Data Summary - Online technology news not founded and operated as an extension of a media grou

6 Online technology news not founded and operated as an extension of a media group (i.e. AOL founded websites)

4.1 Data Summary

國

立政治大學

‧

N a tio na

l C h engchi U ni ve rs it y

33 CHAPTER IV

FINDINGS AND ANALYSIS

4.1 Data Summary

Figure 8: Sample structure by gender

46%

54%

Sample Structure by Gender

Male Female

‧

國

立政治大學

‧

N a tio na

l C h engchi U ni ve rs it y

34 Figure 9: Sample structure by age

Figure 10: Sample structure by education level

89%

Sample Structure by Age

＜ 20 20-30 31-40

63%

33%

Sample Structure by Education Level

＜ College College

＞＝Master

‧

國

立政治大學

‧

N a tio na

l C h engchi U ni ve rs it y

Figure 11: Sample structure by cell phone brands ever used

Figure 12: Sample structure by monthly cell phone usage frequency

67%

31%

Sample Structure Number of Brands Ever Used

1-3 brands 4-6 brands 7-9 brands

24%

15%

12%

43%

Sample Structure by Monthly Cell phone Usage Frequency

＜ 1 hour 1-3 hours 3-5 hours 5-7 hours

＞ 5 hours

‧

國

立政治大學

‧

N a tio na

l C h engchi U ni ve rs it y

Figure 13: Sample structure by monthly cell phone expense

Figure 14: Sample structure by source of event gathering (More than one answer from a participant may be possible)

67%

30%

Sample Structure by Monthly Cell Phone Expense

< $200

$ 201-600

$ 601-1000

32%

21%

18%

15%

14%

Sample Structure by Respondents’ Frequent Source of Information Gathering

The Internet Television

Friends and Relatives Newspaper

Magazine

‧

國

立政治大學

‧

N a tio na

l C h engchi U ni ve rs it y

37 Data Description

Demographic Structure

From the demographic structure, it is found that almost equal percentage of female and male sample respondents answered this survey, while almost 90% of the respondents are aged 20-30, which fit the target audience in this study. Over 90% of the respondents have college or above degree.

Cell phone Usage Experience

From the cell phone usage experience figure, it is found that about 70% of respondents have used 1-3 cell phone brands, while about 40% of respondents use the cell phone for over 5 hours per month, but about 70% spent only $200 per month.

Source of Event

From respondents’ frequent source of event gathering the event gather source, it is found that the most frequent source used by respondents is the internet, followed by television, friends and relatives, newspaper, and magazine. It is not clear that why

magazine as the least frequent source used is rated the most reliable source in comparison with other sources when doing pretest and posttest for this study.

‧

4.2 Data Analysis and Discussion

This part will present tables of tests of between-subjects effects by individual dependent variable, then analyze results and discuss why or why not each hypothesis is supported and not supported hypothesis along with appropriate supporting graphs of estimated marginal means. Highlighted rows in tables of tests of between-subjects effects are those with significance over 0.1 and worth to be discussed.

Table 10: Tests of between-subjects effects by attitude change

Source

Attachment * Source * Severity

.343 1 .343 .462 .497

Error 134.485 181 .743

Total 1789.636 189

Corrected Total 224.700 188

‧

國

立政治大學

‧

N a tio na

l C h engchi U ni ve rs it y

Table 11: tests of between-subjects effects by product evaluation change

Source

Type III Sum of

Squares df Mean Square F Sig.

Corrected Model 105.248^a 7 15.035 17.061 .000

Intercept 1541.549 1 1541.549 1749.257 .000

Attachment 81.019 1 81.019 91.936 .000

Source .074 1 .074 .083 .773

Severity 3.383 1 3.383 3.839 .052

Attachment * Source .032 1 .032 .037 .849

Attachment * Severity .000 1 .000 .000 .989

Source * Severity 5.450 1 5.450 6.184 .014

Attachment * Source * Severity

.025 1 .025 .028 .868

Error 159.508 181 .881

Total 1896.456 189

Corrected Total 264.756 188

‧

國

立政治大學

‧

N a tio na

l C h engchi U ni ve rs it y

Table 12: Tests of between-subjects effects by perceived risk change

Source

Type III Sum of

Squares df Mean Square F Sig.

