Menu
Home
Log in / Register
 
Home arrow Engineering arrow Measuring Electronic Word-of-Mouth Effectiveness: Developing and Applying the eWOM Trust Scale
Source

Nomological Validity

Responses from the main study were used to partly assess the measure’s nomological validity. Table 42 reports the Pearson correlations between eWOM trust and measures dealing with variables representing eWOM trust antecedents, correlates, and consequences. Positive correlations were predicted between customer trust in online reviews and reviewer credibility (H4), as well as distrust in reviewers' motives (H12) as antecedents, between eWOM Trust and eWOM involvement (H13) as a correlate, and between eWOM trust and eWOM usage, (H14) as well as eWOM avoidance (H15) as consequences. By using data from study 4, all of these correlations were proven to be significant and in the hypothesized directions. Specifically, the correlation coefficient between scores on the eWT-S and the scale assessing perceived reviewer credibility in general was .70 (p < .001) and positive. In line with these research predictions, the correlation with perceived reviewer distrust was negative, substantial (r = -.33) and significant (p < .001). According to the data, respondents who have heightened trust in online customer reviews perceive this kind of market communication as more relevant. This fact was manifested by a strong positive correlation between the new measure of the focal construct and eWOM Involvement (r = .69, p < .001). The new scale also showed its ability to predict hypothesized outcomes such as general eWOM usage (r = .76, p < .001) as well as general eWOM avoidance (r = -.21, p < .001). Thus, all these hypotheses were supported. The relationship between the individual constructs also behaved in a manner reasonable to the rationale - supporting, as a consequence, the validity of the scale’s validity.

Mean

SD

eWOM

Trust

eWOM

Usage

eWOM

Involvement

eWOM

Avoidance

Reviewer

Distrust

Reviewer

Credibility

eWOM Trust

3.73

.89

(.95)1

eWOM Usage

3.39

1.19

76***

(.87)1

eWOM

Involvement

3.86

.94

.69***

.64***

(.91)1

eWOM Avoidance

1.56

1.23

_21***

_.87*

_.42***

(.86)1

Reviewer Distrust

2.43

1.18

_ 33***

_19***

_ 41***

53***

(.77)1

Reviewer

Credibility

3.63

.80

.70***

.54***

.70***

_ 27***

_ 37***

(.90)1

Notes: n=526; 1= Internal consistency estimates (Cronbach’s alpha); Pearson correlation, 2-tailed; »»» = p < .001; * = p < .05.

Further evidence for the nomological validity of the scale was provided by a series of street- intercept interviews among Austrian Internet users. These samples were later later pooled to represent sample 5. The separate sub-samples were used to evaluate the relationships between the scale of the focal construct and measures of different constructs. In general, most of the proposed relationships were supported by sample data. Specifically, with data from a subsample (n = 268) the correlation between eWOM trust and two hypothesized antecedents, namely perceived tie strength (measured with 10 items (De Bruyn & Lilien, 2008; Frenzen & Davis, 1990; Money et al., 1998; 7-Point Likert scale) and perceived (behavioral, attitudinal) homophily with online reviewers (measured with 22 items adopted from McCroskey et al. (2006) and De Bryn and Lilien (2008); 7-Point Likert scale). Results showed that in line with the proposed hypotheses (H7 and H8), the relationship among the measures of the constructs was significant. That is, the correlations coefficient between eWOM trust and perceived tie strength was .77 (p < .001), and .68 (p < .001) between eWOM trust and perceived reviewer homophily.

A positive correlation was predicted between perceived reviewer altruism (H5) (construct measured with 11 items adopted from Cheung and Lee (2012), Fang and Chiu (2010), Podsakoff et al. (1990), and Price et al. (1995); 7-Point Likert scale) and a negative correlation between perceived reviewer egoism (H6) (construct measured with 3 items adopted from Power et al. (2008); 7-Point Likert scale) and eWOM trust in general. While a significant relationship could be identified in the first case (r = .76, p < .001) by using data from another sub-sample (n = 274), H6 did not find empirical evidence (r = -.12, p > .05). However, the results found a negative but insignificant relationship.

