7. Analysis
7.2. Perceived quality
and if the flight patterns and facilities that are offered by more airline companies are similar the brand will be determining what flight ticket the customer will end up purchasing. But if these factors are not considered similar then the brand will have no effect on the purchase decision. This implies that brand awareness do have an impact but only when companies are offering similar products.
the flight arrives on time, and consistent service are all indicators of the quality of the airline to the respondents.
7.2.2. Perceived quality of BA
To assess the perceived level of assurance of BA all respondents were asked two questions concerning assurance. They were asked to what extend do you agree with the following statements 1: It is safe to travel with BA, and 2: BA’s staff are professionals in their line of work.
The median for both questions are 4, indicating that most people answered agree to both statements. However there were also a large amount of the respondents who answered don’t know.
In relation to the flight patterns a single question was asked: to what extent do you agree that BA offers an attractive range of travels. The results for the flight patterns are similar to the level of assurance. This means that the median is 4, and that a high number of respondents answered don’t know. So, the people who do didn’t answered don’t know evaluate the quality of the airline to be good.
To assess the level of reliability one question was asked: to what extent do you agree that BA is a reliable airline? The respondents answered in a similar way where the median is 4, indicating a positive assessment of the reliability of BA, but also a large number of respondents answered don’t know to this question as well.
Strongly disagree Disagree Neither agree nor disagree
Agree Strongly agree Don't know
Figure 7.6. Perceived quality of BA
Assurance Reliability Flight patterns
The levels of perceived quality of all three factors can be seen in figure 7.6. The large number of respondents who answered Don’t know indicate, that even though the respondents are familiar with BA, they do not necessarily have an opinion about the quality of their brand. Thus, it seems that those who are familiar with the brand have a positive understanding of the level of quality in BA, but a large number of the respondents are not able to evaluate the level of perceived quality.
Approximately a third of the respondents have not evaluated the brand quality, indicating a general lack of perceived quality. That does however not mean respondents have a negative opinion – as only very few answered that they have a negative opinion about the brand quality. It only means that they have not yet formed any opinion on the level of quality of the brand. As quality indicators are considered important when customers consider which product to purchase, the lack of knowledge about the brand is considered to be negative for BA.
The mean for the quality indicators are 4.00 (assurance), 4.02 (reliability), and 3.83 (flight patterns), indicating that BA is valued a little lower in quality based on flight patterns, than on the two other indicators. In table 7.4 there is an overview of the means of the quality indicators from both the questionnaire and the two experiments. It seems that all means have decreased as a result of the experiments, except the level of assurance after experiment A, where the level of perceived assurance has increased from 4.00 to 4.08.
Assurance Reliability Flight patterns
Questionnaire 4.00 4.02 3.83
Experiment B 3.88 4.00 3.67
Experiment A 4.08 3.94 3.79
Table 7.4. Mean levels for perceived quality
An independent samples t-test for experiment B compared with the questionnaire reveals that F = 10.26 for assurance, F = 5.21 for flight patterns, and F = 1.90 for reliability. The degrees of freedom are (2, 167), (2, 157) and (2, 161) respectively and only the F-value for assurance and flight patterns are above the critical value of 3.04 (Cozby, 2009) indicating that there is a significant difference in the variances. However, with equal variances not assumed for assurance and flight patterns and equal variances assumed for reliability the significance level for the t-test for equality of means are .368 (assurance), .261 (flight patterns), and .851 (reliability) indicating that none of the differences in the means can be explained by the experiment.
The same method has been applied for experiment A, and all the results can be seen in table 7.5 together with the results from experiment B. The important point here is, that none of the differences in the mean values were caused by the experiments.
Experiment A F Sig. t df Sig. (2-tailed)
Assurance 1.443 0.231 -0.758 171 0.45
Flight patterns 0.099 0.754 0.311 152 0.757
Reliability 0.272 0.603 0.692 160 0.49
Experiment B F Sig. t df Sig. (2-tailed)
Assurance 10.264 0.002 0.91 167 0.368
Flight patterns 5.206 0.024 1.138 157 0.261
Reliability 1.902 0.17 0.188 161 0.851
Table 7.5. Independent samples t-test
The fact that the responses from both the product placement and nudging experiments indicates that the direction of the relationship is different that hypothesized. A negative relationship between product placement or nudging and the perceived quality actually indicates that the marketing initiatives are not helpful to the level of perceived quality – it is actually decreasing the value of the brand. But as none of the results are significant it must be assumed that the negative change in perceived quality is random. Regardless of the differences measured in the mean values the overall perceived quality seems to be high among the respondents who actually had an opinion on the subject.
