, 2009a) (Pigg & Crank, 2004)(Wasko & Faraj, 2005). Trust towards a social

media site plays a signi cant role in a rac ngconsumers for dissemina on of informa onand knowledge, which in essence is electron-ic word-of-mouth (eWOM) communica on

(Leonard & Onyx, 2003). Consequently theseinterac ons enhance the creditability of so-cial media sites which means a higher e ff ecton electronic word (eWOM) communica on

(Nahapiet & Ghoshal, 1998) (Robert Jr, Den-nis, & Ahuja, 2008). Consumers past expe-rience with a social site also plays a cri calrole in developing and maintaining trust withit (Jansen, Zhang, Sobel, & Chowdury, 2009).Literature also suggests that trust on socialmedia plays a key role in promo ng electron-ic word-of-mouth (eWOM) communica on(Chu, 2009a) (Wasko & Faraj, 2005). Thus thefollowing hypothesis has been generated.

H1: Trust has a posi ve eff ect on electron-ic word-of-mouth communica on (eWOM).

2.2.3. Homophily and Electronic Word-of-mouth (eWOM)

Homophily is an another antecedent thateff ects electronic word-of-mouth (eWOM)communica on on social media (Kawakami,Kishiya, & Parry, 2013). In essence it is thelevel of similarity between message receiv-er, sender and social media (Kawakami etal., 2013) (Kwak, Lee, Park, & Moon, 2010).Homophilous consumers, more o en thannot, voluntarily provide personal informa-

on with the objec ve of developing socialnetworking with individuals that have simi-lar needs, social life style, and consump onbehavior (Aiello et al., 2012) (M. Y. Cheunget al., 2009). Consequently, they feel more

conformable in exchanging advices and in-forma on which of course is an electronicword of (eWOM) communica on. Social me-dia forums such as research, health and en-



5/18


19Research

tertainments have played a signi cant role inpromo ng the rela onship of homophily andelectronic word of (eWOM) communica on(Brown & Reingen, 1987) (Dellande, Gilly, &

Graham, 2004) (Feldman & Spencer, 1965).

Studies on this rela onship found thatthat perceptual homophily has a posi veeff ect and demographic homophily has anega ve e ff ect on electronic word-of-mouth(eEOM) communica on (Gilly, Graham, Wolf-inbarger, & Yale, 1998). Other studies, whileinves ga ng the in uence of homophily onelectronic word-of-mouth (eWOM) com-

munica on found that the creditability andhomophily are the two fundamental aspectswhich consumers consider for selec ng so-cial forum (Wang, Walther, Pingree, & Haw-kins, 2008)

Social networking sites thus are able to at-tract homophlious consumers with commoninterests for conveying product informa-

on and crea ng electronic word-of-moutheWOM communica on (Thelwall, 2009). Lit-erature also suggests that social media userswith a higher level of perceived homophilywill have a stronger par cipa on and e ff ecton electronic word-of-mouth (eWOM) com-munica on (Chu & Choi, 2011) (Chu & Kim,2011). Thus it has been postulated that:

H3: Homophily posi vely eff ects electron-ic word-of-mouth communica on (eWOM).

2.2.4. Interpersonal In uence and Elec-tronic Word-of-mouth (eWOM)

Researchers since decades have suggest-ed that interpersonal in uences signi cantly

aff ect consumers decision making. Thus in-terpersonal in uence also a ff ects consumerbehavior through social media (Bearden &Etzel, 1982b) (DRozario & Choudhury, 2000).

Interpersonal in uence could be norma veor informa ve. Norma ve consumers are in- uenced by the peer groups, whereas infor-ma ve consumers seek informa on from theexperts prior to making their purchase deci-sion (Deutsch & Gerard, 1955).

Consumer vulnerability to interpersonalin uence (Bearden & Etzel, 1982b) plays asigni cant role in explaining social rela on-

ships and electronic word-of-mouth (eWOM)communica on (Chu, 2009a) (McGuire,1968). Norma ve and informa ve in uencedespite being two di ff erent constructs a ff ectelectronic word-of-mouth (eWOM) behav-ior on social media sites, collec vely andindividually (Chu, 2009a). Informa ve con-sumers are generally a racted to those so-cial media sites which transmit informa vevalues, whereas norma ve consumers pre-fer those social media sites which promoterela onship and social networking (Laroche,Kalamas, & Cleveland, 2005).

