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

Profiling the Segments

Subsequently, the identified segments were first profiled in terms of group demographics and followed by perceptual, attitudinal and habitual characteristics. Analysis of variance (ANOVA) showed a significant difference in age between the groups (F(3,513) = 5.88, p < .001). Post-hoc analyses using Tukey’s HSD indicated that consumers belonging to the fourth segment (i.e., high trusters) were the youngest (M = 39.81, SD = 13.39) when compared to the three remaining groupings. By using a chi-squared difference test, it was found that clusters did not significantly vary either in age (x2(3) = 4.41, p > .05) or in education (x2(18) = 27.13, p > .05). The same was true for maritial status (%2(6) = 6.33, p > .05), gender (x2(3) = 4.41, p > .05) (however, pure review trusters seem to slightly tend to be female), as well as ethnicity of cluster members (X2(12) = 16.00, p > .05). By using a one-way ANOVA, the clusters were found to not differ in net income (F(3,490) = .64, p > .05). However, by applying a similar procedure, this research found a significant difference in current employment among the segments (x2(21) = 40.65, p < .01). That is, white-collar workers and public servants were typically well represented in the high-truster cluster, while retired persons were very unlikely to be found in this segment. On the other hand, unemployed persons are often pure review trusters. Another chi-square test showed that the clusters significantly varied in the living area (x2( 6) = 21.94, p < .001). High trusters were well represented in the urbanized area, while the members of the remaining segments more typically lived in the suburban area.

While some of these findings provide first insights into the characteristics of online consumers, the results also suggested that demographic variables only could not meaningfully differentiate the consumer segments. Prior segmentation studies agree in this aspect and advocate the inclusion of further contextual variables (Schoefer & Diamantopoulos, 2009; Walsh et al., 2010).

Table 51: Cluster Demographics

Cluster 1

Cluster

2

Cluster

3

Cluster

4

Pure review trusters

Nontrusters

Moderately

trusting

consumers

High trusters

CLUSTER SIZE

Absolute

99

54

210

154

Percentage

19%

10%

41%

30%

DEMOGRAPHICS

Age

Mean

44.76

47.52

45.46

39.81

SD

15.81

16.50

14.84

13.39

Gender

male

37.4%

50.0%

49.0%

48.7%

female

62.6%

50.0%

51.0%

51.3%

Employment

Student

5.2%

5.6%

6.2%

3.9%

Employed for wages (white-collar worker, public servant)

35.1%

27.8%

31.6%

47.4%

Employed for wages (blue-collar worker)

8.2%

13.0%

15.8%

14.9%

Self-employed

9.3%

13.0%

11.0%

7.1%

Retired

17.5%

22.2%

20.6%

9.7%

Unemployed (looking for work)

2.1%

7.4%

5.7%

5.8%

Unemployed (not looking for work, unable to work)

19.6%

5.6%

5.3%

7.8%

Other

3.1%

5.6%

3.8%

7.8%

Monthly Net Income

Under USD 20,000

15.8%

14.0%

10.6%

8.0%

USD 20,000-29,999

14.7%

10.0%

11.1%

12.7%

USD 30,000-39,999

10.5%

12.0%

12.1%

12.0%

USD 40,000-49,999

10.5%

6.0%

14.6%

11.3%

USD 50,000-74,999

18.9%

34.0%

26.1%

27.3%

USD 75,000-99,999

16.8%

10.0%

17.6%

15.3%

Above USD 100,000

12.6%

14.0%

8.0%

13.3%

Cluster

1

Cluster

2

Cluster

3

Cluster

4

Pure

review

trusters

Nontrusters

Moderately

trusting

consumers

High trusters

Marital Status

Single

19.4%

24.5%

26.7%

25.8%

Married, living with another person

68.4%

58.5%

57.8%

64.9%

Divorced, separated, widowed

12.2%

17.0%

15.5%

9.3%

Ethnicity

African-American (Non-

6.1%

11.1%

4.8%

13.2%

Hispanic)