Corrected Model 95.344^a 7 13.621 14.804 .000

Intercept 1496.590 1 1496.590 1626.641 .000

Attachment 54.488 1 54.488 59.223 .000

Source 1.302 1 1.302 1.416 .236

Severity 15.347 1 15.347 16.680 .000

Attachment * Source .114 1 .114 .124 .725

Attachment * Severity .745 1 .745 .809 .370

Source * Severity 3.165 1 3.165 3.440 .065

Attachment * Source * Severity

.003 1 .003 .004 .952

Error 166.529 181 .920

Total 1887.468 189

Corrected Total 261.873 188

‧

國

立政治大學

‧

N a tio na

l C h engchi U ni ve rs it y

41 Hypotheses results

Hypothesis 1: Main effect of brand attachment to all dependent variables

▪ Statement: When exposed to a negative event, consumers with higher brand attachment will show less negative attitude, product evaluation, and perceived risk changes comparing to consumers with lower brand attachment

▪ Result: supported

▪ Discussion:

Table 13: Average dependent variable score by level of attachment

High attachment Low attachment

Average negative attitude change 2.24 3.59

Average negative product evaluation change 2.26 3.69 Average negative perceived risk change 2.34 3.59

Average brand attachment 3.53 2.07

From the table 10, 11 and 12, it is found that the brand attachment as an

independent variable has significant effect to consumers’ attitude, product evaluation and perceived risk changes. Furthermore, in table 13, it is found that high brand attached consumers tend to show less negative changes in attitude, product evaluation and lower perceived risk comparing to low brand attached consumers.

‧

國

立政治大學

‧

N a tio na

l C h engchi U ni ve rs it y

Consistent with the literature findings, that consumers tend to pay more attention to positive events related to their attached brand (Feldman & Lynch, 1988) to maintain altitude change in line with their previous perception about the brand (Lord, et al., 1979;

Zanna, 1993), this research finds out that high brand attachment customers tend to show less negative attitude, product evaluation and severity changes when exposed to a

negative event, comparing to low brand attached group. In other words, brand attachment has a negative relationship with customers’ overall attitude changes when consumers are exposed to negative events.

Hypothesis 2: Main effect of event source to all dependent variables

▪ Statement: When exposed to a negative event from a more credible source,

consumers will show more negative attitude, product evaluation, and perceived risk changes comparing to from a less credible source

▪ Result: not supported

▪ Discussion:

From the table 10, 11 and 12, it is found that the event source as an independent variable has no significant effect to consumers’ attitude, product evaluation and perceived risk changes.

According to the literature, source credibility will affect the level of an event’s persuasiveness, such that high credibility sources elicit more persuasion than low credibility sources. As a result, magazine source considered as a credible source with

‧

國

立政治大學

‧

N a tio na

l C h engchi U ni ve rs it y

experts’ knowledge by the respondents should have better persuasion power than forwarded email source as a less credible source. However, credibility did not affect consumers’ attitude after analyzing the survey data.

Possible reasons might be that source credibility should be considered with consumers’ elaboration process in order to make the persuasion of the credible source take effect. Heesacker et al. (1983) and DeBono et al. (1988) indicate that, source credibility works as a reference to consumers whose elaboration is neither high nor low, or is low. Since in hypothesis 2, consumers’ responses are analyzed purely from the perspective of source but not take into account of their elaboration levels, source is therefore not initiated by consumers as a credible reference. As a result, one cannot tell whether the source credibility would affect consumers’ negative attitude change, negative product evaluation and perceived risk change.

