• No results found

Development of the Dial for Conducting Continuous Response Measurement

CHAPTER 3: DEVELOPMENT AND JUSTIFICATION OF THE PRINCIPAL MEASURES

3.2 Continuous Measurement of Fear and Relief

3.2.3 Development of the Dial for Conducting Continuous Response Measurement

A fear-relief continuous measurement dial (see Figure 3.1), providing 10 readings per second, was constructed to record viewers' reactions during the entire commercial in terms of tenseness felt (fear) and relief experienced. The dial and program were

developed by a computer programmer, David Webster, from the Department of

Psychology, University of Wollongong. Its use is best explained by outlining the procedure.

Figure 3.1 - Dial (CRM)

3.2.4 Procedure for Use of the Dial

Participants were informed that they were about to watch a road safety commercial.

The following statement was then read to them: "Some road safety advertisements are designed to make people feel a bit tense, and we want to map your reactions. You start off with the dial in the middle. As you feel more tense, turn the dial to the right. The more tense you feel, the more you should turn the dial to the right, then back towards the centre if you're feeling less tense. If and when you experience a feeling of relief while watching the advertisement, turn the dial to the left of the centre point. The more relieved you feel, the more you should turn the dial to the left. When you feel neutral, turn the dial back to the centre."

The dial was labelled on the left of the "neutral" point with the numbers 1 to 5, with

"very relieved" being located near 5. Again, to the right of the neutral point were the numbers 1 to 5, with "very tense" located near 5. From the resulting data, the fear pattern of the commercial can be mapped, both within and across participants.

3.2.5 Possible Limitations of the CRM Measure

It is possible that the CRM dial is both a cognitive measure of a participant's reaction to an advertisement and a measure of his or her emotional response to the advertisement,

in that the viewer has to reflect on how he or she feels while watching the advertisement and translate this into moving the dial accordingly towards the labels of relief or

tension. Additionally, the viewer's attention is divided between the advertisement and the task of moving the dial (Hazlett and Hazlett, 1999). In a realistic setting the viewer may not translate emotions into verbal terms, as emotion is not primarily a

language-based experience and cognitive effort is required to put experience into words (Rogers,

1983) as required in this exercise. Rather, the viewer m a y simply "feel" the emotions of fear and relief. Bagozzi, Gopinath and Nyer (1999) noted that research needs to investigate the relationship between these two aspects (cognitive appraisals and physiological processes) of emotion.

Study 3 in this thesis measures fear-pattern responses in two different ways, cognitively and physiologically, in an attempt to explore this largely under-researched issue.

3.3 Electrodermal Responses (EDR)

A psychophysiological measure is included in Study 3. Electrodermal response (EDR) measurement, specifically skin conductance level (SCL), is the alternative method of determining differences in levels and patterns of emotion.

Traditionally, skin conductance recording was referred to as galvanic skin response (GSR). Skin conductance recording detects involuntary changes in the electrical

activity of the skin that is measured by the change in skin resistance (Krugman, 1965) that results from fluctuations in electrical conductivity of the skin from a base level.

Electrical activity is produced by the activity of the sweat glands that are widely distributed throughout the skin and are regulated by the autonomic nervous subsystems (LaBarbera and Tucciarone, 1995).

Skin conductance response (SCR) refers to phasic data (short-term, that is, only a few seconds), with phasic fluctuations referring to the fluctuations in electrical activity of the skin from a base level. Skin conductance level (SCL) refers to tonic data (long-term averages of reactions to stimuli lasting a minimum of 30 seconds) (Andreassi, 2000;

Hopkins and Fletcher, 1994). A s 30 to 60 second television commercials are the stimuli used for the studies in this thesis, SCL will be reported.

Skin conductance recording is a sensitive indicator of the internal psychological activity of participants and is one of the most frequently applied activation indices (Kroeber-Riel, 1979). Kroeber-Riel's (1979) research used 'psychobiologicaP approaches

(including skin conductance) to measure activation (also termed arousal or alertness), explaining that the effect of human behaviour, such as activation, is physically

demonstrated in the form of physiological responses because of the central nervous system. In this paper, Kroeber-Riel (1979, p.248), after conducting an empirical investigation, concluded that "psychobiological measurements are in many cases more valid than other methods, and are thus indispensable for scientific research". Kroeber-Riel (1979, p.243) also claimed that "extensive tests of the validity and reliability of EDR have yielded satisfactory results" and referred interested readers to the works of Lykken and Venables (1971), Burstein, Frenz, Bergeron, and Epstein (1965) and

McCurdy(1950).

Other studies that support the validity and reliability of skin conductance recording include Kohan's (1968) study and Greenwald, Cook and Lang's (1989) study.

Furthermore, there have been other marketing studies that have successfully used skin conductance measures, these include studies conducted by Mewborn and Rogers (1979), Hopkins and Fletcher (1994), La Barbera and Tucciarone (1995), Lang, Zhou,

Schawrtz, Bolls and Potter (2000) and Kilbourne, Painton and Ridley (1985. Each of these studies will be briefly reviewed in the following two sections.

K o h a n (1968) investigated twenty participants' psychophysiological responses (via EDR) to three advertisements. The study found that the combined (averaged) EDR scores of the participants showed peaks where "it was expected that attention would rise" (p.47), for example, skin conductance increased significantly in correspondence with scenes where lights were suddenly turned on or when dialogue suddenly started.

Greenwald, Cook and Lang (1989) compared 48 participants' psychophysiological responses (including skin conductance, facial electromyographic and heart-rate measures) to their verbal affective judgements of 21 pictorial stimuli (shown using photographic slides). The findings of the study, in relation to skin conductance, were that larger changes in skin conductance were directly reacted to increased arousal ratings, particularly for slides showing erotic and mutilation images.