Thus both norma ve and informa vein uence a ff ects electronic word (eWOM)communica on on social networking sites.Literature also suggests that both informa-

ve and norma ve consumers u lize net-working sites as a media for electronic word(eWOM) communica on (Chu, 2009a) (Laro-

che et al., 2005). Thus it can be argued that:

H4: Interpersonal in uence posi vely ef-fects electronic word-of-mouth communica-



6/18


20 Research

on (eWOM).3. METHODOLOGY

The conceptual framework developed anddiscussed in earlier sec on comprised of fourexogenous models which are social capital,trust, interpersonal in uence, and homoph-ily, and one endogenous model electronicword-of-mouth (eWOM). The methodologyadopted for tes ng the model is discussed inthe following sec ons.

ProcedureThe data was collected by preselected

enumerators though mall intercepts meth-od. This procedure was adopted as the con-sumers who congregate to malls were thetarget audience. The ques onnaire for thesurvey was self-administered. Ini ally, a pre-test of the ques onnaire was carried out tosee the wording, ow of the ques ons andto check social desirability issue. Social desir-ability issue is an imported issue in the Asiancontext, and if not pretested could adverselyaff ects the results. Based on the inputs re-ceived, required rec ca ons were made.

Addi onally, the enumerators a ended atraining session in which the objec ves andpurpose were explained to them and theirqueries were also a ended. The respondedwho par cipated in pretests were not part ofthe main survey.

SampleThree hundred and thirty respondents

of all groups were approached and 300 re-

sponded on voluntary basis. The responserate was 90%. The sample size was higherthan the minimum sample size suggested bysome for studies based on Structural Equa-

on Modeling. (Anderson & Gerbing, 1988).Addi onally for unde ned popula on thesuggested sample size is 285 (Kline, 2005).Thus 300 sample size used in this study isappropriate. In terms of gender 180(60%)were male and 120 (40%) were female andtheir age ranged from 19 to 60 years (M =22.25, SD = 2.78). In terms of marital status,120 (40%) were single and 180 (60%) weremarried. In terms of profession, 90 (30%)were students, 210 (70 %) were employed.In terms of educa on, 90 (30%) had educa-

on up to secondary school cer cate (SSC),105 (35%) had a higher educa on cer cate(HSC), 75 (25%) had bachelors degrees, and

the rest 45 (15%) had at least masters de-gree.

Measures :

Social Capital Scale:Social capital refers to social rela on-

ships in social media sites (Gil de Ziga etal., 2012). Social capital scale in this study isbased on two factors: bridging social capi-

tal (three items) and bonding social capital(three items) all taken from the social capitalmeasure developed by Chu (2009). Reliabil-i es for social capital in previous researchwas .87, and for bonding social capital was.84 (Chu, 2009b). The respondents rated thestatements on a scale of seven (very highagreement) and one (very low agreement).Average mean score of the six items re ectsrespondents level of social capital.

Trust ScaleTrust refers to expecta on of honest and

coopera ve behavior that conforms to the



7/18


21Research

norms of the community (Rahn & Transue,1998). Trust measure (scale) for this paperhas been adopted from trust measure (scale)developed by Chu (2009). The reliability of

the trust measure was 0.93 (Chu, 2009b).The respondents rated the statements ona scale of seven (very high agreement) andone (very low agreement). Average meanscore of the six items re ects respondentslevel of trust scale.

Homophily Scale:Homophily refers similarity and traits

and a ributes between individuals who in-

teracts with each other (Aiello et al., 2012).Homophily scale in this study has four fac-tors a tude, background and morality, andappearance. In all there were eight items inhomophily scale two from each factor, all ad-opted from the measure(scale) developed byChu (2009). The reliability of the homophilyscale ranged 0.85 to 0.89 (Chu, 2009b). Therespondents rated the statements on a scaleof seven (very high agreement) and one(very low agreement). Average mean scoreof the eight items re ects respondents levelof homophily.