Asian or Pacific Islander

3.0%

1.9%

4.8%

7.9%

White/Caucasian (Non-Hispanic)

85.9%

81.5%

85.1%

73.7%

Latino or Hispanic

3.0%

3.7%

3.8%

4.6%

Native American or Aleut

2.0%

1.9%

1.4%

0.7%

Living Area

Urban

19.2%

32.1%

21.0%

40.5%

Sub-urban

54.5%

49.1%

52.4%

39.9%

Rural

26.3%

18.9%

26.7%

19.6%

Accordingly, the four clusters were compared in terms of their perceptions, attitudes and habits concerning online customer reviews and online advertising in general. As discussed earlier in this thesis, consumers with heightened trust in eWOM were theoretically expected to exhibit positive reactions towards online customer reviews than non-trusters. Therefore, this research expected to identify higher positive perceptions and attitudes as well as cooperative behaviors among pure review trusters and high trusters compared to the other two segments. However, the circumstance that trust in B2C communication was found to differ in the high eWOM trust segments makes certain differences likely.

Using one-way ANOVA, significant divergence in the consumers’ attitude towards online customer reviews (a = .92) was found across the four clusters (F(3,513) = 97.75, p < .001). In particular - and in line with expectations - the two clusters having high scores of eWOM trust (i.e., pure review and high trusters) were shown to score significantly higher on the trait than the other two clusters. Additionally, pure review trusters and high trusters exhibited similar attitudes towards eWOM in general (see Table 52). Another insight of the nomological network is that consumer’s trust in online customer reviews is formed by experiences over time, implying that the perceived valence of these interactions has a rudimentary impact on later developed generalized trust. This research therefore expected that individuals who have earlier benefited from online customer reviews are therefore likely to perceive heightened trust (due to operational conditioning) (Hogg & Vaughan, 2011). To investigate this assumption, another one-way ANOVA was conducted showing significant differences between the clusters

(F(3,513) = 76.76, p < .001), in particular demonstrating that pure review trusters as well as high trusters both seemed to have more positive experiences (a = .71) with eWOM than consumers in the other two segments in the past.

It was also discussed earlier that positive relationships between eWOM trust and review usage habits are likely to exist. The research at hand was able to demonstrate such linkages by using a series of one-way ANOVAs. That is, variations in review seeking behavior (a = .90) between the segments were identified (F(3,513) = 84.47, p < .001). There was also a significant impact of segment membership and the impact of online reviews on purchase decisions (a = .91) (F(3,513) = 117.67, p < .001). In both cases, pure review trusters, as well as high trusters, showed a significant heightened positive behavior when compared to the other segments, but did not vary essentially from each other in these respects. Another assumption was that high eWOM trust will guide consumers to forward review content to others (i.e., review passing). This research was able to demonstrate that this expectation is also mirrored by empirical data (F(3,513) = 66.55, p < .001). However, the consumers who are most likely to forward comments and recommendations from fellow shoppers are not pure review trusters but high trusters (M = 4.88, SD = 1.17). Pure review trusters, in contrast, possessed the second highest likeability but they did not significantly differentiate from moderate-trusting consumers. Conducting another ANOVA analysis provided the insight that differences did exist between the identified segments in terms of review posting (F(3,513) = 25.46, p < .001). Here, high trusters were (again) the most active consumers.

A further investigation of the segments’ characteristics revealed differences concerning the perceived levels of risk in using online customer reviews (a = .87) (F(3,513) = 11.51, p < .001). Here, as theorized, post-hoc analyses indicated that pure review trusters exhibited significantly lower levels of risk when compared to the other segments. High trusters showed the highest risk perceptions. So this group of consumers seems to be quite aware of eWOM’s perils. Using another one-way ANOVA also showed significant differences between the segments in terms of review avoidance behavior (a = .87) (F(3,513) = 13.23, p < .001). In line with expectations, the segment of pure review trusters showed the smallest fear of contact (M = 1.01, SD = 1.19), which was significantly smaller than in the other three clusters. High trusters showed a significantly higher review avoidance behavior when compared to this group but also when compared to moderately trusting persons.