Kilbourne, Painton and Ridley (1985) successfully used skin conductance measures to determine psychophysiological reactions to different versions of subliminally embedded sexual images in advertising. Skin conductance measurements were obtained from 36 participants who viewed two versions of two advertisements, one containing a sexual embed and one without. Skin conductance scores were higher for the embed versions of both ads.

Hopkins and Fletcher (1994) developed an ad-testing measure that used skin conductance response (SCR) to predict the effectiveness of sales messages for commercial products and services. LaBarbera and Tucciarone (1995) conducted a

similar study to Hopkins and Fletcher (1994), using two decades of a USA Fortune 500 packaged goods company's general and direct response advertising, sales data and

galvanic skin responses scores. LaBarbera and Tucciarone (1995, p.33) conclude that

marketing practitioners should reconsider "the potential value of GSR methodology in evaluating the motivational power of marketing communications".

Lang, Zhou, Schwartz, Bolls and Potter (2000) used skin conductance responses to measure arousal of participants when viewing television messages. Their study focused upon the effect of low or high rates of camera edits within the same visual scene on viewers' arousal and memory. Faster editing resulted in increased skin conductance, self-reported arousal and an increase in memory.

For the interested reader, Bagozzi (1991) provided a comprehensive summary of the role of psychophysiology in consumer research. A more general discussion of

psychophysiological methods was presented by Andreassi (2000) in his book Psychophysiology: Human Behaviour and Physiological Response.

A major reason for the frequency of its use is that skin conductance measures are easier to obtain than other psychophysiological measures, and have high correlation with other psychophysiological measures. Mewborn and Rogers' (1979) study is an example of

one of the few experiments that have used both self-reports of fear and physiological measures of emotion when investigating the effect of fear appeals. They found high correlation (r = 0.7) between heart rate (not used in this study) and skin conductance (used in this study). The justification given for using autonomic measures of fear was that an essential part of the definition of emotion is physiological measures. For example, Shaver (1987, p. 1076) described a person's state when they feel fearful, as

being an arousal of the autonomic nervous system in preparation for fright, "The person feels jittery and jumpy, perspires, trembles and looks quickly around".

LaBarbera and Tucciarone (1995, p.48) noted that "because consumers have little voluntary control over their autonomic nervous systems, changes in bodily functions can be used by researchers to indicate the actual, unbiased amount of activation or arousal resulting from marketing stimuli". It is important to note that EDR measures general arousal, that is not necessarily "fear" arousal. For example, SCL detects changes in electrical activity of the skin caused by fear and excitement.

Emotions have valence (positive and negative) and intensity (low and high arousal). A major limitation of EDR is the inability to determine the valence component. But in the instance of fear appeals it is highly likely that the arousal is mainly caused by fear arousal (general arousal accompanied by negative affect).

Study 3 again tests the fear-as-acquired-drive model (Hovland et al, 1953) by using continuous EDR (SCL) recording in addition to continuous response measurement dial recording.

3.3.1 Description of the EDR Recording Apparatus

The EDR recording device was UFI's Model 2701-SC Simple Scope (August 2001)

(see Figure 3.2). Participants consented to have finger-cuff electrodes fitted to their middle and index fingers. They were then instructed to sit comfortably and quietly while they watched a black television screen with a white cross (+) on it for 3 minutes. The advertisement immediately followed this 3-minute baseline period.

Figure 3.2: UFI's M o d e l 2701-SC simple scope ( E D R recording instrument)

Similar to the dial, the E D R recording equipment takes 10 recordings per second.

Therefore, for a 60-second advertisement there are 600 data points and for a 30-second advertisement there are 300 data points. The EDR data was calculated as follows: from the initial data-point numbers that the Simple Scope provides (example of a data-point, 459) this number is multiplied by 100 and then divided by 4095 (example resulting in a number of 11.21, 2.d.p.) to produce the converted SCL (in ohms). The digital output range for the data recording was 0-4095 that equates to 4096 possible points in a 0-20 ohm range.

An average converted baseline figure was calculated from the middle 2 minutes of the 3 minute baseline period (example, 11.08) as this allowed participants time to settle down and obtain a normal reading prior to the start of the advertisement. This average

baseline figure was then deducted from all of the converted SCL scores to produce adjusted SCL scores (11.21 - 11.08 = 0.13). This means that once the average baseline was deducted from the data obtained during the advertisement the difference in scores should have reflected a result of the psychophysiological reactions to stimuli within the

advertisement. A n example of this conversion process is presented in Figure 3.3. For

this individual recording of a participant's reaction to an advertisement, the first graph (upper left) shows the original raw data collected for the duration of the advertisement.

The second graph (upper right) shows the same responses, but they are now expressed in ohms (SCL converted). The third graph (lower left) shows the 2-minute baseline output (also in SCL converted) and the fourth graph (lower right) shows the SCL

adjusted score that was derived from subtracting the average 2-minute baseline score (11.08) for this particular participant from their SCL converted data (in the second graph).

Figure 3.3: Examples of how adjusted SCL scores are derived

Original data - Participant No.56 S C L converted - Participant N o . 5 6

3 O

Pi

1000 800 600 400 200

25 20

t! 10 o

>

c i

O Duration of ad

2-minute converted SCL baseline • Participant No.56

25.00 20.00 15.00 10.00 5.00 0.00

Duration of ad

S C L minus baseline - Participant N o . 56

Duration ofbaseline period -2.00

Duration of ad