Investigating Electronic Word-of-Mouth Effects on Online Discussion Forums: The Role of Perceived Positive Electronic Word-of-Mouth Review Credibility

Abstract

Electronic word of mouth (eWOM) has been an important factor influencing consumer purchase decisions. Using the ABC model of attitude, this study proposes a model to explain how eWOM affects online discussion forums. Specifically, we propose that platform (Web site reputation and source credibility) and customer (obtaining buying-related information and social orientation through information) factors influence purchase intentions via perceived positive eWOM review credibility, as well as product and Web site attitudes in an online community context. A total of 353 online discussion forum users in an online community (Fashion Guide) in Taiwan were recruited, and structural equation modeling (SEM) was used to test the research hypotheses. The results indicate that Web site reputation, source credibility, obtaining buying-related information, and social orientation through information positively influence perceived positive eWOM review credibility. In turn, perceived positive eWOM review credibility directly influences purchase intentions and also indirectly influences purchase intentions via product and Web site attitudes. Finally, we discuss the theoretical and managerial implications of the findings.

Introduction

The impact of word of mouth (WOM) on a consumer's decision-making process has grown with the development of the Internet.^1,2 The characteristics of the Internet, such as immediacy and interaction, facilitate consumers' information searches and communications with others.³ WOM communication in virtual communities is much more influential in terms of speed and scope than ever. Due to an increase in user popularity, social networking sites (SNS) have become resources and new platforms for marketers to disseminate and exchange information with their targeted population. Electronic word of mouth (eWOM) has become one of the important routes for consumers to obtain product information. Recent research shows that 80% of consumers trust online recommendations. This percentage is far beyond that of consumers who trust traditional advertisements.⁴ Investigating the factors that influence eWOM credibility will be of great interest to scholars and practitioners.

Consumer motives relevant to traditional WOM can also be expected to be of relevance for eWOM.³ Prior research has identified speakers' motives for positive WOM,^5,6 as well as characteristics of WOM speakers and listeners of positive eWOM usage.⁷ Few studies have identified the factors that influence listeners' perceived positive eWOM review credibility (PPERC) given its practical and academic interests and importance. The WOM literature has also called for research in understanding whether and how WOM influences listeners' behaviors.^8,9 In order to fill these research gaps, based on the C-A-B hierarchy of the ABC model of attitude, we propose a model that identifies the factors that influence PPERC (cognition) and how it impacts purchase intentions (behavior) via product and Web site attitudes (affect). Specifically, we propose that platform-based (Web site reputation, WR, and source credibility, SC) and customer-based (obtaining buying-related information, OBRI, and social orientation through information, SOTI) motivations influence PPERC. In addition, product and Web site attitudes (PAs and WAs) mediate the effects of PPERC on purchase intentions (PIs).

Literature Review and Hypotheses Development

eWOM refers to user-generated content on the Internet.¹⁰ It differs from traditional WOM in several aspects, such as information amount and format.¹¹ The development of the Internet has provided eWOM listeners with great sources of product information for purchase considerations³, such as product review Web sites.¹²

Research in WOM has focused on investigating the WOM speakers' motives for spreading positive WOM reviews, such as product involvement, self-involvement, and other involvement.^5,6,13 Researchers have identified the characteristics of WOM listeners and WOM speakers that influence WOM usage.⁷ From an eWOM listener's perspective, this research investigates the factors that lead to PPERC and its impact on PIs. Based on the ABC model of attitude, being formed by affect, behavior, and cognition,^14,15 we propose a model that integrates the antecedents and consequences of PPERC. In regard to cognition, this research proposes platform- and customer-based motivations as the antecedents of PPERC. The platform-based motivations include WR and SC, as purchase decisions have been deeply influenced by the Internet. WR and SC are likely to influence PPERC.¹⁶ Consumers seek to minimize their time spent and efforts used to search for product information.^17,18 They also look for opportunities to socialize with others via the Internet.⁴ Thus, customer-based motivations include OBRI and SOTI as antecedents to PPERC. Following the ABC model of attitude, consumers' cognitive processing of information influences their affect and then behavior. This research proposes that PPERC will influence PIs via PAs and WAs.

The proposed model is presented in Figure 1. Previous research has indicated that listeners respond to positive and negative WOM reviews differently.^19,20 eWOM is a double-edged sword, as it can make or break a product.²¹ Although some studies have shown that the effects of negative eWOM reviews are stronger than positive ones,²² other studies have found the opposite.^23–25 Some researchers have found that the effect of the review valence on persuasiveness depends on individual consumption goals²⁶ and product category.²⁷ Nevertheless, both positive and negative eWOM reviews are important to product or service marketing. This research focuses only on listeners' responses to positive eWOM reviews for the several reasons. First, researchers have indicated that positive eWOM reviews are more likely to occur than negative ones.^22,28 Second, studies have shown that the motive of self-enhancement leads consumers to generate positive WOM,²⁹ and moderately positive WOM reviews are positively related to review value.³⁰ Third, positive eWOM reviews are likely to increase customers' PIs³¹ because they reduce the risks involved in the purchase.⁵ Fourth, positive WOM reviews can help create favorable company and brand images, which eventually reduce promotional expenditures.^32,33 Therefore, it is essential to examine the effects of positive eWOM reviews.

FIG. 1.

Proposed model.