Referring to the usage of online shops providing customer reviews, the analysis of variance showed an effect of segment membership on site visits, F(3,513) = 10.46, p < .001. Post-hoc analyses indicated that shop usage was higher within the segments with moderate and high eWOM trust. Here, high trusters exhibited the most positive behavior, being more favourable than in most other segments. However, there was no significant difference in respect to the pure review trusters' cluster. Low trusters are more likely to refrain from visiting commercial websites.

Table 52: Cluster Description

Cluster

1

Cluster

2

Cluster

3

Cluster

4

Pure

review

trusters

Nontrusters

Moderately

trusting

consumers

High trusters

I. SEGMENTATION VARIABLES

Trust in Online Reviews (eWOM Trust)

Mean

4.57

2.69

3.96

5.13

SD

.58

.82

.63

.50

Trust in Online Advertising (Ad Trust)

Mean

1.49

1.17

3.28

4.79

SD .72

.88

.50

.68

II. ONLINE REVIEWS ATTITUDES AND BEHAVIORS Attitude towards Online Reviews

Mean

5.14

3.18

4.41

5.24

SD

.77

1.20

.87

.67

Positive Review Experience

Mean

4.76

2.86

4.12

5.03

SD

.87

1.21

.98

.93

Perceived Review Usage Risk

Mean

1.58

2.52

2.18

2.66

SD

1.29

1.05

1.21

1.96

Review Avoidance

Mean

1.01

1.96

1.57

2.20

SD

1.18

1.15

1.26

2.08

Review Seeking

Mean

5.12

3.15

4.44

5.33

SD

.85

1.30

1.06

0.61

Review Purchase Influence

Mean

4.90

2.66

4.15

5.23

SD

.88

1.25

1.01

.70

Review Forwarding

Mean

3.37

1.83

3.34

4.88

SD

1.79

1.55

1.52

1.17

Review Posting

Mean

2.32

2.06

2.47

3.19

SD

.91

1.00

.99

1.17

Visiting Online Shops with Online Customer Reviews

Mean

3.87

3.43

3.75

4.11

SD

.77

1.00

.87

.78

(Continued on next page.)

III. ONLINE ADVERTISEMENT ATTITUDES AND BEHAVIORS

Attitude towards Online Advertisement

Mean

1.54

1.13

3.13

4.62

SD

.92

.83

.71

.86

Online Advertising Usage

Mean

2.12

2.04

2.52

3.45

SD

.87

.96

.93

1.03

IV. INTERNET ATTITUDES AND BEHAVIORS General Internet Attitude

Mean

5.27

4.60

5.00

5.52

SD

.89

1.29

.94

.61

General Attitude towards Online Shopping

Mean

5.13

4.41

4.69

3.69

SD

1.20

1.73

1.47

2.35

Online Shopping Behavior

Mean

3.72

3.26

3.49

3.77

SD .98

1.14

.98

.96

V. MARKET PERCEPTIONS AND CONSUMER PSYCHOGRAPHICS Consumer Alienation

Mean

4.27

3.91

3.68

.87

SD

1.02

1.49

.88

1.26

Perceived General Product Quality

Mean

2.62

2.80

2.84

2.59

SD

1.29

1.41

1.06

1.63

Perceived General Price Fairness

Mean

2.52

2.11

3.20

4.22

SD

1.22

1.49

1.00

1.34

Self-Esteem

Mean

5.13

4.49

4.51

5.26

SD

.80

1.32

1.00

.67

Self-Confidence (Decision Making)

Mean

4.02

3.98

3.43

2.75

SD

1.23

1.11

1.10

1.91

Self-Confidence (Persuasion Knowledge)

Mean

4.91

4.28

4.41

4.93

SD

.83

1.38

.88

.78

CSII (Informational Influence)

Mean

3.06

2.44

3.39

4.32

SD

1.38

1.24

1.13

1.31

CSII (Normative Influence)