Interpersonal In uenceIn uence of others refers to norma ve

(peers) and informa ve (experts) in u-ences on consumers a tude and behav-ior. (Bearden & Etzel, 1982b) (DRozario &Choudhury, 2000). Interpersonal in uencescale for this study has been adopted from

from the measure (scale) developed by (Chu,2009b). The scale for this study has two fac-tors which are informa ve (three items) andnorma ve (three items). Reliability of the

Interpersonal in uence in previous researchranged 0.94 to 0.94 (Chu, 2009b). The re-spondents rated the statements on a scale ofseven (very high agreement) and one (very

low agreement). Average mean score of theeight items re ects respondents level of in-terpersonal in uence.

Electronic Word-of-mouth (eWOM) ScaleElectronic word-of-mouth (eWOM) for

the present study has three factors whichare opinion leadership, opinion seeking andpass along behavior with six items all takenfrom the measure developed by Chu (2009).

Reliability of the Electronic word-of-mouth(eWOM) ranged 0.68 to 0.93 (Chu, 2009b).The respondents rated the statements ona scale of seven (very high agreement) andone (very low agreement). Average meanscore of the eight items re ects respondentslevel of electronic word-of-mouth (eWOM)communica on.

Data Analysis TechniqueTwo so ware SPSS-v19 and AMOS-v18

have been used in this study. The former hasbeen used for reliability, descrip ve and nor-mality analyses and the later for tes ng theendogenous model and derived hypotheses(D. Byrne, London, & Reeves, 1968) (Cabal-lero, Lumpkin, & Madden, 1989).The bene tof using Structural Equa on Model (SEM) isthat it has the capacity for assessing theo-ries and tes ng derived hypotheses simul-taneously (Hair Jr, Black, Babin, Anderson,

& Tatham, 2010). The tness of the modelwas improved based on the following crite-ria: Standardized Regression Weight of la-tent variables 0.40; Standardized Residual



8/18


22 Research

Covariance < 2.58 and Modi ca on Index 9.0 > 0.9 > 0.95 > 0.50 > 0.50

Note. 2 = Chi Square; 2/df= Relative Chi Sq; CFI= Comparative Fit Index, NFI- Normed FixedIndex; IFI= Incremental Fixed Index, PNFI= Parsimonious Fit Index, PCFI is Parsimonious Fit Index.

Table 4.2 Descriptive and Reliability of Initial Constructs

Measures Mean Std. Dev. Skewness Kurtosis Variance Reliability Social Capital 3.48 1.06 -0.70 -0.28 1.12 0.96

Trust 3.55 0.84 -0.71 0.84 .70 0.88

Homophily 3.58 0.96 -0.40 2.40 .93 0.89

Inter Per. Influence 3.70 0.88 -0.82 0.58 .78 0.92Elect. Word of Mouth 3.60 0.92 -1.03 0.68 .85 0.94


9/18


23Research

of univariate normality (B.M Byrne, 2001)(Hair Jr et al., 2010).

Bivariate Correla on

Inter item correla on was carried out tocheck whether the variables are separateand dis nct concepts or not. The summa-rized results depicted in Table 4.3 show thatnone of the inter-item correla on is greaterthan 0.90 (Kline, 2005) thus indica ng thatall the variables/ constructs used in thisstudy are separate and dis nct and do nothave Mul collinearity issues.

Construct ValidityConstruct validity is necessary if instru-ment developed in one country is adoptedand administered in other country (Bhard-waj, 2010). Since the instrument used in thisstudy has also been adopted therefore con-struct validity has been ascertained throughconvergent and discriminant validity (Bhard-waj, 2010). CFA results (Refer to Table 4.5)show that most of indices outputs exceedprescribed criteria. Addi onally the factor

loading of all indicator variables loading areat least 0.40 (Refer to Figure 4.2). Thus it isinferred that the data ful ll convergent va-lidity requirements (Hsieh & Hiang, 2004)

(Shammout, 2007).