According to literature, a negative relationship between the eWOM trust scale and a scale assessing the different dimensions of consumer self-confidence (measured with 18 items taken mainly from Bearden et al. (2001) and Rosenberg (1965); 7-Point Likert scale) was expected

271

(H2). By using data from an alternative sample (n = 80), however, a significant but positive interrelation was discovered for information acquisition or the consumer’s confidence in his/her ability to obtain needed marketplace information) (r = .29, p < .05), persuasion knowledge (i.e., the individual’s confidence in his/her knowledge regarding the tactics used by marketers in order to persuade consumers) (r = .24, p < .05), and consideration set formation, which is the defined as the consumer’s confidence in one’s ability to identify acceptable choice alternatives (r = .28, p < .05) (Bearden et al., 2001). Bearden et al. conceptualize self-confidence as a construct consisting of self-efficacy and self-esteem. Self-confidence and esteem are hence highly intertwined. Radecki and Jaccard (1995) argue that persons with a low level of selfesteem typically hesitate to reveal their knowledge gap, which ultimately decreases information search. Similarly, Bishop and Barber (2012) hypothesize that (i) individuals with low selfesteem tend to depend more on internal sources of information (i.e., themselves), and further that (ii) individuals with low self-esteem are more likely to consult external resources, including other persons (e.g., kin) or impersonal sources of information (e.g., Internet, product discussions, advertisement). Both hypotheses were supported by empirical data. The same relationship seems to surface in this study.

By using the same sample, nevertheless, H3 was supported by a significant positive relationship (r = .51, p < .001) between eWOM trust and consumer risk propensity (measured with 2 items adopted from Chang et al. (2008); 7-Point Likert scale). This implied that the more a consumer is a risk taker, the more he/she is willing to trust information given in online customer reviews.

This research also predicted that consumer susceptibility to interpersonal information (H9), attention to social comparison information (H10), eWOM influence (H14a), eWOM seeking (H14b), and eWOM passing (H14c) would be positively correlated with eWOM trust. The following scales were used: 16 items taken from Bearden et al. (1989) to measure CSII; 7 items from Bearden and Rose (1990) to measure ATSCI; and a total of 18 items adopted from Park and Lee (2009), Chu and Kim (2011), Bailey (2005), and Lee and Koo (2012) to measure the different kinds of eWOM usage. All constructs were measured on a 7-Point Likert scale ranging from 0 (“I strongly disagree”) to 6 (“I strongly agree”). According to the data obtained from the sample (n = 120), all the correlations pertaining to these predictions were significant. The weakest correlation was observed between eWOM trust and ATSCI (r = .19); however, it was still significant on the p < .05 level.

H11 predicted that eWOM trust and eWOM skepticism (measured in the questionnaire with 9 items adopted from Obermiller and Spangenberg (1998); 7-Point Likert scale) were negatively related. Accordingly, the sub-sample (n = 135) provided adequate support for this relationship (r = -.93, p < .001). In total, as the majority of the hypothesized relationships behaved as predicted, strong empirical evidence for the nomological validity of the scale across separate samples is suggested. Table 43 provides an overview of the hypotheses tests.

Mean

SD

Internal

Consistency1

Correlation with eWOM

Trust

Sample

eWOM Trust

2.78

1.21

.97

n.a.

(n=268)

Perceived Tie Strength

1.68

1.37

.95

77***

Reviewer Homophily

2.26

1.31

.98

.68***

eWOM Trust

3.15

1.12

.97

n.a.

(n=274)

Reviewer Altruism

3.60

1.19

.95

76***

Reviewer Egotism

2.10

1.12

.72

-.12

eWOM Trust

3.34

1.16

.97

n.a.

(n=80)

Consumer Self-confidence

2.49

.95

.83

.29*

Risk Propensity

2.03

1.44

.81

51***

eWOM Trust

2.82

.98

.94

n.a.

(n=120)

CSH (Informative)

3.17

1.16

.86

.34***

ATSCI

2.49

1.39

.88

.19*

eWOM Influence

2.30

1.42

.95

74***

eWOM Seeking

2.42

1.54

.95

75***

eWOM Passing

2.35

1.72

n.a.

59***

eWOM Trust

3.33

1.19

.97

n.a.

(n=135)

eWOM Skepticism

3.37

1.20

.91

-93***

Notes: != Internal consistency estimates (Cronbach’s alpha); Pearson correlation, 2-tailed; »»» = p < .001; » = p < .05.

 
Source
Found a mistake? Please highlight the word and press Shift + Enter  
< Prev   CONTENTS   Next >
 
Subjects
Accounting
Business & Finance
Communication
Computer Science
Economics
Education
Engineering
Environment
Geography
Health
History
Language & Literature
Law
Management
Marketing
Mathematics
Political science
Philosophy
Psychology
Religion
Sociology
Travel