The impact of WR and SC on PPERC

A good reputation is a powerful means for persuasion.³⁴ Consumers usually use reputations to infer the credibility of information received.^15,34–36 A good WR will, in turn, usually have high trust from its consumers.³⁷ Consumers can infer the quality of the contents of a Web site based on its reputation. Therefore, a reputable Web site is more readily accepted by consumers than a lesser known Web site.³⁸ Similarly, the eWOM effect is higher for Web sites with established reputations than for those without established reputations (i.e., consumers are more likely to trust eWOM reviews posted on reputable Web sites).²⁰ Thus, when consumers perceive that a Web site has a high reputation, then they will consider the positive eWOM reviews on the Web site as more credible.

Given the substantial amount of information on the Internet, consumers rely on source credibility to distinguish between Web sites with and without reputable content.³⁹ WOM speakers with positive characteristics are more persuasive than those with fewer positive characteristics.⁴⁰ The credibility of the WOM speakers themselves is an important antecedent to the credibility of WOM review.⁴¹ For instance, expert recommendations are more helpful and readily accepted in making decisions due to their high level of credibility.⁴² The same SC effect could be applied to online contexts.⁴³ When eWOM speakers are more credible, consumers are more likely to trust them.⁴⁴ Therefore, when the source of the eWOM review is more credible, consumers perceive the positive eWOM review as also being more credible.

H1: WR has a positive effect on the PPERC.

H2: SC has a positive effect on the PPERC.

The impact of OBRI and SOTI on PPERC

The purpose of OBRI is to reduce purchase risk and information search time.⁴⁵ Prior research has shown that more than 80% of the individuals who provide product reviews in online discussion forums intend to help others make decisions.⁴⁶ Competent reviewers provide consumers with useful information, which increases the PPERC.⁴⁷ Thus, consumers' motivations for reading eWOM reviews influence their perceptions of the review credibility. When consumers are motivated to search for information on a SNS, they tend to perceive the site as being more credible,⁴⁸ and such a motivation influences eWOM review credibility.^49–51 Researchers have also found that information seekers perceive online reviews as being more credible than noninformation seekers.⁵² It is expected that when consumers have a higher level of motivation to OBRI, they have a higher PPERC.

Prior research has indicated that interpersonal influence is associated with eWOM effects on SNS.⁵³ Research shows that tie strength positively influences the effect of WOM reviews on consumer decisions.⁴⁹ Consumers trust WOM reviews more if the WOM reviews are from their family or close friends.^54,55 Thus, we propose that SOTI impacts PPERC. Interactions with and obtaining confirmation from other users can maintain an individual's social status and reduce cognitive dissonance with regard to purchase decisions.⁵⁶ Consumers perceive reviews from other users as more credible.⁵⁷ Researchers also indicated that an information receiver's social location in a virtual community influences his perception of the eWOM review credibility.³⁹ When the goals of social status and dissonance reduction can be achieved through online eWOM review reading, consumers perceive the eWOM reviews as more credible. Therefore, SOTI should have a positive impact on PPERC.

H3: OBRI has a positive effect on the PPERC.

H4: SOTI has a positive effect on the PPERC.

The impact of PPERC on PIs—the mediating role of PA and WA

PA is defined as an eWOM review reader's evaluation of a product and PIs as the likelihood that a consumer will make a purchase.⁵⁸ WA is defined as an eWOM review reader's evaluation of a Web site that provides a platform for posting eWOM reviews.³⁸ Following the C-A-B hierarchy of the ABC model of attitude, PAs and WAs are proposed to mediate the effects of PPERC on PIs. That is, PPERC (cognition) will influence PIs (behavior) via both PAs and WAs (affect). After reading eWOM reviews, consumers make judgments on the eWOM reviews' credibility.^44,59 When consumers perceive positive WOM reviews as more credible, they form more positive PAs.⁶⁰ Consumers' attitudes toward reviews are transferred to the product. The same effect of traditional WOM occurs with positive eWOM reviews.²⁰ Researchers indicated that accessing product reviews from online discussions influences consumers' PAs.^61,62 When positive eWOM reviews are perceived as being credible, then consumers are more likely to form favorable PAs. It is expected that PPERC has a positive effect on consumers' PAs. Previous research has shown that product reviews from online discussions increase consumers' interests in the product topics.⁶¹ After receiving product information, consumers develop positive PAs when they perceive it as having a high value and, therefore, will increase their PIs.^63,64 Research showed that eWOM review readers' PAs influenced their PIs and behaviors.⁴⁴ Therefore, it is expected that consumers' PAs will be positively associated with PIs.

H5: PPERC has a positive effect on PAs.

H6: PAs have a positive effect on PIs.

Similar to the effect of PPERC on PAs,^2,50,65 consumers' attitudes toward WOM reviews could be transferred to Web sites. When consumers perceive eWOM reviews on a Web site as being credible, they are more likely to trust and form favorable attitudes toward the sponsored Web site.⁶⁶ Researchers have shown that eWOM review credibility influences consumers' adoptions of eWOM reviews on a Web site.^26,67 When consumers perceive eWOM reviews as being more credible, they are more likely to adopt the eWOM review recommendations. Researchers also indicated that eWOM review favorableness impacted consumers' WAs.⁶⁸ When consumers perceive positive eWOM reviews as being more credible, they are more likely to form positive WAs. Thus, it is expected that PPERC will have a positive effect on WAs. In addition, previous research has indicated that consumers' WAs are highly associated with their PIs. For example, researchers have shown that interactions between consumers and firms in electronic commerce mainly take place through the company Web sites. When consumers are satisfied with a Web site's design, product information, and services, they are more likely to make a purchase on that Web site.⁶⁹ Some researchers have demonstrated that using technology to enhance interactions positively influenced consumers' WAs and subsequently positively influenced their PIs.⁴⁶ Similar findings of the positive effects of WAs on PIs were observed in other studies.^70,71 Following the same logic, we expect that when consumers hold favorable WAs, they are more likely to make a purchase.