Mean

1.49

1.21

2.35

3.53

SD

1.35

1.12

1.32

1.86

Disposition to Trust

Mean

3.86

3.31

3.94

4.81

SD

1.25

1.24

1.00

.92

Due to a similar reasoning, this research also expects that high trusters or consumers that have high beliefs in the reliability of online advertising also have favourable attitudes and show positive approach behaviors towards this medium. Drawing on an analysis of variance, this research was able to demonstrate that consumers’ attitude towards online advertising (a = .92) is not equally distributed over the four samples (F(3,513) = 407.35, p < .001). More specifically, ad attitude had a significant positive relationship with oADmist. That is, samples with the lowest oADTrust (i.e.,pure review trusters, non-trusters) had a more negative attitude, while moderate and high trusting individuals were characterized by a reasonably higher level of devotion. As expected from ad literature, high trusters scored highest on this trait (M = 3.06, SD = 1.49).

Referring to the usage of online advertising information, the same literature proposes that consumers that posses a heightened level of generalized oADTrust also more often seek advice from this kind of market information (e.g., Soh, 2007). Assuming that high trusters find online advertising information more valuable and are deliberately looking to consume advertised products, it seems reasonable that these consumers belong to the segment of consumers who are likely to use ad information for their purchase decisions. Therefore, this research expected a positive relationship between ad trust and the indicated use of online advertising. Investigating this assumption using a one-way ANOVA, significant differences between the clusters were found (F(3,513) = 55.20, p < .001). In particular, the data demonstrated the hypothesized increase in usage patterns, with high trusters showing significantly more attention to advertising information than the other three segments. Usage behaviors did not differ significantly between the segments with low oADTrust.

Another group of segment variables targeted the description of the perceptions of the Internet as a research and shopping instrument. According to a variance analysis, significant differences among the clusters existed in respect to general Internet attitude (F(3,513) = 18.54, p < .001). More specifically, the two segments with increased trust in eWOM (i.e., pure review trusters and high trusters) both also had a more favourable attitude towards the online medium. Compared to this, low trusters were more opposed to the Internet. Another interesting characteristic of different groups of online consumers is their attitude towards and usage of online shopping. This research found that the four segments also significantly differ in these respects. More precisely, pure review trusters showed the most favourable shopping attitude which was, however, insignificantly higher compared to non-trusters and moderately trusting consumers. In contrast, high trusters demonstrated an online shopping attitude that was significantly lower when contrasted to the remaining three segments. The research further demonstrated significant differences concerning online shopping behaviors (F(3,513) = 4.98, p < .01). Subsequent analysis, however, showed that only the difference between the high trusters and the low trusters segment was significant, where the former was found to have increased usage patterns. The rest of the sample was homogeneous in this respect. In the course of the empirical investigation, the consumers’ attitude towards online shopping in general was also examined and a one-way ANOVA was able to find significant differences (F(3,513) = 15.72, p < .001). Here, pure review trusters exhibited the most positive attitude (M = 5.13, SD = 1.20), which nevertheless was only significantly higher when compared with the high truster cluster. The latter was positioned on the other end of the continuum and was characterized with very low scores (M = 3.69, SD = 1.84). In addition, non-trusters as well as moderately trusting consumers were both found to feel more positively towards online shopping compared to high trusters.