Table 7.6 shows that in the questionnaire 41.1% of the respondents answered don’t know to the question concerning facilities, 31.6% to the question concerning reliability and 38.5% answered don’t know to either one or two of the questions concerning assurance.
Assurance Reliability Facilities
Don't know Answered Don't know Answered Don't know Answered
Questionnaire 41.1% 58.9% 31.6% 68.4% 38.5% 61.5%
Experiment A 25.0% 75.0% 15.0% 85.0% 25.0% 75.0%
Experiment B 40.8% 59.2% 28.6% 71.4% 40.8% 59.2%
Table 7.6. Level of respondents who answered don’t know
The change in the level of respondents that answered don’t know have not changed much as a result of experiment B. But the levels of all three factors seem to have decreased considerably as a
result of experiment A and therefore a chi-square test has been made to establish if the changes are significant. Pearson’s chi-square for the reliability-factor is 4.417 with an asymptotic significance of .036. This tells us that the level of respondents who answered don’t know to the question concerning reliability has changed at a significant level as a result of experiment A.
However, for the other two factors the chi-square values are 2.604 (assurance) and 3.605 (facilities) with p-levels of .107 and .058 respectively, which tells us that the changes in these two factors are not significantly different from the questionnaire at a 95% confidence level. So even though fewer respondents answered don’t know as a result of the experiment, only the changes in the reliability-factor were significant, and can thus be described as an effect of the independent variable.
7.2.3. Price as a quality indicator
Additionally to the established quality indicators in the industry theory on perceived quality argues that price can be an indicator of quality. Thus, products that are considered to be of high quality will also be perceived to be more expensive. To evaluate how respondents evaluate the price-level of BA they were asked to answer the question to what extent do you agree that it is cheap traveling with BA?
The mean value of the questionnaire is 3.10 and the median is 3, which means that generally respondents neither agree nor disagree that the prices traveling with BA are low. Again a high share of the respondents have answered don’t know indicating that they have no perception of the price range at BA. The mean for experiment B is 2.62 with a median of 3, and therefore initially the same conclusion as from the questionnaire is valid for experiment B. Furthermore, it can be seen in figure 7.7 that the level of people who disagree has increased as a result of experiment A from
1%
8%
26%
13%
1%
37%
5%
50%
20%
15%
0%
10%
2%
22%
31%
4%
0%
41%
Strongly disagree Disagree Neither agree nor disagree
Agree Strongly agree Don't know
Figure 7.7. Evaluation of price
Questionnaire Experiment A Experiment B
8% to 50%. The mean value for experiment A is 2.5 and the median is 2, suggesting that the participants in experiment A disagrees that it is cheap traveling with BA. As the dependent variable is interval-scaled and the independent variable is nominal in Analysis of variance (ANOVA) have been made to test whether the experiments have had any significant impact on the mean values.
The results from the ANOVA can be found in appendix G. The F-value is 10.53 with a significance of .000. The F-value is greater than the critical value why we have to assume that there are some differences between the groups. However a test of the homogeneity of variances tells us, with a p-level of .195 that we cannot reject the possibility of some of the groups have similar variances, and we therefore have to apply Scheffe’s test for variability to determine the differences in mean values.
95% Confidence Interval Mean Difference Std. Error Sig. Lower Bound Upper Bound
Product placement .48211 0.1608 .013 0.085 0.8792
Nudging .60280 0.14799 0 0.2373 0.9683
Table 7.7. Scheffe’s Test for multiple comparisons (Price)
Scheffe’s test actually shows that there has been a significant decrease in the mean values both as a result from experiment A and B. As the mean has decreased it indicates that both experiments have affected participants to believe that the price for flying with BA are high. Thus, if the perceived level of price is indicating a perceived quality the participants in both experiments are expressing an increased level of perceived quality.