H7: PPERC has a positive effect on WAs.

H8: WAs have a positive effect on PIs.

Methodology

Sample and data collection

We selected users of a popular beauty and fashion Web site (Fashion Guide) as the study population (more than 2,360,000 users). Following the C-A-B hierarchy of the ABC model of attitude, this study focused on consumers who were highly involved in this discussion forum Web site. The sample consisted of users who had read online reviews on products or services on the Fashion Guide within the past 6 months and participated in the forum discussions at least twice a week. The Fashion Guide is one of the most visited Web sites in Taiwan in regard to online reviews on beauty and fashion products/services.⁴²

Online surveys were distributed to potential respondents via MY3Q, an online survey distribution platform, between July and August 2010. A total of 462 surveys were returned, and 353 surveys were valid after eliminating the 109 incomplete surveys. Of the respondents, 83.57% were female. In addition, 67.42% of the respondents were between the ages of 18 and 30 years. Of the respondents, 84.13% of the respondents had been using the Internet for more than 4 years. With regard to daily Internet usage, 34.28% of the respondents use the Internet for more than 4 hours, and 36.54% for between 2 and 4 hours. Moreover, 68.5% of the respondents spent more than 6 hours on the Fashion Guide Web site per week.^a

Measures

All of the constructs included in the proposed model were measured using multi-item scales drawn from previous studies that reported high statistical reliability and validity. The items used to measure each of constructs are presented in Appendix 1. The scale for measuring WR was adapted from Bart et al.³⁷ The scale for SC was derived from Sussman and Siegal.⁷² The items used to measure OBRI and SOTI were from Henning-Thurau and Walsh.⁵⁶ PPERC was adapted from Cheung et al.⁴⁴ The scales for PA and PIs were adapted from Jiang and Benbasat,³⁴ while WA was from Shamdasani et al.³⁸ In order to minimize the problem of possible multicollinearity, WR, PA, WA, and PIs were measured on a 7-point Likert scale (1=“strongly disagree,” 7=“strongly agree”), whereas OBRI, SOTI, and PPERC were measured on a 5-point Likert scale (1=“strongly disagree,” 5=“strongly agree”). SC was measured on a 7-point semantic differential scale.

Data Analysis and Results

In order to test the proposed model and research hypotheses, structural equation modeling (SEM) was used. This study employed the two-stage approach suggested by Anderson and Gerbing.⁴⁵ First, the measurement model is estimated with CFA to test reliabilities and validities of the research constructs. Then, the structural model is used to test the strength and direction of the proposed relationships among research constructs.

Measurement model

The measurement model showed adequate fit:⁷³ x²/df=1.557, goodness-of-fit index (GFI)=0.891, comparative fit index (CFI)=0.956, and root mean square error of approximation (RMSEA)=0.040. As shown in Table 1, the composite reliability for each construct was greater than 0.702, demonstrating a reasonable degree of internal consistency between the corresponding indicators.⁷⁴ The squared multiple correlations were all greater than 0.2.⁷⁵ Results showed support for convergent and discriminant validity. As evidence of convergent validity, each item loaded significantly on its respective construct.⁷³ Evidence of discriminant validity exists when the square root of the average of variance extracted (AVE) in each construct exceeds the coefficients representing its correlation with other constructs.⁷⁶ The AVE of each construct exceeded 0.50,⁷⁷ except for OBRI, SOTI, and WA. As presented in Table 2, the results indicate acceptable discriminant validity.

Table 1.

Analysis of Measurement Model

	MLE estimates
Constructs	Factor loading (λ_x/λ_y)	Measurement error (δ/ɛ)	Squared multiple correlation (SMC)	Composite reliability	Average variance extracted (AVE)
Web site reputation				0.897	0.636
WR1	0.725^***	0.475	0.525
WR2	0.719^***	0.484	0.516
WR3	0.870^***	0.243	0.757
WR4	0.861^***	0.259	0.741
WR5	0.799^***	0.361	0.639
Source credibility				0.860	0.607
SC1	0.700^***	0.510	0.490
SC2	0.692^***	0.521	0.479
SC3	0.850^***	0.277	0.723
SC4	0.859^***	0.263	0.737
Obtaining buying-related information				0.706	0.376
OBRI1	0.553^***	0.694	0.306
OBRI2	0.649^***	0.579	0.421
OBRI3	0.593^***	0.648	0.352
OBRI4	0.653^***	0.573	0.427
Social orientation through information				0.702	0.371
SOTI1	0.592^***	0.650	0.350
SOTI2	0.611^***	0.627	0.373
SOTI3	0.597^***	0.644	0.356
SOTI4	0.635^***	0.597	0.403
Perceived positive eWOM review credibility				0.842	0.641
PPERC1	0.837^***	0.300	0.700
PPERC2	0.781^***	0.390	0.610
PPERC3	0.782^***	0.388	0.612
Product attitude				0.801	0.574
PA1	0.747^***	0.442	0.558
PA2	0.844^***	0.287	0.713
PA3	0.673^***	0.548	0.452
Web site attitude				0.856	0.499
WA1	0.684^***	0.533	0.467
WA2	0.737^***	0.457	0.543
WA3	0.753^***	0.432	0.568
WA4	0.704^***	0.505	0.495
WA5	0.636^***	0.595	0.405
WA6	0.719^***	0.483	0.517
Purchase intentions				0.893	0.677
PI1	0.792^***	0.490	0.627
PI2	0.813^***	0.322	0.661
PI3	0.853^***	0.502	0.727
PI4	0.831^***	0.405	0.691
Fit statistics (N=353)
χ²=727.066
df=467
CFI=0.956
GFI=0.891
RMSEA=0.040

***

All factor loadings are significant at the p<0.001 level.