Uniqueness of the variables was test-ed through Discriminant validity (Hair et al.2010) by comparing the square root of aver-age variance extracted (AVE) with the squarecorrela on coe ffi cient. The summarizedresults depicted in Table 4.4 show that thevalues of average variance extracted is lesserthan square of all possible pairs of constructs

therefore the variables are unique and dis-nct (Fornell & Larcker, 1981)

1) Diagonal entries show the square-rootof average variance extracted by the con-struct (2) O ff -diagonal entries represent thevariance shared (squared correla on) be-tween constructs

Con rmatory Factor AnalysisIn CFA the factors and items (indicators)


Table 4.3 Correlation

SC_T IN_T T_T HT_T EW_TSocial Capital 1.00

Int. Influence 0.46 1.00

Trust 0.55 0.61 1.00

Homophily 0.47 0.48 0.58 1.00

Electronic Word of Mouth 0.63 0.54 0.70 0.71 1.00

**. Correlation is significant at the 0.01 level (i-tailed)


10/18


24 Research

are tested based on theory therefore it is alsoknown as a test for measuring theories (Hairet al, 2006, p. 747). The summarized CFA re-sults of the four constructs are presented inTable 4.4.

Table 4.5 above shows that the t indicesexceed the prescribed criteria. Addi onal-ly, factor loading of indicator variables aregreater than 0.40 and standardized residu-

als are below 2.58 con rming the tness ofeach CFA model (Hair Jr. et al., 2007).

Overall ModelThe tested model has four exogenous

variables including social capital, trust, ho-mophily, and interpersonal in uence andone endogenous variable electronic word-of-mouth communica on (eWOM) (Refer toFigure 4.1)


Table 4.4 Discriminant Validity

SC_T IN_T T_T HT_T EW_TSocial Capital 0.75

Int. Influence 0.21 0.81

Trust 0.30 0.37 0.82

Homophily 0.22 0.23 0.34 0.81

Elect Word of Mouth 0.40 0.29 0.49 0.50 0.74

Table 4.5 Confirmatory Factor Analysis

Absolute Relative Parsimonious

2 2/df DOF(p) CFI NFI IFI PNFI PCFISocial Capital 5.058 5.058 1(0.025) 0.980 0.960 0.980 0.325 0.327

Trust 4.979 2.490 2(0.083) 0.990 0.984 0.990 0.328 0.330

Homophily 4.216 2.198 2(0.121) 0.995 0.990 0.995 0.330 0.335

Int. Influence 6.751 3.376 2(0.034) 0.992 0.992 0.992 0.330 0.331

e.Word of Mouth 28.612 5.772 5(0.006) 0.997 0.973 0.997 0.586 0.589

Criteria Low < 5.0 n/a > 9.0 > 0.9 > 0.95 > 0.50 > 0.50

Note. 2 = Chi Square; 2/df=; DOF(p)= Degree of Freedom and probability, CFI= Comparative FitIndex, NFI- Normed Fixed Index; IFI= Incremental Fixed Index, PNFI= Parsimonious Fit Index, PCFIis Parsimonious Fit Index.


11/18


25Research

Figure 4.1 for the overall model showsthat each factor loading of each observedvariable is atleast 0.40 and standardized re-sidual are within the range of 2.58 (Hair Jr.,Anderson, Tatham, & Black, 2007). Addi on-ally all the t indices exceed the prescribedcriteria as discussed in the following para-graph.

The Chi Square value ( 2 = 175.325, DF= 94, p= 0.003 < .05), is signi cant, and 2/df (rela ve) was 1.865 < 5. These resultsmeet the absolute criteria. Rela ve t in-

dices are also within the prescribed limit (CFI = 0.975 > 0.900; and NFI = 0.948 > 0.900and IFI=0.975>=.95). Parsimony AdjustedNormed Fit Indices are also meet the pre-scribed criteria (PNFI =0.743 > 0.50 and PCFI= 0.764 > 0.50. Thus the CFA results con rmsthat the overall hypothesized model is agood t.

Hypothesized ResultsThe summarized SEM output in the con-

text of regression weight is depreciated inTable 4.6.


Figure 4.1

Final SEM Model


12/18


26 Research

Table 4.6 shows that trust (M= 4.71, SD=1.55, SRW= 0.512, CR= 2.640, P= 0.008 0.05).This nding is consistent to earlier literature.For example several studies while validat-

ing the e ff ect of social capital on electronicword-of-mouth (eWOM) communica on,found that the rela onship of of senders andreceivers signi cantly a ff ects the consumer

a tude and behavior in general and par c-ularly towards a brand or/and product (M. Y.Cheung et al., 2009) (Kiecker & Cowles, 2002)(Park & Kim, 2009). Addi onally, studies sug-gested that the rela onship of social capi-tal and electronic world of mouth (eWOM)also a ff ects consumers pre and post evalu-a on of products and brands (Goldenberg etal., 2001) (Litvin et al., 2008) (Price & Feick,1984). Others while elabora ng on the e ff ectof social capital on electronic world of mouth(eWOM) communica on observed that so-cial media helps their users to ful ll theirneeds such as valida ng informa on, build-ing and maintaining social rela onships (Chu& Kim, 2011) (Stephen & Lehmann, 2008)

