• No results found

Method .1 Overview



4.3 Method .1 Overview

There were three stages involved in the process of identifying the fear patterns within the anti-speeding TV commercials: first, selecting the TV advertisements for fear

patterning; second, data collection using a group of participants to undertake the continuous measurement dial procedure; and third, data collection from a separate group

of participants for the static measurement of the advertisements (to allow for

comparison of dynamic and static measurements). These stages are detailed below.

4.3.2 Stage I: Selection of TV Commercials

Television commercials were chosen as the stimuli for the study because they are the main medium used by road safety advertisers.

Initially, a large set of road safety TV commercials (n=120) were collected from road safety authorities around Australia (Roads and Traffic Authority, New South Wales;

Transport Accidents Commission Victoria; Queensland Department of Transport;

Western Australian Office of Road Safety; Transport South Australia and the

Tasmanian Road Safety Authority). Anti-speeding advertisements made up the largest percentage (38%, that is, 45 of 120) of the road safety advertisements and were chosen as the applied focus of the thesis. A subset of 36 of the 45 anti-speeding advertisements was judged to involve a threat and potentially evoke fear.

Content analysis of the advertisements was undertaken, using a set of criteria for the selection of the television commercials to produce a set of anti-speeding advertisements that had consistent content for the presented threat. The final set of advertisements was selected based on the following criteria: first, the bad behaviour portrayed was

primarily speeding, and was not confounded by other acts of dangerous driving such as ranning stop signs, running red lights, not wearing seatbelts, inattention, fatigue, and intoxication; second, the advertisement focused on driving within the speed limits on local (suburban) streets ; third, the negative physical consequence was the portrayed death of another person (not driver death); fourth, the psychological consequences

portrayed b y the driver were grief and remorse; and fifth, the driver w a s visible in the advertisement, so that identification with the driver could be measured.

Advertisements from the NSW Roads and Traffic Authority were eliminated from

consideration for the advertising experiment because it was to be conducted in that state and the avoidance of prior exposure to the advertisement was considered very

important. From the remaining set of out-of-state anti-speeding TV advertisements, four were chosen based upon their common theme of speeding in local streets and

hitting pedestrians as a result of speeding. More importantly, the advertisements were expected to represent fear-only and fear-relief"patterns (two of each pattern). The test advertisements were from those made by the Victorian and Western Australian road

safety authorities.

4.3.3 Description of the TV Advertisements used in the Study

Written summaries of the test advertisements are provided below. Copies of the advertisements are included in Appendix 4 that is a CD-ROM. Expected Fear-Relief Advertisements

The first advertisement, Advertisement 1 - "Pizza" (60 seconds), showed a pedestrian being hit by a speeding car, then a surgeon commenting on how speed caused the fatal injuries, followed by a second sequence of visuals re-enacting in slow motion the

pedestrian's body being hit by the car, concluding with a further recommendation by the surgeon to reduce speed, while at the same time a scenario was shown of a car travelling below the speed limit and avoiding hitting a pedestrian.

T h e second advertisement, Advertisement 2 -"Pram" (30 seconds) showed three

different speeding scenarios: a driver was shown avoiding, frightening, then hitting a pedestrian who was pushing a pram. The voice-over commentary in the advertisement provided an explanation of the different consequences of faster and slower speeds. The visuals for the final part of the advertisement were a black screen with white writing that slowly revealed the statement "Every 1 Okm/hr can increase your braking distance by up to 50%". Expected Fear-Only Advertisements

The third advertisement, Advertisement 3 - "Trike" (30 seconds), began with

youngsters riding tricycles on a driveway and ended with one of the children riding onto the road and being run over and killed by a speeding motorist.

The fourth advertisement, Advertisement 4 - "4WD" (60 seconds), showed a young mother who was mnning late picking her child up from school driving a four-wheel drive (4WD) and speeding recklessly through local streets, and hitting and killing another person's child.

4.3.4 Stage 2: Research Process for CRM Responses

Two groups of 30 participants from a first-year undergraduate marketing class (age

range 18-25 years) were recruited for the CRM phase of the study (refer to Chapter 3 for a description of the CRM device), and they watched six anti-speeding advertisements during individual sittings. Thus a total of 12 anti-speeding advertisements were fear patterned using the CRM procedure (note: most of the advertisements generally fell into the classifications of either fear-relief or fear-only advertisements). However, for the

purposes of continuity in this study only four advertisements were selected and tested in

all three studies and only these four are described in this chapter. The graphs for the remaining advertisements have been included in Appendix 5. Three differently ordered rotations of the advertisements were used to average out potential carryover or desensitising effects of watching six commercials.