CFI, comparative fit index; GFI, goodness-of-fit index; RMSEA, root mean square error of approximation.

Table 2.

Correlation Matrix for Measurement Scales

Constructs	WR	SC	OBRI	SOTI	PPERC	PA	WA	PI
WR	0.797
SC	0.507	0.779
OBRI	0.496	0.243	0.613
SOTI	0.452	0.343	0.330	0.609
PPERC	0.570	0.509	0.285	0.330	0.800
PA	0.611	0.404	0.482	0.470	0.419	0.758
WA	0.630	0.407	0.546	0.434	0.414	0.541	0.706
PI	0.693	0.473	0.472	0.500	0.453	0.663	0.618	0.823

Diagonal elements in bold face type are the square roots of the average variance extracted.

WR, Web site reputation; SC, source credibility; OBRI, obtaining buying-related information; SOTI, social orientation through information; PPERC, perceived positive eWOM review credibility; PA, product attitude; WA, Web site attitude; PI, purchase intentions.

Model fit

The fit of data to the proposed model was adequate:⁷⁸ χ²=1050.832, df=481, CFI=0.916, GFI=0.843, AGFI=0.816, and RMSEA=0.058. Consistent with H1–H4, the results showed that WR (γ₁₁=0.511, p<0.001), SC (γ₁₂=0.275, p<0.001), OBRI (γ₁₃=0.173, p<0.01), and SOTI (γ₁₄=0.169, p<0.01) had a significant and positive effect on PPERC. PPERC had a significant and positive effect on PAs (β₂₁=0.792, p<0.001) and WAs (β₃₁=0.760, p<0.001), supporting H5 and H7. PAs (β₄₂=0.487, p<0.001) and WAs (β₄₃=0.299, p<0.001) had a significant and positive effect on PIs, supporting H6 and H8. The results are shown in Table 3.

Table 3.

Results of Proposed Model

Hypothesis relationships				Hypothesized model	Alternative model
γ₁₁	Website reputation	→	Perceived positive eWOM review credibility	0.511^***	0.537^***
γ₁₂	Source credibility	→	Perceived Positive eWOM review credibility	0.275^***	0.257^***
γ₁₃	Obtaining buying-related information	→	Perceived positive eWOM review credibility	0.173^**	0.168^**
γ₁₄	Social orientation through information	→	Perceived Positive eWOM review credibility	0.169^**	0.184^**
β₂₁	Perceived Positive eWOM review credibility	→	Product attitude	0.792^***	0.773^***
β₃₁	Perceived Positive eWOM review credibility	→	Web site attitude	0.760^***	0.761^***
β₄₂	Product attitude	→	Purchase intentions	0.487^***	0.292^***
β₄₃	Web site attitude	→	Purchase intentions	0.299^***	0.154^*
β₄₁	Perceived positive eWOM review credibility	→	Purchase intentions	—	0.338^***
Fit index:
χ² (df)				1050.832 (481)	1037.291 (480)
p value				0.000	0.000
CFI				0.916	0.918
GFI				0.843	0.844
AGFI				0.816	0.818
RMSEA				0.058	0.057
1. Hypothesized model: perceived positive eWOM review credibility→R²=0.859; product attitude→R²=0.627; Web site attitude→R²=0.577; purchase intentions→R²=0.502.
2. Alternative model: perceived positive eWOM review credibility→R²=0.892; product attitude→R²=0.598; Web site attitude→R²=0.579; purchase intentions→R²=0.509

***

p<0.001; ^**p<0.01; ^*p<0.05.

We further performed an analysis to test the overidentifying restrictions of the model. This analysis was done by freeing the path between PPERC and PIs that was constrained to zero. Surprisingly, this alternative model, shown in Figure 2, seemed to fit the data better than the proposed model (χ²=1037.291, df=480, CFI=0.918, GFI=0.844, AGFI=0.818, RMSEA=0.057), and PPERC had a positive and significant effect on PIs (β₄₁=0.338, p<0.001). We take into consideration both model parsimony and fit in the selection of the best-fitting model.^b Table 4 provides the fit statistics for the hypothesized and the alternative models. The results of comparing the hypothesized model with the alternative model suggest that PAs and WAs partially mediate the effects of PPERC on PIs.

FIG. 2.

Alternative model.

Table 4.

Fit Statistics for Hypothesized and Alternative Models

Model	x ²	df	Δx ²	Δdf	GFI	TLI	CFI	RMSEA	AIC	CAIC	ECVI	PNFI
Null model (independence model)	7,299.077	528	–	–	0.188	0.000	0.000	0.191	7,365.077	7525.670	20.924	0.000
Saturated model	0.000	0	–	–	1.000	–	1.000	–	1,122.000	3852.089	3.188	0.000
Hypothesized model	1,050.832	481	6,248.245	47	0.843	0.908	0.916	0.058	1,210.832	1600.150	3.440	0.780
Alternative model
Hypothesized model+one direct path between PPERC and purchase intentions	1,037.291	480	13.541	1	0.844	0.909	0.918	0.057	1,199.291	1593.475	3.407	0.780

Note. We compare alternative models with the hypothesized model. df, degree of freedom, GFI, goodness of fit index; TLI, Tucker–Lewis index; CFI, comparative fit index; RMSEA, root mean square error of approximation; AIC, Akaike information criterion; CAIC, consistent ACI; ECVI, expected cross-validation index; PNFI, parsimonious normed fit index. For AIC, CAIC, and ECVI, lower value indicates a better fit and a more parsimonious model; for PNFI, greater value indicates a better fit and a more parsimonious model.