Hypothesis (two) on the e ff ect of trust(M= 3.55, SD= 0.84) and electronic word-of-mouth (eWOM) communica on (M=3.60, SD= 0.92) failed to be rejected (SRW=0.512, CR= 2.640, P= 0.008 eWOM .175 .055 3.157 .002

Trust ---------------> eWOM .512 .194 2.640 .008

Homophily ---------------> eWOM .241 .095 2.546 .011

I. Influence ---------------> eWOM .061 .091 .667 .505

*Standardized Regression Weight


13/18


27Research

several studies while valida ng the e ff ect oftrust on electronic word-of-mouth (eWOM)communica on suggest that a higher levelof trust between consumers and social me-

dia leads to more meaning full interac on,(Chu, 2009a) (Leonard & Onyx, 2003) (Pigg &Crank, 2004) (Wasko & Faraj, 2005). Conse-quently these interac ons enhance the cred-itability of social media sites which meansa higher e ff ect on electronic word (eWOM)communica on (Nahapiet & Ghoshal, 1998)(Robert Jr et al., 2008). Studies also suggestthat consumes past experience with a socialsite also plays a cri cal role in developing and

maintaining trust and promo ng electron-ic word-of-mouth (eWOM) communica on(Chu, 2009a) (Jansen et al., 2009) (Wasko &Faraj, 2005)

Hypothesis (three) on the e ff ect of ho-mophily (M= 3.58, SD= 0.96) and electron-ic word-of-mouth (eWOM) communica on(M= 3.60, SD= 0.92) failed to be rejected(SRW= 0.241, CR= 2.546, P= 0.011.05. This nding iscontrary to the literature and earlier studies.Studies and literature suggests that consum-er vulnerability to interpersonal in uence(Bearden & Etzel, 1982b) plays a signi cant

role in explaining social rela onships andelectronic word-of-mouth (eWOM) commu-nica on (Chu, 2009a) (McGuire, 1968). Liter-ature also suggest that both norma ve andinforma ve in uence a ff ect electronic word(eWOM) (Chu, 2009a) (Laroche et al., 2005).

6. ConclusionThis model on antecedents to electron-

ic word-of-mouth (eWOM) communica onempirically tested through SEM will helpthe in understanding consumers a tudeand behavior towards this new medium ofcommunica on. This new medium and es-pecially electronic word-of-mouth (eWOM)has brought challenges and opportuni es tothe marketer. Of the four hypotheses threefailed to be rejected and one was rejected.Trust was found to be the strongest predic-tor of electronic word-of-mouth (eWOM)communica on, followed by homophily andsocial capital. Interpersonal in uence has norela onship with electronic word-of-mouth(eWOM) communica on.



14/18


28 Research

Implica ons for MarketersThree of the social factors social capital,

amophily, and trust have posi ve impacton the electronic word-of-mouth (eWOM)communica on. Thus the marketer mustconcentrate in developing social networkingsites which are able to a ract homophhliousconsumers with common product interests(Thelwall, 2009). Since this is not possiblewith one social media site so they shoulddevelop hyperlinks of di ff erent forums toinduce par cipa on and exchange of infor-ma on which lead to social capital (bondingand linkages). Di ff erent hyperlinks of di ff er-ent forum will help the diversi ed consum-

ers to a ract homophilous consumers. Themore individuals go on these social sites the

trust and creditability will also increase(Chu,2009b) (Khan & Bha , 2012)

Limita on and Future ResearchThis study was limited to higher income

group of Karachi. Individuals behavior in thecontext of social capital, amophily, trust andinterpersonal rela onship may vary from de-mographic which could be incorporated infuture studies. This study is restricted to theeff ect of social variables on electronic word-of-mouth (eWOM) communica on. Futurestudies could measure the e ff ect of the vari-ables used in this study on a tude and be-havior towards brands, product category and

adver sements. Incorpora on of culture andmul -cultural study could also be explored.