Ethics approval was provided by the University of Wollongong's Human Research Ethics Committee. Healthy members of the population anonymously took part in the study, hence there were no major ethical considerations related to this study.

Participants were offered movie money (to the value of $ AUDI 3) for approximately 10 minutes of their time.

4.3.5 Stage 3: Research Process for Static Measures of Fear and Relief

A separate sample of first-year undergraduate marketing students was recruited for the static ratings phase of the study. Static measures of advertisement reactions were obtained from a post-exposure questionnaire for the four advertisements. Participants watched one TV commercial in classroom groups (n ranged from 51 to 116 per TV

advertisement). The advertisement reactions measures included ratings of the extent to which the participants felt "tense" and "relief, respectively, while watching the commercial. The answer options were 0 = "not at all", 1 = "slightly", 2 = "quite", 3 =


There were several reasons for using this four-point scale. First, the popularly used

semantic differential scale, that requires bipolar opposites, was not relevant in this case, for example the opposite of "extremely tense" could not be "extremely not tense". It

obtaining reactions to fear appeal advertisements did not produce relaxed feelings. It

was also believed that five-point scales could have a tendency to encourage neutral or middle of the range responses, thus using a four-point scale avoided this potential problem. Furthermore, participants were still offered a more 'neutral' option by the response category of "not at all".

4.4 Results 4.4.1 Overview

The outcome of the fear patterning dial recordings is that there is a definite sequence of emotions taking place when viewers are watching these anti-speeding advertisements

that use fear appeals. The CRM provides temporal (continuous) response patterns for each of the four TV commercials. The advertisements, based on average scores, are

classified into one of the two distinct fear pattern groups: fear-relief 'and fear-only. The general characteristic of fear-only advertisements is that there is no or very slight

reduction in fear at the end of the advertisement. The viewer feels increasingly tense as the advertisement progresses, and is left at the end of the advertisement feeling very tense. The general characteristic of fear-relief advertisements is that during the middle component of the advertisement, peak fear is reached, and then the fear is reduced

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

4.4.2 Fear Patterns (Dynamic Ratings)

Each participant's fear pattern response to every advertisement he/she watched was inspected (thus there were 30 CRM patterns for each advertisement). Examples of individual graphs are first shown for each advertisement in Figures 4.1 to 4.4. The individuals' response patterns for each advertisement were highly consistent, although

the precise magnitudes of peaks and valleys were slightly different for each individual.

A few obvious outlying recordings for each of the advertisements were deleted from the data and then average scores for each advertisement were used to form a graph of the overall pattern of fear, and where relevant relief, felt by viewers when watching a particular advertisement. The outliers that were discarded from the analysis mainly included "flatliners" who did not turn the dial at all during their viewing of the ad.

Fortunately, this only occurred with a small percentage of respondents.

The individual graphs remaining did not necessarily all conform to the expected fear-only or fear-relief pattern for each specific advertisement type tested. These results could not be considered outliers as it is possible that jfear-re/z'e/advertisements may cause some viewers to still feel fearful at the end of the advertisement. For example, in regard to Advertisement 1 - Pizza, despite 90% of responses showing a downward

movement towards relief at the end of the advertisement, the responses of participants' 15 and 16 stayed high. The combined scores for Pizza still resulted in a fear-relief pattern. This study was focused on average or group responses versus individual responses as advertisers send messages to target audiences rather than to specific individuals. Future studies could investigate these individual responses.

The group mean patterns of felt fear and relief are shown for the four advertisements in Figures 4.1 a to 4.4a. It should be noted that while the range on the dial for the CRM process was from -240 to +240, corresponding to "very relieved" through to "very

tense" (respectively), the minimum score for the group mean patterns was -50 and therefore only the range of -50 to +240 is shown on the group mean graphs.

