H5: The commercials lead to different degrees of purchase intention

84

6.1.5. H5: THE COMMERCIALS LEAD TO DIFFERENT DEGREES OF PURCHASE

85 Figure 22. Comparison of languages in terms of purchase intention. Y-axis showing Kruskal-Wallis mean rank value.

Table 11. Display of what purchase intention the languages evoke.

PURCHASE INTENTION

BUY BRAND BUY BRAND

AS GIFT INQUIRE ABOUT

BRAND RECOMMEND

BRAND

SWEDISH COMMERCIAL

INTENTION SIGN. DIFFERENCE

Overall purchase intention Inquire about brand Buy brand

Buy brand as gift Recommend brand

ENGLISH COMMERCIALS

INTENTION SIGN. DIFFERENCE

Overall purchase intention Inquire about brand Buy brand

Buy brand as gift Recommend brand Significantly higher than the other language

Significantly lower than the other language Not significantly different from the other language

Significantly higher than the other language Significantly lower than the other language Not significantly different from the other language

86

Different English accents

In order to see if different English accents affect purchase intention the following hypothesis have been tested:

H5b: At least one of the English commercials creates a higher likelihood of purchase than the others

The results reveal a generally greater likelihood of purchase among people who have watched the American commercial. The results are significant on an overall level (p=0.002), but also for the sub-variables. Respondents who watched the American commercial scores higher compared to those who watched the Australian commercial, both in terms of likelihood of inquiring about the brand (p=0.020), buying the brand (p=0.012), purchasing the brand as a gift to someone else (p=0.001) and recommending the brand (p=0.009). In addition to this, the American commercial also scores higher than the British commercial in terms of likelihood of buying the brand (p=0.010). An overview of the results are found in Table 12. The outcome leads to H5b being supported. A presentation of the results can be found in Figure 23 and Table 12.

Figure 23. Comparison of accents in terms of purchase intention. Y-axis showing Kruskal-Wallis mean rank value.

Significantly higher than at least one other accent Significantly lower than at least one other accent Not significantly different from the other accents

PURCHASE INTENTION

BUY BRAND BUY BRAND

AS GIFT INQUIRE ABOUT

BRAND RECOMMEND

BRAND

87 Table 12. Display of what purchase intention the accents evoke.

Conclusion: H5 is supported, because there is evidence for the commercials resulting in different likelihood of purchase. However, differences can only be found when comparing the English accents.

Analysis H5

The clear differences between the English accents, together with the insignificant difference between Swedish and English, raise the question of whether the scores from the English commercials erase each other (in H5a) and that a Swedish commercial can actually result in a higher or lower likelihood of purchase than the other commercials. Even though testing this is not in the scope of the hypotheses, it is interesting to run the analysis from a statistical standpoint. Doing this reveals that the American commercial itself actually creates higher level of purchase intention than the Swedish one. The result makes it less obvious as of whether or not H5a is supported. However, to keep the analysis as objective as possible, it is decided to maintain the decision of not accepting the first sub-hypothesis.

Scholars who have previously investigated the relationship between accent and purchase intention agree that it is not the accent per se that affect whether consumers buy a product or not. Instead there are mediating factors which are dependent on accent and in turn affect purchase intention. What these mediating factors are, are however disagreed upon. For example, Tsalikis et al. (1991) claim that competence is a mediating factor. DeShields et al. (1996), on the other hand believe source credibility is an important factor, while DeShields et al. (1997) in another article argue for the relationship strength

BRITISH COMMERCIAL

INTENTION SIGN. DIFFERENCE

Overall purchase intention Inquire about brand Buy brand Buy brand as gift Recommend brand

AMERICAN COMMERCIAL

INTENTION SIGN. DIFFERENCE

Overall purchase intention Inquire about brand Buy brand Buy brand as gift Recommend brand

AUSTRALIAN COMMERCIAL

INTENTION SIGN. DIFFERENCE

Overall purchase intention Inquire about brand Buy brand Buy brand as gift Recommend brand

Significantly higher than at least one other accent Significantly lower than at least one other accent Not significantly different from the other accents

88

towards the accent being an indicator of purchase intention. With this as foundation of knowledge, correlation with purchase intention was tested for several variables (brand personalities, credibility and respondent’s accent) in order to see if any mediating factors could be found. In accordance with previous research, both competence and credibility correlate positively with purchase intention (competence:

τ=0.435, p<0.001; credibility: τ=0.403, p<0.001). Looking specifically at the English commercials, there is also higher likelihood of purchase if the respondent speaks the same accent as the person in the commercial (p=0.001). The result indicates that all three (competence, credibility, relationship strength) can be predictors of purchase intention. When comparing the Swedish and English commercials, however, the predictors become less straightforward. Most noteworthy, is that there is a significant difference in perceived competence between the Swedish and English versions, but no difference in purchase intention can be found.

In addition to what previous scholars mention, the correlation between likability and purchase intention was also explored, because of the subjective (but hopefully accurate) estimate that consumers rather buy brands they like than brands they dislike. However, similar to the competence variable, the results contravene. There are differences in likability between the Swedish and English commercial, even though no difference is found in purchase intention. Furthermore, while there is no support for respondents liking any English accent more than another, they are more likely to buy a product if they have seen the American commercial.

Because of the contradictions in the results, it is hard to interpret whether there are mediating variables between accent and purchase intention, and if so which these variables are. The only thing that could be said with certainty is that there are differences in purchase intention depending on what English accent the respondents have listened to. Why these differences exist is less obvious. The main purpose of this study, however, was to investigate accents effect on brand perception. The analysis of purchase intention is rather an extension of the research question to better understand accents’ effects on firms’ financial equity, i.e. do differences in brand perception actually have effect on firms’ financial value. A deeper analysis of the underlying factors for the differences in purchase intention was therefore considered to be beyond the scope of this study.

89