15/18


16/18


30 Research

in uence. Journal of Consumer Marke ng, 17(4), 290-307.

Dellande, S., Gilly, M. C., & Graham, J. L. (2004). Gaining compliance and losing weight: the role of the service provider in health care services. Journal of Marke ng, 68(3), 78-91.

Deutsch, M., & Gerard, H. B. (1955). A study of norma ve and informa onal social in uences upon individual judgment. The journal of abnormal and social psychology, 51(3), 629.

Feldman, S. P., & Spencer, M. C. (1965). The E ff ect of Personal In uence on the Selec on of Consumer Services:Center for regional studies.

Fornell, C., & Larcker, D. F. (1981). Evalua ng structural equa on models with unobservable variables andmeasurement error. Journal of Marke ng research, 39-50.

Gil de Ziga, H., Jung, N., & Valenzuela, S. (2012). Social media use for news and individuals social capital, civicengagement and poli cal par cipa on. Journal of Computer-Mediated Communica on, 17(3), 319-336.

Gilly, M. C., Graham, J. L., Wol nbarger, M. F., & Yale, L. J. (1998). A dyadic study of interpersonal informa onsearch. Journal of the Academy of Marke ng Science, 26(2), 83-100.

Goldenberg, J., Libai, B., & Muller, E. (2001). Talk of the network: A complex systems look at the underlying process of word-of-mouth. Marke ng le ers, 12(3), 211-223.

Hair Jr, J., Black, W., Babin, B., Anderson, R., & Tatham, R. (2010). Mul variate data analysis New Jersy: PearsonEduca on.

Hair Jr., J. F., Anderson, R. E., Tatham, R. L., & Black, W. C. (2007). Mul variate data Analysis (6th Ed.). . New Jersey: Pren ce Hall Interna onal, Inc.

Hennig-Thurau, T., Gwinner, K. P., Walsh, G., & Gremler, D. D. (2004). Electronic word-of-mouth via consumer-opinion pla orms: What mo vates consumers to ar culate themselves on the Internet? Journal of interac vemarke ng, 18(1), 38-52.

Hsieh, Y. C., & Hiang, S. T. (2004). A study of the impacts of service quality on rela onship quality in search-experience-credence services. Total Quality Management and Business Excellence, 15(1), 43-58.

Huang, J. H., Lee, B. C. Y., & Ho, S. H. (2004). Consumer a tude toward gray market goods. Interna onalMarke ng Review, 21(6), 598-614.

Hung, K. H., & Li, S. Y. (2007). The in uence of eWOM on virtual consumer communi es: social capital, consumerlearning, and behavioral outcomes. Journal of Adver sing Research, 47(4), 485.

Jansen, B. J., Zhang, M., Sobel, K., & Chowdury, A. (2009). Twi er power: Tweets as electronic word of mouth. Journal of the American Society for Informa on Science and Technology, 60(11), 2169-2188.

Joreskog, K. G., & Sorbom, D. (1988). Prelis. A Program for Mul variate Data Screening and Data Summariz on.Users Guide. Chicago, Scien c So ware Interna onal.

Kawakami, T., Kishiya, K., & Parry, M. E. (2013). Personal word of mouth, virtual word of mouth, and innova onuse. Journal of Product Innova on Management, 30(1), 17-30.

Khan, S. A., & Bha , R. (2012). Applica on of social media in marke ng of library and informa on services: Acase study from Pakistan. Webology, 9(1), 1-8.

Kiecker, P., & Cowles, D. (2002). Interpersonal communica on and personal in uence on the Internet: A framework for examining online word-of-mouth. Journal of Euromarke ng, 11(2), 71-88.



17/18


31Research


Kline, T. (2005). B, Sulsky, LM, & Rever-Moriyama, SD (2000). Common method variance and speci ca on errors: A prac cal approach to detec on. Journal of Psychology, 134, 401421.

Kwak, H., Lee, C., Park, H., & Moon, S. (2010). What is Twi er, a social network or a news media? Paper presentedat the Proceedings of the 19th interna onal conference on World wide web.

