• No results found

Fear Patterns using EDR (arousal) vs. CRM (fear)

SI &

6.6 Results

6.6.2 Fear Patterns using EDR (arousal) vs. CRM (fear) CRM-fear and EDR arousal of Advertisement 8-"4 WD-Relief

The fear and arousal responses to Advertisement 8 - "4WD-Relief (refer to Figure

6.9) show a fear-reliefpattern, despite the cognitive fear graph reflecting more sustained fear after the major shock component of the advertisement. The fear measure indicates that at the end of the advertisement the audience experienced significant fear reduction that was similar to the feedback provided by the arousal measure.

Figure 6.9 - Advertisement 8 "4WD-relief' - Comparison of C R M - f e a r and E D R arousal

C R M - F e a r E D R - Arousal


Duration of ad

Fletcher (1994) noted that "current tendencies in psychological findings suggest that electrodermal data should not be expected to correspond closely to such cognitive experiences as conscious remembering or opinions, especially when verbalised"


The fear and arousal responses to the fourth advertisement, 4WD, were the most similar in comparison to the other seven advertisements tested. It could be that the other advertisements were stimulating other emotions (not labelled on the dial) that caused arousal, whereas the fourth advertisement stimulated only fear (expressed as tension) that caused arousal.

It can be seen, particularly demonstrated by the Trike advertisements, that a major limitation of EDR is the inability to determine the directionality of the affect

(like/dislike) component and in the instance of fear appeals it is not definite (although is highly likely) that the arousal is caused by fear. All that can be assumed, if changes in skin conductance occur, is that the viewer is feeling increased arousal. It is important to note that when using EDR to test fear appeal advertisements that are designed to evoke fear, researchers should not use the label "fear" as EDR may indicate arousal caused by fear and arousal caused by other emotions, such as excitement. Given that the EDR is not able to determine directionality of affect when testing fear appeal commercials, it would seem that the cognitive fear CRM-dial measure may be a better way of

measuring fear and relief responses to fear appeal commercials.

There are various sources of research that argue differently and support the use of EDR measures, such as skin conductance, over cognitive measures that require verbalising

feelings, such as the dial measure (Hopkins and Fletcher, 1994). LaBarbera and

Tucciarone (1995, p.49) interviewed commercial marketing practitioners and found that

"it is more accurate to rely on the GSR ratings" (compared to self-report data), using market-place responses as the dependent variable for commercial products and services.

Similarly, Vaughn (1980, p.31) stated that "emotional arousal (autonomic,

psychogalvanometer) tests may be more helpful in determining advertising effect".

EDR has mainly been used in commercial applications that have tested mostly positive appeal commercials. In this instance it is generally safe to assume that arousal is indicating the viewer's positive affect and interest in components of the advertisement.

A negative appeal commercial, such as the anti-speeding fear appeal commercials tested in this experiment, may cause interest, excitement, fear and relief, thus when testing the theory of the effect of fear patterns within these commercials an instrument that

measures viewers' fear and relief responses, as opposed to other emotions, is required.

The CRM-dial instrument provides this opportunity. It is possible, although highly unlikely, that the differences between audiences, rather than the ads, may have contributed to the differences in results.

Overall, it does appear from most of the graphs that viewers' cognitive appraisal of their emotions are slightly different to their actual arousal, however, most of the graphs show that the patterns of affect are similar enough to confirm the fear pattern within each advertisement. That is, the fear patterning component of Study 3 showed that when using either measure of fear or arousal the advertisements can still be classified according to "patterns" of fear-relief or fear-only. That is, fear-only pattern

advertisements (Advertisements 3,4, 5 and 6) demonstrate that there is no or very slight reduction in fear at the end of the advertisement. Whereas, the general characteristic of

fear-relief pattern advertisements (Advertisements 1, 2, 7 and 8) is that during the middle component of the advertisement, peak fear is reached, and then the fear is

reduced towards the end, that is often related to recommendations, resulting in relief felt by the viewer.

EDR for the control advertisement were not obtained as the dishwashing detergent advertisement had been a priori chosen as a neutral advertisement. In hindsight, it would have been useful to have this data so as to separate testing effects (i.e., any emotional reaction to the experimental situation) from exposure effects (i.e., responses to fear-based advertisements). Thus comparisons cannot be made between the

experimental advertisements and the control advertisements in regard to psychophysiological reactions.