Hypothesis 3: Main effect of event severity to all dependent variables

▪ Statement: When exposed to a negative event with higher level of severity,

consumers will show more negative attitude, product evaluation, and perceived risk changes comparing to with lower level of severity

▪ Result: mainly supported

‧

國

立政治大學

‧

N a tio na

l C h engchi U ni ve rs it y

▪ Discussion:

From the table 10, 11 and 12, it is found that the severity level as an independent variable has significant effect to consumers’ product evaluation and perceived risk

changes, but not attitude change. Furthermore, in table 14, it is found that consumers who are exposed to high severity negative events tend to show higher average negative

product evaluation change and higher average negative perceived risk change.

In the literature, fear contained in the negative event can bring up individual’s level of problem elaboration, depending on the level of fear appeal. A low-fear appeal event reduces individual’s level of problem elaboration and high-fear appeal increases individual’s level of problem elaboration (Keller & Block, 1996). The findings in the research support the literature, and imply that increasing problem elaboration process would affect consumers’ product evaluation and perceived risk, but not brand attitude.

Perhaps the fear level is more related and determined by severity level, whereas brand attitude is more affected by consumers’ perception about the brand and cannot by simply shaken by a single negative event.

Table 14: Average dependent variable score by level of severity

High severity Low severity

Average negative attitude change 3.07 2.67

Average negative product evaluation change 3.2 2.66 Average negative perceived risk change 3.31 2.53

‧

國

立政治大學

‧

N a tio na

l C h engchi U ni ve rs it y

Hypothesis 4: Interactive effect between event severity and event source

▪ Statement:

4 (a): When consumers receive a negative event with higher level of severity from a more credible source, they will show more negative attitude, product evaluation, and perceived risk changes comparing to from a less credible source

4 (a): When consumers receive a negative event with lower level of severity from a more credible source, they will make no difference in negative attitude, product evaluation, and perceived risk changes comparing to from a less credible source

▪ Result: supported

▪ Discussion:

Figure 15: Estimated marginal mean of attitude by severity and source

high severity

low severity

less credible more credible

‧

國

立政治大學

‧

N a tio na

l C h engchi U ni ve rs it y

From the table 10, 11 and 12, it is found that interactive effect between event severity and event source does exist. Graphs of estimated marginal mean of each dependent variable are provided to enable more analysis.

Figure 15 that provide interactive effect between event’s severity and source to attitude change show that the p-value of the two points along the vertical axis of less credible source is more than 0.25, and imply that regardless of high or low severity, consumers would behave similarly when the piece of event is from a less credible source.

Furthermore, figure 15 show that when respondents receive a negative event from a more credible source, high severity involved event would trigger more attitude change to customers comparing to low severity event.

Figure 16: Estimated marginal mean of evaluation by severity and source

high severity

low severity

less credible more credible

‧

國

立政治大學

‧

N a tio na

l C h engchi U ni ve rs it y

Figure 16 that provide interactive effect between severity and source to attitude product evaluation change, show the p-value of the two points along the vertical axis of less credible source is between 0.2-0.25, and imply that regardless of high or low severity level, consumers would behave similarly when the piece of event is from a less credible source. Furthermore, figure 16 show that when respondents receive a negative event from a more credible source, high severity involved event would trigger more product

evaluation change to customers comparing to a low severity event.

Figure 17: Estimated marginal mean of severity by severity and source

Figure 17 provide interactive effect between severity and source to perceived risk change show the p-value of the two points along the low severity involved line on less

high severity

low severity

less credible more credible

‧

國

立政治大學

‧

N a tio na

l C h engchi U ni ve rs it y

credible source and more credible source is found to be greater than 0.25, which imply that consumers behave similarly when receiving low severity involved event.

Furthermore, figure 17 show that when respondents receive a high severity involved event, a more credible source would trigger more perceived risk change comparing to a less credible source.

In sum, it can be inferred from figure 15, 16, and 17, when consumers receive a high severity involved negative event, they tend give more negative rating in favorability, attitude, evaluation and severity dimensions, and such negative rating becomes stronger as the source of the events moves from a less credible source to a more credible source.