Laroche, M., Kalamas, M., & Cleveland, M. (2005). I versus we: How individualists and collec vists useinforma on sources to formulate their service expecta ons. Interna onal Marke ng Review, 22(3), 279-308.

Leech, N. L., Barre , K. C., & Morgan, G. A. (2005). SPSS for intermediate sta s cs: Use and interpreta on:Lawrence Erlbaum Assoc Inc.

Leonard, R., & Onyx, J. (2003). Networking through loose and strong es: an Australian qualita ve study.Voluntas: Interna onal Journal of Voluntary and Nonpro t Organiza ons, 14(2), 189-203.

Li, P. P. (2007). Social e, social capital, and social behavior: Toward an integra ve model of informal exchange. Asia Paci c Journal of Management, 24(2), 227-246.

Li, S., & Li, J. (2014). Web and social media dynamics, and evolu onary and adap ve branding: theories and ahybrid intelligent model. Paper presented at the 13th interna onal conference on ar cial intelligence, knowledgeengineering and data bases, , Poland.

Litvin, S. W., Goldsmith, R. E., & Pan, B. (2008). Electronic word-of-mouth in hospitality and tourism management.Tourism management, 29(3), 458-468.

McGuire, W. J. (1968). Personality and suscep bility to social in uence. In E. F. Borga a & William W. Lambert(Eds.), Handbook of personality theory and research (pp. 1130-1187). Chicago: Rand McNally.

Nahapiet, J., & Ghoshal, S. (1998). Social capital, intellectual capital, and the organiza onal advantage. Academyof Management Review, 23(2), 242-266.

Oh, H., Labianca, G., & Chung, M.-H. (2006). A mul level model of group social capital. Academy of ManagementReview, 31(3), 569-582.

Park, D.-H., & Kim, S. (2009). The e ff ects of consumer knowledge on message processing of electronic word-of-mouth via online consumer reviews. Electronic Commerce Research and Applica ons, 7(4), 399-410.

Pigg, K. E., & Crank, L. D. (2004). Building community social capital: The poten al and promise of informa onand communica ons technologies. The Journal of Community Informa cs, 1(1).

Price, L. L., & Feick, L. F. (1984). The role of interpersonal sources in external search: an informa onal perspec ve. Advances in consumer research, 11(1), 250-255.

Putnam, R. (1993). The prosperous community: social capital and public life. The american prospect, 13(Spring),Vol. 4. Available online: h p://www. prospect. org/print/vol/13 (accessed 7 April 2003).

Rahn, W. M., & Transue, J. E. (1998). Social trust and value change: The decline of social capital in Americanyouth, 19761995. Poli cal Psychology, 19(3), 545-565.

Robert Jr, L. P., Dennis, A. R., & Ahuja, M. K. (2008). Social capital and knowledge integra on in digitally enabledteams. Informa on Systems Research, 19(3), 314-334.

Sen, S., & Lerman, D. (2007). Why are you telling me this? An examina on into nega ve consumer reviews onthe web. Journal of interac ve marke ng, 21(4), 76-94.

Shammout, A. B. (2007). Evalua ng an Extended Rela onship Marke ng Model for Arab Guests of Five-StarHotels. (PhD Disserta on), University Melbourne, Victoria


18/18

Ste ff es, E. M., & Burgee, L. E. (2009). Social es and online word of mouth. Internet Research, 19(1), 42-59.

Stephen, A. T., & Lehmann, D. R. (2008). Recipient characteris cs and product-related-transmission: the role ofsocial capital. Available at SSRN 1150996.

Thelwall, M. (2009). Homophily in myspace. Journal of the American Society for Informa on Science andTechnology, 60(2), 219-231.

Wang, Z., Walther, J. B., Pingree, S., & Hawkins, R. P. (2008). Health informa on, credibility, homophily, andin uence via the Internet: Web sites versus discussion groups. Health Communica on, 23(4), 358-368.

Wasko, M. M., & Faraj, S. (2005). Why should I share? Examining social capital and knowledge contribu on inelectronic networks of prac ce. MIS quarterly, 35-57.

Zhang, J. Q., Craciun, G., & Shin, D. (2010). When does electronic word-of-mouth ma er? A study of consumer product reviews. Journal of Business Research, 63(12), 1336-1341.


32 Research


Documents

Social Media in Virtual Marketing