Post analysis: tests of mediation effects

In order to test the mediating effect of the PAs and WAs on the relationship between PPERC and PIs, this research conducted a series of regression analyses⁷⁹ using PPERC as the independent variable (IV), PAs and WAs as the mediators, and PIs as the dependent variable (DV). As Baron and Kenny's procedure only tests the effect of one mediator, we ran tests of the mediation effect of the PAs and WAs separately. First, we regressed the mediator on the IV. Then we regressed the DV on both the IV and the mediator variable. The results in Table 5 showed that PPERC significantly influenced PAs (β=0.523, p<0.001) and WAs (β=0.305, p<0.001). In addition, PPERC influenced PIs (β=0.569, p<0.001). However, this effect was reduced when PAs/WAs were included in the regression equation (PA: β=0.266, p<0.001; WA: β=0.299, p<0.001), while the effect of PAs/WAs on purchase intentions remained significant (PA: β=0.579, p<0.001; WA: β=0.884, p<0.001). The Sobel tests⁷⁹ confirmed that the reduction of the PPERC effect was significant (PA: z=7.32, p<0.001; WA: z=4.66, p<0.001), suggesting that the PAs and WAs partially mediated the impact of PPERC on PIs.^c

Table 5.

Mediation Analysis

Mediator	Equation	Dependent variable	Independent variable	Unstandardized β	t	p
Product attitude	1	Product attitude	PPERC	0.523	8.653	0.001
	2	Purchase intentions	PPERC	0.569	9.519	0.001
	3	Purchase intentions	PPERC	0.266	4.982	0.001
			Product attitude	0.579	13.487	0.001
Web site attitude	1	Web site attitude	PPERC	0.305	8.519	0.001
	2	Purchase intentions	PPERC	0.569	9.519	0.001
	3	Purchase intentions	PPERC	0.299	5.363	0.001
			Web site attitude	0.884	11.694	0.001

Discussion

The purpose of this research was to propose and empirically test an integrated eWOM effect model, developed using the ABC model of attitude as its base. We investigated the antecedents and consequences of PPERC in an online community. Based on the research findings, a revised eWOM effect model was suggested.

The results show that a WR positively influences PPERC. When the WR is high, the PPERC increases. SC positively influenced the PPERC, which is consistent with the finding of Cheung et al.⁴⁴ Our findings suggest that SC plays an important role in consumer trust in eWOM reviews. The Fashion Guide Web site provides information on the number of recommendations for posted reviews. This Web site feature might provide a good way to build PPERC. Our results also indicated that OBRI positively influences PPERC. When consumers read eWOM reviews in order to OBRI, they want to reduce their purchase risks and time spent searching.⁵⁶ As such, they are more likely to give higher credibility to eWOM reviews.⁴⁸ The results showed that SOTI positively influenced the PPERC. Online community members provided product/service reviews to others in order to make contributions to the community. By doing so, they not only help others make better decisions, but also obtain social identification. As such, they show their involvement in the community.⁸⁰ eWOM Web sites consist of individuals with similar interests who share and exchange experiences. As such, these individuals are likely to trust each other.²

PAs and WAs partially mediate the effects of the PPERC on PIs. When consumers perceive the eWOM reviews as being credible, they form favorable PAs and WAs that, subsequently, impact PIs. PPERC also directly positively influences PIs. Although the direct effect of PPERC on PIs was not expected in our proposed model, previous research has shown that when consumers trust eWOM reviews, they are more likely to adopt the reviews and make a purchase decision.^44,81 PIs are determined by consumers' perceived value of a product.⁸² When consumers perceive positive eWOM reviews as being more credible, they are more likely to rely on the eWOM reviews in order to make purchase decisions.⁸¹ The findings suggest that PPERC is an important predictor of PAs and WAs as well as PIs.

This research makes theoretical contributions to the WOM literature in several aspects. First, unlike previous research that has focused on external information sources, such as source expertise⁷ and review ratings,⁷ this research considered the impact of individual and social factors on eWOM effects. We identified the impact of platform- (WR and SC) and customer-based (OBRI and SOTI) motivations for PPERC. Second, based on the ABC model of attitude, this research demonstrated that platform- and customer-based motivations influence the PPERC and then subsequently PIs. Different from previous research that regarded the affect component as a single dimension,²⁰ this research showed how PAs and WAs partially mediated the effects of PPERC on PIs. In addition, previous studies did not clearly differentiate between the elements of the eWOM effect, such as PPERC, PAs, and WAs.⁵⁰ The underlying process of the eWOM effect is revealed.

Our findings provide managerial implications in online community management and marketing. Our findings suggest that consumer-opinion platform providers need to develop good WRs and a speaker rating system in order to show its objectiveness in regard to the reviews on the platform. The Fashion Guide Web site uses a rating system to evaluate the speakers' influence. Such a system shows the SC of the eWOM reviews and helps consumers make purchase decisions. Our research also indicated that social interactions with others about products/service consumption experiences may influence consumers' PIs via PPERC as well as PAs and WAs.

