3 - MEASURING THE EVALUATION OF MUSICAL STIMULI

[ 3 ]
MEASURING THE EVALUATION OF MUSICAL STIMULI

3.1 The convergent validity of the semantic differential technique.

Methodologically, all the empirical studies carried out on music perception are based on the use of rating scales, typically presented in the context of the semantic differential technique.
This technique uses many rating scales to measure affective meaning (Osgood, Suci, Tannenbaum, 1957).
The technique assumes that a stimulus being rated will elicit affective associations detectable on independent dimensions, and that the aggregate scores deriving from separate bipolar scales are surface variables of a deeper meaning dimension.
It is interesting to point out that, while the semantic differential technique has been applied to the study of many different domains, one of its authors (Osgood, 1980) noted a link between his early studies of color-music synesthesia and the development of the semantic differential.
Tannembaum (1956), co-author of the semantic differential technique, applied this measuring instrument in an early study of the effects of music on television and drama.
Therefore, it is plausible that at a subsequent time more researchers converged on the use of the rating scales technique in the context of music.

3.2 Evidence for a bi-dimensional structure

Bruner (1990) reviewed all the literature dealing with mood being induced by musical stimuli. This review reported a small number of adjectives, all produced by participants, referring to singular characteristics of music able to induce mood. The most frequent occurrences were: exciting/intense, tranquil/peaceful, solemn/dignified, joyful/happy, sad/tragic.
Baumgartner (1992) also came up with a very similar set of terms, which were used to construct rating scales in an empiric research on the role of music in evoking emotions.
These two findings are strongly correlated, and, entered in multi-dimensional analysis, prove that a two-dimensional solution is feasible and interpretable.
The two dimensions are Pleasure (deriving from the combination of the opposite pairs playful/tragic, happy/sad, joyful/solemn) and Arousal (deriving from the combination of the opposite pairs exciting/tranquil, interesting/boring, etc.).
These axes seem to correspond exactly to the first two factors in Mehrabian and Russell’s (1974) three-dimensional model of emotions, also known as PAD.
Originally developed within the context of environmental psychology, Mehrabian and Russell’s three-dimensional model of emotional responses has been successfully used for research on marketing, advertising videos (consisting of both images and music) and consumer behavior (North and Hargreaves, 1996 and 1998).
As a matter of fact a film or a piece of music are stimuli possibly not less complex than an environment, and the movie viewing experience is probably as cognitively demanding as the exposure to a environment. Within the PAD model, emotional reactions to such complex stimuli can be characterized by three evaluative dimensions:

1) Pleasure

2) Arousal

3) Dominance.

These dimensions are conceptualized to be relatively independent from one another.
The results of a very recent study in the field of cognitive neuropsychology also seem to fit well with a bi-dimensional model.
Schmidt and Trainor (2001) found that the pattern of asymmetrical frontal EEG activity distinguished affective valence of musical stimuli. Participants exhibited greater relative left frontal EEG activity to joyful/happy musical excerpts and greater relative right frontal EEG activity to sad/tragic musical excerpts. They also found that participants exhibited significantly greater overall activity in the frontal region of the brain as the musical stimuli varied from calm to intense.
These two response patterns, left vs. right frontal activity corresponding to affective valence and magnitude of overall frontal activity corresponding to the intensity of the stimuli, seem to be connected to the Pleasure and Arousal bi-dimensional structure of music appraisal.
After all, according to several musicologists, meaning within music lies in the emotions it is able to arise (Meyer, 1956; Cooke, 1959).

3.3 Measuring responses: Our rating scales

We created a response sheet where participants could fill in their ratings for the stimuli (music only, film only, or a combination of the two, depending on the phase of the research they took part to).
Our questionnaire was composed of semantic differential scales scored on a 9-point (–4 to +4) interval. Items used the opposite adjective pairs gathered from the previously mentioned literature:

1) Sad/Happy
2) Calm/Exciting
3) Playful/Solemn
4) Interesting/Boring
5) Tragic/Joyful
6) Relaxing/Stimulating

The Pleasure dimension was represented by odd items (1, 3, and 5) while even items (2, 4, and 6) represented the Arousal dimension. Item scores were then averaged within dimension to form indices.

Participants were also asked to self-report their musical skill:

How do you consider your musical knowledge and abilities?
[none] [basic] [experienced] [expert]

[ next ]-[ index ]