The final group of segment characteristics emphasized perceptions and attitudes towards the marketplace and additional consumer psychographics. Dramatically changing consumer attitudes make the topic of consumer alienation worth considering. According to literature, an alienated person bears a profound feeling of separation, exclusion or estrangement towards a particular social institution (Seeman, 1959). Interactions with this social entity are at the same time typically associated with unpleasant and unfavourable feelings (Krishnan et al., 2009). Some authors argue that alienation is a domain-specific phenomenon; as a consequence the topic was discussed among a variety of disciplines, including marketing (e.g., Balasubramanian & Kamakura, 1989; Burns, 2010; Johnson, 1996; Krishnan et al., 2009. Marketing scholars have heavily drawn on Seeman’s conceptualization and have investigated the extent to which shoppers feel alienated by manufacturers and vendors in the market place (Allison, 1978; Lambert, 1980). This thesis adopts the definition of Mady (2011) and views consumer alienation as the consumer’s “feelings of separation from the norms and values that characterize the typical marketplace” (p. 194). Here marketplace includes all social institutions that are involved with the offering of products/services to the consumer and all activities these entities conduct (Johnson, 1996). The shift in market power has led to an increased importance of the topic within the last years. Classic as well as current literature suggests that consumer alienation causes several negative business outcomes. For instance, Lambert (1980) argues that consumer alienation from the marketplace leads to increased dissatisfaction, as well as mistrust towards the company. In addition, alienated consumers demonstrate a variety of avoidance behaviors, such as switching to alternative market offerings, minimizing the interactions with the company and negative word-of-mouth. These consumers refrain from identifying with the market institutions, their outputs (e.g., products, brands), and are unlikely to accept usual market practices (e.g., advertising) (Pruden et al., 1974; Shuptrine et al., 1977). Mady (2011) is able to demonstrate a negative relationship between consumer alienation and sentiment toward marketing. Therefore, this research expected consumer groups that have more adversarial feelings towards the marketplace to be over-represented in the segments that show low levels of oADimst. A one-way analysis of variance showed a significant effect of segment membership on the degree of consumer criticism of business principles (i.e., consumer alienation, a = .85) (F(3,513) = 6.46, p < .001). As expected, pure review trusters were the most alienated consumers when compared to moderately and high trusting segments. They strive to bypass the information given by companies.

In contrast, further data analysis was not able to identify differences among the clusters concerning perceived product quality; however, the segments showed divergences in respect to perceived general price fairness (F(3,513) = 61.84, p < .001). While there was no difference between the pure review trusters and the non-trusting segment, perceived price fairness seemed to increase with trust in reviews/ads. Here, high trusters were most likely to believe that the general market prices are justified.

As discussed earlier, generalized trust is likely to be dependent on a person’s self-esteem, as well as self-confidence (see Chapter 3). Accordingly, these variables were also included in the segmentation approach (a: self-esteem = .86; self-confidence (decision making) = .89; selfconfidence (product knowledge) = .88) and significant differences were identified due to a series of variance analyses. Here, one of the outcomes was that pure review trusters, as well as high trusters, were both characterized by a heightened degree of self-esteem, which differentiates them from the remaining two segments but not from each other. As discussed before, this implies that individuals with high self-esteem have no problem consulting external information sources and demonstrating their lack of knowledge. In contrast, individuals scoring low on the trait are more likely to use internal sources (Bishop & Barber, 2012). Segments which showed a minor degree of oADTrust (i.e., pure review trusters and non-trusters) were characterized by a high self-confidence in making adequate purchase decisions, while high trusters had little confidence. In contrast, the segments also varied concerning the level of selfconfidence in persuasion knowledge (F(3,513) = 15.22, p < .001). Here, the two segments which put high trust in online reviews scored significantly higher. This implies that these consumers feel confident that they are able to identify misleading and fake messages.

As eWOM and online advertising are both likely to represent different kinds of social influence, it also was reasonable to include variables measuring this degree. Specifically, variations among the segments could be demonstrated concerning consumer susceptibility to interpersonal influence (CSII) both for informational (CSIhnfo, a = .86) (F(3,513) = 40.01 , p < .001) as well as normative influence (CSIINorm, a = .91) (F(3,513) = 53.32, p < .001). In respect to the former, non-trusters scored significantly lower, while high trusters indicated the highest level (M = 4.32, SD = 1.31). Normative influence was also greatest in the high truster segment, as expected. Pure review trusters and non-trusters scored relatively low on this trait. The high truster segment is additionally characterized by the desire to trust others. Disposition to trust (a = .90) varied significantly between the four segments (F(3,513) = 36.10, p < .001) and was highest (as expected) in the high truster cluster, followed by moderately trusting consumers. However, this segment did not differ from the level of interpersonal generalized trust measured in the pure review truster group.

This thesis closes with a discussion of the results and an interpretation of the above-mentioned findings in the final chapter.

 
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