This research demonstrates that the success of eWOM should be determined by satisfying the consumers' functional and social needs. Online community members usually seek useful product information (functional need) and build social relationships with others (social need). In order to meet these needs, consumer-opinion platform providers can offer customized features so that users can build their own discussion forums, which reduce their information search efforts. They may also host social activities that facilitate interactions between members. As such, eWOM review listeners are more likely to trust the eWOM reviews on the Web site.

This research also suggests that PAs and WAs play an important role in consumer PIs and that PAs have a higher impact on PIs than WAs. When eWOM speakers have positive reviews toward a product, consumers are more likely to form positive PAs. Internet marketers should seek reviews from credible and influential eWOM speakers on a particular Web site by asking them to try products. Positively reviewed products should generate favorable attitudes and high PIs. On the other hand, platform service providers need to ensure that users are satisfied with the Web site services and that the Web site is easy to navigate so that they will desire to return to search for additional information.

Despite the progress made by this research, it has some limitations. First, we only considered a select number of platform- and customer-based factors as antecedents of PPERC. Many possible factors, such as browsing time and response frequency, might also influence PPERC.⁵⁹ In addition, we did not control for the listeners' involvement⁸³ and expertise⁸⁴ or the speakers' expertise⁸⁵ and review quantity and quality.⁵⁵ Future research may want to control for or manipulate some of these variables in order to see how they influence the observed effects in our research. Third, following previous research,⁴⁴ we focused on the effects of positive eWOM reviews. We did not consider consumer revenge⁸⁶ and dysfunctional behavior.^87,88 Therefore, future research is encouraged to investigate effects of negative eWOM reviews.

Footnotes

Notes

Author Disclosure Statement

No competing financial interests exist.

Appendix

References

Sun

, Youn

, Wu

, Kuntaraporn

. Online word-of-mouth (or mouse): an exploration of its antecedents and consequences. Journal of Computer-Mediated Communication, 2006; 11:1–22.

Lee

, Youn

. How eWOM platforms influence consumer product judgment. International Journal of Advertising, 2009; 28:473–99.

Hennig-Thurau

, Gwinner

, Walsh

, Gremler

. Electronic word-of-mouth via consumer-opinion platforms: what motivates consumers to articulate themselves on the Internet. Journal of Interactive Marketing, 2004; 18:38–52.

Forrester Research. 2011. The five methods of online marketing through social networks. http://advertising.microsoft.com/taiwan/NewsAndEvents/PressRelease.aspx?Adv_PressReleaseID=826. 2011 May 9.

Dichter

. How word-of-mouth advertising works. Harvard Business Review, 1966; 44:147–66.

Engel

, Blackwell

, Miniard

. 1993. Consumer behavior. Fort Worth, TX: Dryden Press.

Martin

, Lueg

. Modeling word-of-mouth usage. Journal of Business Research, 2011. https://dx-doi-org.web.bisu.edu.cn/10.1016/j.jbusres.2011.06.004

Huang

, Chou

, Lin

. The influence of reading motives on the responses after reading blogs. CyberPsychology & Behavior, 2008; 1:351–55.

Sweeney

, Soutar

, Mazzarol

. Factors influencing word of mouth effectiveness: receiver perspectives. European Journal of Marketing, 2008; 42:344–64.

10.

Riegner

. Word of mouth on the Web: the impact of Web 2.0 on consumer purchase decisions. Journal of Advertising Research, 2007; 47:436–47.

11.

Hoffman

, Novak

. Marketing in hypermedia computer-mediated environments: conceptual foundations. Journal of Marketing, 1996; 60:50–68.

12.

Gelb

, Sundaram

. Adapting to word of mouse. Business Horizon, 2002; 45:21–5.

13.

Sundaram

, Mitra

, Webster

. Word-of-mouth communications: a motivational analysis. Advances in Consumer Research, 1998; 25:527–31.

14.

Breckler

. Empirical validation of affect, behavior, and cognition as distinct components of attitude. Journal of Personality & Social Psychology, 1984; 47:1191–205.

15.

Lutz

. 1985. Affective and cognitive antecedents of attitude toward the ad: a conceptual framework. Alwitt

, Mitchell

. Psychological processes and advertising effects: theory, research and applications. Hillsdale, NJ: Lawrence Erlbaum, 45–63.

16.

Ifeng Forum. 2011. Maintenance for online reputation: businesses and individuals cannot ignore. http://big5.ifeng.com/gate/big5/bbs.ifeng.com/viewthread.php?tid=6417409&extra=page-ifbase4-base8-JTNEMQ. 2011 May 22.

17.

Dhar

, Nowlis

. The effect of time pressure on consumer choice deferral. Journal of Consumer Research, 1999; 25:369–84.

18.

Spears

. Time pressure and information in sales promotion strategy: conceptual framework and content analysis. Journal of Advertising, 2001; 30:67–76.

19.

East

, Hammond

, Lomax

. Measuring the impact of positive and negative word of mouth on brand purchase probability. International Journal of Research in Marketing, 2008; 25:215–24.

20.

Park

, Lee

. Information direction, Website reputation and eWOM effect: a moderating role of product type. Journal of Business Research, 2009; 62:61–7.

21.

Helm

. Viral marketing—establishing customer relationships by “word-of-mouse.” Electronic Markets, 2000; 10:158–61.

22.

Chevalier

, Mayzlin

. 2003. The effect of word of mouth on sales: online book reviews. NBER Working Papers 10148. National Bureau of Economic Research, Inc.

23.

Gershoff

, Mukherjee

, Mukhopadhyay

. Consumer acceptance of online agent advice: extremity and positivity effects. Journal of Consumer Psychology, 2003; 13:161–70.

24.