Figure 4.1: E x a m p l e s of individual participants' C R M responses to advertisement



150 S



'articipant 1 - Pizza

'A^A 4

i a *> - . . ? H . . a u o n i > . >

— H


1 2 5 0


50-I •

1 "Pizza"

Participant 4 - Pizza




Dmiion of ad

Participant 5 - Pizza

250 T 1






r ^ v — v

1 ^ 1

Duration ofad

Participant 6 - Pizza


Participant 10 - Pizza Participant 15 - Pizza

Participant 16 - Pizza Participant 18 - Pizza Participant 2 0 - Pizza

Participant 23 - Pizza Participant 26 - Pizza






1 £

Participant 2 9 - Pizza

Figure 4.1a: G r o u p m e a n C R M pattern Advertisement 1 "Pizza'



CO <JO fl*

L A r- o*

m irt u>

Duration of ad

Figure 4.1a (Advertisement 1 - Pizza) was a^ear-refte/advertisement, and was

specifically classified as a "double fear-relief pattern as the following occurred: fear was aroused (the pedestrian steps onto the road and is hit by a car) then was slightly reduced (visuals of surgeon), fear was aroused again (by showing the slow motion re-enactment of the pedestrian's body being hit by the car), and then fear was reduced significantly at the end of the advertisement (by showing the good behaviour of a driver who is not speeding and avoids hitting a pedestrian who steps onto the road).

The group mean graph showed a maximum fear (peak fear) of 178 (of a possible

maximum of 240), and a base relief score of 33 (base relief was the lowest recording of fear after peak fear).

Figure 4.2: Examples of individual participants' C R M responses to Advertisement 2 " P r a m "

Participant 4 - Pram Participant 5- Pram Participant 10 - Pram

250 200

£ f 3

I 16 Jl 46 « » 91 IN 111 136 III 166 III 1M HI 226 241 ZH 271 2*5

Drain) of ad

1 17 33 49 65 «1 97 113 129 145 161 177 193 2W 225 241 257 273 Dtfatkra of ad

l 17 33 49 65 81 97 113 129 I4S 161 177 193 209 225 241 257 273 EXntfen of ad

Participant 11 - Pram Participant 15 - Pram Participant 16 - Pram



17 J3 * « (1 97 IU !» 145 161 177 TO a» 22S 341 2S7 273



Participant 19 - Pram Participant 20 - Pram Participant 21 - Pram


Jl J\



I \J





Participant 23 - Pram Participant 24 - Pram Participant 25 - Pram




n i "U

• " r


Figure 4.2a: G r o u p m e a n C R M pattern for Advertisement 2 " P r a m "



i/t t O r>

Duration of ad

Figure 4.2a (Advertisement 2 - Pram) was also a/ear-re/je/advertisement, and was

specifically described as an "escalating double fear-relief pattern. The following responses occurred: at the beginning of the advertisement, viewers were neutral as the pedestrian crossed at a safe distance in front of the oncoming car; fear was aroused as the car nearly hit the pedestrian during the second driving scenario, but, there was also fear reduction as the car did not hit the pedestrian; fear was aroused again as the car hit the pedestrian; fear was reduced at the end of the advertisement by information about

stopping distances. However, residual fear was left at 91, three times the level of


The group mean graph showed peak fear reached 176 and was reduced to a fear arousal figure of 91.

Figure 4.3: Examples of individual participants' C R M responses to Advertisement 3 "Trike"

Participant 3 - Trike Participant 4 - Trike Participant 5 - Trike




I IOO 50

>•,»«<—» ••«<(

Participant 7 - Trike Participant 10-Trike Participant 15 - Trike


Participant 16 - Trike Participant 17 - Trike Participant 2 2 - Trike

Participant 2 4 - Trike Participant 28 - Trike Participant 29 - Trike

Diratioo of ad

Figure 4.3a: G r o u p m e a n C R M pattern for Advertisement 3 "Trike"

Duration of ad

Figure 4.3 a (Advertisement 3 - Trike) was a fear-only advertisement, that was more

specifically labelled as a "late shock" pattern. At the beginning of the advertisement, participants indicated that they were feeling relaxed (not tense) because of the upbeat jingle and happy children playing on their colourful tricycles. Halfway through the advertisement a child started to ride down a driveway towards the road, on which there was a driver who was speeding. The child rode onto the road and was hit by the car.

Fear was aroused and remained with the participants, thus making the "Trike"

commercial an example of a fear-only pattern.

The group mean graph indicated that peak fear reached 177 and was only slightly reduced to 138 at the end.

