4 - APPRAISAL OF STIMULI

4.1 Preliminary appraisal of music

The aim of this first experiment was to obtain a measure of the Pleasure and Arousal dimensions for the musical stimuli alone.
In the second and third part of the research we needed a set of musical tracks whose elements varied on the Pleasure and Arousal axes (i.e. different means), but each stimulus had to receive consistent ratings between subjects (i.e. a low variance).
We also suspected that musically trained participants, when presented specific stimuli, could give ratings that were different from the ones produced by the non-trained participants.
For this reason, musical expertise was used as a criterion for the final selection of stimuli.

4.2 Collecting and editing the auditory material

A first bulk of music was gathered in a way such as to cover a wide number of different genres (classical, jazz, pop, etc).
Another important criterion was that the participants of the experiments should not have heard the stimuli before. Therefore all music was selected from the repertoires of foreign artists not known in the country and not covered by the national media. Some other we collected directly from still unreleased material by local professional musicians.
Each of the 22 tracks has been edited and trimmed down to a 30-second segment that featured a key part of the song (i.e. the chorus and not the introduction or the coda).

4.3 Participants and experimental design

Participants (f=87, m=20) were students of University of Padova; they sat in a classroom while music was diffused through the speaker system. There was no perceivable difference in loudness from one row of seats to another, as several speakers were attached to each side of the room.
Participants were asked to listen to the music for its entire length, and then to start filling in the form with the scales of the semantic differential.
The order of the tracks was randomized and there was a 45-second interval after each presentation. Participants rated each musical excerpt using the 9-point semantic differential scales on the response sheet.
After the presentation, we also asked if anyone was familiar with some of the soundtracks. Nobody had ever heard any of the music before.

Before calculating Pleasure and Arousal ratings, we recoded the collected data. The polarities of three of the scales were inverted, according to the rule that the highest rating was associated to the maximum score on the Pleasure or Arousal dimension.
Each 9-point rating scale originally ranged from –4 to +4.
We shifted the range to 0 to 8 by adding 4 to each score.
The result of this operation was a translation of all the ratings to positive values, ranging from minimum (0) to maximum (8).
At this point we calculated the Pleasure and the Arousal indices.
Scores for items sad/happy, solemn/playful, and tragic/joyful were averaged to form the Pleasure index. Similarly, scores for items calm/exciting, boring/interesting, and relaxing/stimulating were averaged to form the Arousal index.

MUSIC	Pleasure	Arousal
01. lydiom	3.49 (1.24)	2.28 (1.33)
02. 5guys	6.46 (1.11)	5.69 (2.13)
03. dhall	4.66(.96)	3.99(1.03)
04. sunshine	5.65(1.27)	5.04(1.20)
05. fefa	4.10(1.51)	3.40(2.05)
06. chili	5.84(1.48)	5.54(1.28)
07. Armani	1.13(1.14)	3.40(2.87)
08. haze	5.25(1.81)	3.35(2.14)
09. rocksolo	4.85(1.91)	5.60(1.36)
10. fading	3.07(1.46)	2.20(1.94)
11. gangrap	6.12(1.45)	6.16(1.63)
12. justthe	3.57(.89)	2.68(1.00)
13. throw	5.97(1.44)	6.48(1.52)
14. divpro	2.20(1.68)	3.91(2.18)
15. crazy	4.38(.69)	3.16(1.12)
16. Duncan	3.73(.71)	2.80(1.16)
17. grace	3.11(1.31)	1.87(1.28)
18. never	6.10(1.01)	5.45(1.72)
19. howlong	6.30(1.19)	5.94(1.83)
20. duninstr	2.53(1.18)	2.66(2.13)
21. placidk	4.69(1.60)	4.91(2.07)
22. angela	5.80(.92)	4.67(1.92)

Table 4.4.1: Pleasure and Arousal scores (mean and variance) for music appraisal (n=107).

We compared the male (n=20) and female (n=87) groups using independent samples t-tests.

Group means for track 22 showed a significant difference in Pleasure (t = 2.021, df=105, p<.05).

Group means for tracks 3, 15, 18 and 22 showed a significant difference in Arousal (t=2.496; t=2.165, t=2.054, t=4.793, df=105, p<0.05).

These pieces of music have not been used further on in the research.