On the other hand, consumers who receive a lower severity involved events tend to demonstrate less negative favorability, attitude. Also, they tend to behave similarly regardless of the source credibility.

Hypothesis 5: Interactive effect between brand attachment and event source

▪ Statement:

5 (a): When consumers with higher brand attachment receive a negative event from a more credible source, they will show less negative attitude, product evaluation, and perceived risk changes comparing to from a less credible source

5 (b): When consumers with lower brand attachment receive a negative event from a more credible source, there will show more negative attitude, product evaluation, and perceived risk changes comparing to from a less credible source

‧

國

立政治大學

‧

N a tio na

l C h engchi U ni ve rs it y

▪ Result: not supported

▪ Discussion:

From the table 10, 11 and 12, it is found that the brand attachment has no interactive effect with event source to consumers’ attitude, product evaluation and perceived risk changes. As a result, hypothesis 5 is not supported.

According to biased assimilation theory, brands consumers’ commitment will affect how they perceive negative event about their brands, so it is originally assumed that high committed consumers tend to reject an event that threatens one’s belief to a higher degree, whereas low committed consumers tend to accept negative events.

However, biased assimilation only works when singly speaking about consumers brand attachments’ effects to consumers’ brand attitude, product evaluation and

perceived risk changes, but not when it interacts with events source. Perhaps, consumers did not consider attachment to the brand with the sources of negative events, and vice versa holds true too.

‧

國

立政治大學

‧

N a tio na

l C h engchi U ni ve rs it y

Hypothesis 6: Interactive effect between brand attachment and event severity

▪ Statement:

6 (a): When consumers with higher brand attachment receive a negative event with higher level of severity, they will show less negative attitude, product evaluation, and perceived risk changes comparing to consumers with lower brand attachment.

6 (b): When consumers with lower brand attachment receive a negative event with higher level of severity, they will show more negative attitude, product evaluation, and perceived risk changes comparing to consumers with lower brand attachment.

▪ Result: not supported

▪ Discussion:

From the table 10, 11 and 12, it is found that the brand attachment has no interactive effect with event source to consumers’ attitude, product evaluation and perceived risk changes.

Similar to hypothesis 5, brand attachment is found to play a role when singly speaking to consumers’ brand attitude, product evaluation and perceived risk changes, but not when brand attachment interacts with event source.

Perhaps consumers regard brand attachment separately with negative event severity, thus brand assimilation did not seem to work as a protection wall between consumers and the negative event.

‧

4.3 Hypothesis Result

# Hypothesis Result

1 When exposed to a negative event, consumers with higher brand attachment will show less negative attitude, product evaluation, and perceived risk changes comparing to consumers with lower brand attachment

Support

2 When exposed to a negative event from a more credible source, consumers will show more negative attitude, product evaluation, and perceived risk changes comparing to from a less credible source

Not support 3 When exposed to a negative event with higher level of severity,

consumers will show more negative attitude, product evaluation, and perceived risk changes comparing to with lower level of severity

Mainly support 4 4 (a): When consumers receive a negative event with higher level of

severity from a more credible source, they will show more negative attitude, product evaluation, and perceived risk changes comparing to from a less credible source

4 (b): When consumers receive a negative event with lower level of severity from a more credible source, they will make no difference in negative attitude, product evaluation, and perceived risk changes comparing to from a less credible source

Support

5 5 (a): When consumers with higher brand attachment receive a negative event from a more credible source, they will show less negative attitude, product evaluation, and perceived risk changes comparing to from a less credible source

Not support

6 6 (a): When consumers with higher brand attachment receive a negative event with higher level of severity, they will show less negative attitude, product evaluation, and perceived risk changes comparing to consumers with lower brand attachment.

Not support

Table 15: Hypothesis results

‧

國

立政治大學

‧

N a tio na

l C h engchi U ni ve rs it y

52 CHAPTER V

CONCLUSION

在文檔中事件嚴重性、來源可靠性、品牌依賴程度對消費者態度和產品評估改變的影響 - 政大學術集成 (頁 40-59)