Skowronski

, Carlston

. Social judgment and social memory: the role of cue diagnosticity in negativity, positivity, and extremity biases. Journal of Personality & Social Psychology, 1987; 105:689–99.

25.

Skowronski

, Carlston

. Negativity and extremity biases in impression formation: a review of explanations. Psychological Bulletin, 1989; 105:131–42.

26.

Zhang

, Craciun

, Shin

. When does electronic word-of-mouth matter? A study of consumer product reviews. Journal of Business Research, 2010; 63:1336–41.

27.

Sen

, Lerman

. Why are you telling me this? An examination into negative consumer reviews on the Web. Journal of Interaction Marketing, 2007; 21:76–94.

28.

Rossiter

, Percy

. 1997. Advertising communications and promotion management. New York: McGraw-Hill.

29.

De Angelis

, Bonezzi

, Peluso

et al. On braggarts and gossips: a self-enhancement account of word-of-mouth generation and transmission. Journal of Marketing Research, 2012; 49:551–63.

30.

Schindler

, Bickart

. Perceived helpfulness of online consumer reviews: the role of message content and style. Journal of Consumer Behaviour, 2012; 11:234–43.

31.

Jeong

, Jang

. Restaurant experiences triggering positive electronic word-of-mouth (eWOM) motivations. International Journal of Hospitality Management, 2011; 30:356–66.

32.

Arndt

. Role of product-related conversations in the diffusion of a new product. Journal of Marketing Research, 1967; 4:291–95.

33.

Sundaram

, Mitra

, Webster

. Word-of-mouth communications: a motivational analysis. Advances in Consumer Research, 1998; 25:527–31.

34.

Flanagin

, Metzger

. 2008. Digital media and youth: unparalleled opportunity and unprecedented responsibility. Metzger

, Flanagin

, John

, Catherine

. Digital media, youth, and credibility. Cambridge, MA: MIT Press, 5–28.

35.

Mackenzie

, Lutz

. An empirical examination of the structural antecedents of attitude toward the ad in an advertising pretesting context. Journal of Marketing, 1989; 53:48–65.

36.

Metzger

. Making sense of credibility on the Web: models for evaluating online information and recommendations for future research. Journal of the American Society for Information Science and Technology, 2007; 58:2078–91.

37.

Bart

, Shankar

, Sultan

et al. Are the drivers and role of online trust the same for all Web sites and consumers? A large-scale exploratory empirical study. Journal of Marketing, 2005; 69:133–52.

38.

Shamdasani

, Stanaland

AJS

, Tan

. Location, location, location: insights for advertising placement on the Web. Journal of Advertising Research, 2001; 41:7–21.

39.

Wathen

, Burkell

. Believe it or not: Factors influencing credibility on the Web. Journal of the American Society for Information Science & Technology, 2002; 53:34–144.

40.

Eagly

, Chaiken

. 1993. The psychology of attitudes. Fort Worth, TX: Harcourt Brace Jovanovich College Publishers.

41.

Hovland

, Weiss

. The influence of source credibility on communication effectiveness. Public Opinion Quarterly, 1951; 15:635–50.

42.

Grewal

, Gotlieb

, Marmorstein

. The moderating effects of message framing and source credibility on the price-perceived risk relationship. Journal of Consumer Research, 1994; 21:145–53.

43.

Lim

, Sia

, Lee

MKO

et al. Do I trust you online, and if so, will I buy? An empirical study of two trust building strategies. Journal of Management Information System, 2006; 23:233–66.

44.

Cheung

, Luo

, Sia

et al. Credibility of electronic word-of-mouth: informational and normative determinants of on-line consumer recommendations. International Journal of Electronic Commerce, 2009; 13:9–38.

45.

Schiffman

, Kanuk

. 2000. Consumer behavior, 7th. Upper Saddle River, NJ: Prentice Hall.

46.

Lee

, Fiore

, Kim

. The role of the technology acceptance model in explaining effects of image interactivity technology on consumer responses. International Journal of Retail & Distribution Management, 2006; 34:621–44.

47.

Wiedmann

, Walsh

, Mitchell

. The mannmaven: an agent for diffusing market information. Journal of Marketing Communications, 2001; 7:195–212.

48.

Chen

. Accreditation of website credibility: concept, discourse, and operational framework. Mass Communication Research, 2011; 106:249–83.

49.

Bansal

, Voyer

. World-of-mouth processes within a services purchase decision context. Journal of Service Research, 2000; 3:166–77.

50.

Doh

, Hwang

. How consumers evaluate eWOM (electronic word-of-mouth) messages. CyberPsychology & Behavior, 2009; 12:193–7.

51.

Price

, Feick

. 1984. The role of interpersonal sources in external search: an informational perspective. Kinnear

. Advances in consumer research, 11. Provo, UT: Association for Consumer Research, 250–3.

52.

Armstrong

, McAdams

. Blogs of information: how gender cues and individual motivations influence perceptions of credibility. Journal of Computer-Mediated Communication, 2009; 14:435–56.

53.

Chu

, Kim

. Determinants of consumer engagement in electronic word-of-mouth in social networking sites. International Journal of Advertising, 2011; 30:47–75.

54.

Sweeney

, Soutar

, Mazzarol

. Factors influencing word of mouth effectiveness: receiver perspectives. European Journal of Marketing, 2008; 42:344–64.

55.

Fan

, Miao

, Farn

. 2012. Effect of electronic word-of-mouth on consumer purchase intention: the perspective of gender differences. Proceedings of the International Conference on Innovation and Management. Koror: Republic of Palau.

56.