Musical expertise or training, as self-reported by participants in the final part the experimental session, was also inspected. Most participants (n=65) reported to have no musical training at all, some reported a basic (n=17) or average (n=19) musical expertise. Only a very limited number (n=6) rated themselves as experts.

In order to test whether musical ability influenced appraisal of soundtracks we needed to compare different expertise groups with a consistent population.To accomplish this, the expertise levels 0 and 1 were considered as one group, and the expertise levels 2 and 3 were aggregated in a second group. These two groups have then been compared with independent samples t-tests.

Group means for tracks 7, 10, 12 and 20 showed a significant difference in Pleasure (t=2.626, t=2.458, t=2.220, t=2.190, df=105, p<.05).

Group means for tracks 12 and 20 showed a significant difference in Arousal (t=2.603; t=2.237, df=105, p<0.05).

This material has been excluded from further use in the research.

4.5 Appraising film excerpts

The aim of this experiment was to obtain a measure of the Pleasure and Arousal dimensions for the visual stimuli alone.We were looking for a small number of film excerpts whose elements varied on the Pleasure and Arousal dimensions (different means), but each scene had to get consistent ratings between subjects (a low variance).The structure of the experiment was basically the same as in the previous one, except it was carried on visual material.

4.6 Notes on film material

The film excerpts were taken from the movie “Ascenseur pour l’èchafaud” (Elevator to the Gallows) by director Louis Malle (1958).We chose this particular film for a number of reasons.

First of all, the vast majority of the people participating to the research were between 19 and 24 years old, therefore they had probably not seen the movie before. Thus, they did not know anything about the story, or about the position of a certain scene in the plot.This issue was of great importance to us, since we needed participants to focus only on the visual and auditory material they were presented during the experiment.

A second reason for using Malle’s film was that, although we can say it belongs to the noir genre, it also features heterogeneous material and many open-ended sequences. In fact, the selected scenes were very different from each other and ranged from action sequences (a car chase, a shooting) to more soothing sequences (a young couple talking over the phone, a woman buying flowers). Each scene was edited in order to last for about 30 seconds (the same duration of the music excerpts).

4.7 Participants and experimental design

All participants (f=73, m=19) were students of University of Padova, they all sat in the same room while the film was projected on a large screen. Participants were asked to watch the scenes for all their length, after that they started to fill in the form with their evaluation.The order of the stimuli was randomized and there was a 45-second interval after each presentation. After the presentation participants explicitly said they had never seen the film before, also most of them did not even notice that all the scenes had been taken from the same movie.

4.8 Results

Before calculating Pleasure and Arousal ratings, we recoded all data following the same procedure as in the music appraisal. This produced all positive scores for both the Arousal and the Pleasure factor, ranging from a minimum value of 0 to a maximum of 8. Scores for items sad/happy, solemn/playful, tragic/joyful were then averaged to form the Pleasure index, while scores for items calm/exciting, boring/interesting, relaxing/stimulating were averaged to form the Arousal index.

SCENE	Pleasure	Arousal
01. phone	2.63(1.29)	3.10(1.33)
02. climb	3.95(.87)	4.88(2.13)
03. couple	5.00(1.32)	3.42(1.03)
04. walking	4.07(.73)	4.13(1.20)
05. cars	3.46(.82)	5.63(1.59)
06. cheers	4.70(1.07)	5.63(1.02)
07. elevator	2.90(1.32)	5.97(1.86)
08. shoot	1.75(1.33)	6.17(1.71)
09. crying	2.24(1.60)	4.05(2.8)
10. press	4.05(.92)	3.22(2.05)
11. flowers	4.15(1.26)	2.98(1.32)
12. croissant	4.36(.92)	3.61(2.06)

Table 4.8.1 Pleasure and Arousal scores (mean and variance) for film appraisal (n=92).

Chart 4.8.1: Position of music and film stimuli on the Pleasure and Arousal axes (n=92).

Male and female participants have been compared by means of independent samples t-tests.

Group means for scenes 2 and 10 showed a significant difference in Pleasure (t=2.702; t=2.053, df=90 p<0.05).

Group means for scene 10 showed a significant difference in Arousal (t=2.200, df=90, p<0.05).

This material has been excluded from further steps of the research.