6.6.3 Calculation of Fear Reduction and Arousal Reduction

The graphs can be used to determine arousal reduction, using the EDR measure, or fear reduction, using the cognitive fear measure (the CRM dial). The percentage reduction in fear (or arousal) is calculated by finding the average maximum response, then finding the average minimum response after the maximum. That is:

Fear reduction = (Dialmax - Dialmin after max) / Dialmax * 100 or

ArOUSal reduction = (EDRmax - EDRmin after max) / EDRmax * 100

Table 6.4: Comparison of arousal-reduction and cognitive fear-reduction between fear-relief advertisements and fear-only advertisements


Fear-only Fear-relief


154 157


Peak fear (avg)

194 195

Base relief (avg)

83 -2

Fear reduction




n 80 84

Peak arousal

(avg) 4.09 5.36

Base arousal

(avg) 1.54 0.89

Arousal reduction



Table 6.4 shows the results of C R M - F e a r and EDR-Arousal for the fear-only advertisements combined (Advertisements 3,4, 5 and 6) and the fear-relief advertisements combined (Advertisements 1, 2, 7 and 8). The responses from the CRM-dial cognitive fear measurement show that there was no difference between the sets of advertisements in regard to the average peak fear (maximum) reached (F = 0.02, p = .897). This was an expected result as the manipulation of the advertisements, that is the construction of paired counterparts, was not meant to affect the fear responses of participants between the groups. There was a difference between the sets of

advertisements in regard to the average base relief (minimum) score (F = 56.7, p = .000). This was also expected as the manipulation of the advertisements was meant to alter the relief components in the advertisements. As anticipated, the fear-relief advertisements resulted in lower base relief scores than the fear-only advertisements.

There were also significant differences in fear reduction between the types of

advertisements (F = 47.0, p = .000), with/ear-re//e/advertisements resulting in 101%

fear reduction and fear-only advertisements resulting in 57% fear reduction.

Similar analysis was undertaken for the arousal (EDR) measure. Again, there was no difference in the average peak arousal reached (F = 2.40, p = .123) between the advertisements. There was no difference in the average minimum score (base arousal)

(F = 2.73, p = . 100). There w a s also a difference in arousal reduction (F = 9.26, p = .003), with/ear-re/ze/advertisements resulting in 83% reduction and fear-only advertisements resulting in 62% reduction. While the/ear-re//e/advertisements

produced significantly greater reduction, the difference between the advertisements was not as great as when measuring fear reduction. As discussed previously, this is likely to be due to the other emotions in addition to fear that register as arousal.

In summary, regardless of which measure is seen as superior, both the fear and arousal measures have demonstrated that the manipulation of the advertisements was successful in terms of creating a set of four^ar-re//e/advertisements and four fear-only


6.6.4 Effects of Advertisements on A VST scores

Analysis of variance was used to examine differences between the mean speed scores on the AVST generated by the respective experimental advertisements and the control advertisement. The AVST is known to differ in effects by gender (male vs. female drivers, with male drivers tending to speed more than females in all situations, on average) and by speeder type (speeders vs. non-speeders, with speeders tending to speed more than non-speeders in all situations). The results are presented for the total sample and then by gender and speeder type. Comparisons were as follows:

Between individual advertisements within fear-relief and fear-only pairs, for the total sample of young drivers

Between fear-relief advertisements and fear-only advertisements overall, for the total sample of young drivers

Between/ear-re//e/advertisements and fear-only advertisements, for males vs.


Between/ear-refte/advertisements and fear-only advertisements, for speeders vs.


Betweenyear-re/ze/advertisements and fear-only advertisements, for female speeders vs. female non-speeders

Between/ear-re//e/advertisements and fear-only advertisements, for male speeders vs. male non-speeders. Individual advertisements: Total Sample

AVST results for the individual advertisements in their fear-relief vs. fear-only pairs, for the total sample of young drivers, are shown in Table 6.5. The predicted

relationship between the sets of advertisements was thatTear-re/ze/advertisements would be more effective and hence produce lower speed scores than fear-only

advertisements. Firstly, the control advertisement resulted in an average driving speed increase of+5.1 km/hr. All of the four advertisement pairs showed the predicted

directional relationship, (although only 3 were statistically significant) based on total average speed scores, between the relief'versus only versions with the fear-relief versions tending to result in lower speed scores than their fear-only versions. The statistical differences found between the advertisements (and reported in Table 6.5)

resulted from differences between thejfear-re/te/advertisement group score and the control group score.

Table 6.5: Experiment groups' and control group's A V S T average scores for individual advertisements: Total sample