Henning-Thurau

, Walsh

. Electronic word of mouth: motives for and consequences of reading customer articulations on the Internet. International Journal of Electronic Commerce, 2003; 8:51–74.

57.

Richins

. Negative word-of-mouth by dissatisfied customers: a pilot study. Journal of consumer Research, 1983; 10:73–82.

58.

Jiang

, Benbasat

. Investigating the influence of the functional mechanisms of online product presentations. Information Systems Research, 2007; 18:454–70.

59.

Tsai

, Lu

, Huang

. 2008. The influences of product review credibility and product rating credibility on browsers' product attitudes: an online platform of cosmetics. Proceedings of the 11th Conference on Electronic Commerce, Taipei, Taiwan.

60.

Wilkie

. 1994. Consumer behavior, 3rd. New York: John Wiley.

61.

Bickart

, Schindler

. Internet forums as influential sources of consumer information. Journal of Interactive Marketing, 2001; 15:31–40.

62.

Lin

, Tan

. A study of persuasive effects of electronic words-of-mouth. Marketing Review, 2010; 7:87–208.

63.

Fishbein

, Ajzen

. 1975. Belief, attitude, intention, and behavior: an introduction to theory and research. Reading, MA: Addison-Wesley.

64.

Liebermann

, Flint-Goor

. Message strategy by product-class type: a matching model. International Journal of Marketing Research, 1996; 13:237–49.

65.

Lee

, Park

, Han

. The effect of negative online consumer reviews on product attitude: an information processing view. Electronic Commerce Research & Applications, 2008; 7:341–52.

66.

Prendergast

, Ko

, Yuen

SYV

. Online word of mouth and consumer purchase intentions. International Journal of Advertising, 2010; 29:687–708.

67.

Park

, Wang

, Yao

. 2011. A structural model of eWOM effects in China. International Conference on Social Science and Humanity, Singapore.

68.

Chiou

, Cheng

. Should a company have message boards on its Web sites. Journal of Interactive Marketing, 2003; 17:50–61.

69.

Park

, Kim

. Identifying key factors affecting consumer purchase behavior in an online shopping context. International Journal of Retail & Distribution Management, 2003; 31:16–29.

70.

Karson

, Fisher

. Reexamining and extending the dual mediation hypothesis in an on-line advertising context. Psychology & Marketing, 2005; 22:333–51.

71.

Jee

, Lee

. Antecedents and consequences of perceived interactivity: an exploratory study. Journal of Interactive Advertising, 2002; 3:34–45.

72.

Sussman

, Siegal

. Informational influence in organizations: an integrated approach to knowledge adoption. Information Systems Research, 2003; 14:47–65.

73.

Anderson

, Gerbing

. Structural equation modeling in practice: a review and recommended two-step approach. Psychological Bulletin, 1988; 103:411–25.

74.

Hair

Jr , Black

, Babin

et al. 2010. Multivariate data analysis. 7th. Englewood Cliffs, NJ: Prentice-Hall.

75.

Bentler

, Wu

EJC

. 1993. EQS/Windows user guide, BMDP statistical software. Los Angeles, CA: BMDP Statistical Software.

76.

Fornell

, Larcker

. Evaluating structural equation models with unobservable variables and measurement error. Journal of Marketing Research, 1981; 18:39–50.

77.

Gaski

, Nevin

. The differential effects of exercised and unexercised power sources in a marketing channel. Journal of Marketing Research, 1985; 22:130–42.

78.

Baumgartner

, Homburg

. Applications of structural equation modeling in marketing and consumer research: a review. International Journal of Research in Marketing, 1996; 13:139–61.

79.

Baron

, Kenny

. The moderator–mediator variable distinction in social psychological research: conceptual, strategic and statistical considerations. Journal of Personality & Social Psychology, 1986; 51:1173–82.

80.

McWilliam

. Building stronger brand through online communities. Sloan Management Review, 2000; 41:43–55.

81.

Nabi

, Hendriks

. The persuasive effect of host and audience reaction shots in television talk shows. Journal of Communication, 2003; 53:527–43.

82.

Dodds

, Monroe

, Grewal

. Effect of price, brands and store information on buyers' product evaluation. Journal of Marketing Research, 1991; 28:307–19.

83.

Park

, Lee

, Han

. The effect of on-line consumer review on consumer purchasing intention: the moderating role of involvement. International Journal of Electronic Commerce, 2007; 11:125–48.

84.

Jamal

, Anastasiadou

. Investigating the effects of service quality dimensions and expertise on loyalty. European Journal of Marketing, 2009; 43:398–420.

85.

D'Astous

, Touil

. Consumer evaluations of movies on the basis of critics' judgments. Psychology & Marketing, 1999; 16:677–94.

86.

Zourrig

, Chebat

, Toffoli

. Consumer revenge behaviour: a cross-cultural perspective. Journal of Business Research, 2009; 62:995–1001.

87.

Harris

, Reynolds

. The consequences of dysfunctional customer behavior. Journal of Service Research, 2003; 6:144–61.

88.

Reynolds

, Harris

. Dysfunctional customer behavior severity: an empirical examination. Journal of Retailing, 2009; 85:321–35.

89.

Zhao

, Lynch

, Chen

. Reconsidering Baron and Kenny: myths and truths about mediation analysis. Journal of Consumer Research, 2010; 37:197–206.

90.

Preacher

, Hayes

. 2008. Contemporary approaches to assessing mediation in communication research. Hayes

, Slater

, Snyder

. The Sage sourcebook of advanced data analysis methods for communication research. Thousand Oaks, CA: Sage, 13–54.