Figure 4.4: Examples of individual participants' C R M responses to Advertisement 4 "4WD"

Participant 4 - 4 W D Participant 6 - 4 W D Participant 7 - 4 W D


Participant 9 - 4 W D Participant 1 1 - 4 W D Participant 1 3 - 4 W D



s ioo






-j / — I



Participant 1 5 - 4 W D Participant 1 9 - 4 W D Participant 2 1 - 4 W D

»" — • • • - ! • '

Participant 23 - 4 W D Participant 26 - 4 W D Participant 27 - 4 W D






200-1 » it g







™ — — — '








Figure 4.4a: G r o u p m e a n C R M pattern for Advertisement 4 " 4 W D "

c H


» a t •> s

Ol < * » '

Duration of ad

Figure 4.4a (Advertisement 4 - 4 W D ) was also a fear-only advertisement, and was best described as a "rising fear then shock" pattern. From the initial moments of the

advertisement, viewers started feeling increasingly tense as the driver (a mother) hurried to pick her child up from school, that is, fear was aroused from the beginning of the

advertisement. The driver sped through a roundabout - that was the first rise in the graph (at 24 to 25 seconds) - and then hit and ran over a child who had run onto the road (the shock, at about 32 seconds). Fear never abated after the shock, even rising slightly during the remaining 27 seconds. Residual fear was very high, at 198,

indicating that there was no fear reduction at the end.

4.4.3 Comparison of Static and Dynamic Ratings

Table 4.1 provides comparisons between the static (post-exposure) and continuous

response measurements obtained for the corresponding advertisements. Static ratings of fear and relief, and the relief/fear ratio for each advertisement are shown, as well as CRM mean ratings of peak fear, base relief, fear reduction percentage, and fear level in the final second of the advertisement.

Table 4.1: Comparison of static and dynamic measures of fear and relief

Static measures3 Dynamic measuresb

Relief/ Fear Fear level Fear fear ratio Peak Base reduction final T V ad pattern n Tense Relief (%) n fear relief (%) second 1-Pizza Fear relief 116 1.56 0.51 33 30 178 30 83 33

2-Pram Fear relief 51 1.27 0.39 31 30 176 91 48 91 3-Trike Fear only 52 1.33 0.23 17 30 177 138 22 138 4-4WD Fear only 60 1.84 0.30 16 30 203 198 2 198

a 0 = not at all, 1 = slightly, 2 = quite, 3 = extremely.

b CRM neutral point (and starting point) = 0, and maximum = +/-240.

The static and dynamic results were consistent in regard to the relative rankings for each measure. For example, the advertisement that produced the highest static scores of fear (Advertisement 4 - 4WD) also produced the highest dynamic score of fear. Similarly, lower static relief scores (for the fear-only advertisements) corresponded with dynamic scores of low reduction of fear for the same advertisements. However, while the

rankings were similar, there were several important differences between the measures.

Advertisements 1 and 2 resulted in fairly different "tense" ratings using the static measures (for example 1.56 versus 1.27) compared to the very similar dynamic

measures of peak fear (178 versus 176). Similarly, Advertisements 1 and 3 had

different tense ratings (1.56 versus 1.33), yet very similar peak fear ratings (178 versus 177). Conversely the static relief measure showed less variation between the

advertisements (a range of 0.23 to 0.51, with 0.23 representing less relief than 0.51) compared to the dynamic measure base relief scores (198 to 33, that is, 198 indicated that viewers were still feeling quite tense, whereas a result of 33 meant that viewers were feeling considerably less tense).

The "tense" static measure rating was most likely an average of the fear felt during the advertisement, that is, at the end of the advertisement viewers provided an indicator of their total experience of tenseness with the advertisement. However, the peak fear

dynamic measure was an indication of the peak of fear experienced by viewers during an advertisement and was recorded at the time they were experiencing the tenseness (the dynamic measure "peak fear" score in Table 4.1 represented the average highest

tenseness score for each advertisement group). When calculating fear reduction it is more important to determine the highest rating of fear, versus an average rating of fear for the entire advertisement (that would be more suitable if investigating "levels" of fear).

The relief/fear ratio generated from the static measures did not indicate exactly the same underlying construct as the fear reduction dynamic measure as the two measures were

collected in different ways. The fear reduction dynamic measure was collected using the one instrument (CRM-dial), meaning that the percentage decrease in fear was

measurable. However, the tense and relief static ratings were measured separately, thus attempting to calculate a percentage reduction in fear was not possible. For example, for

Advertisement 1 it w a s not appropriate to deduct a static relief score of 0.51 from a static tense score of 1.56, to then generate a percentage reduction.

While there is almost perfect correlation between the figures when comparing ranks of each measure between the static and dynamic measures, the CRM-dial method was

regarded as superior for obtaining viewer's reactions to fear appeal TV commercials

when conducting research on fear patterns and in particular, the effect of fear reduction.

That is, it is likely that the static and dynamic ratings were measuring similar constructs of fear and relief, but not identically.