Results

The timbre identification task based on pure tones was more difficult than the timbre identification task based on piano tones in Experiment 1. Four participants were excluded from the analyses because of low accuracy (mean percentage of correct responses: 46.4 %).For the remaining participants, accuracy was high overall (mean percentages of correct responses: 89.7 ± 13.2 %). Percentages of correct responses (Table 2) and correct response times (Figure 5) were analyzed with two 2x2 ANOVAs with Tonal Relatedness (Related / Less-related) and Target Timbre (Timbre A / Timbre B) as within-participant factors.

For percentages of correct responses, only the main effect of Target Timbre was significant (F(1,33) = 4.73; MSE = 84.67; p < .05): targets played by Timbre B were judged more accurately than targets played by Timbre A.

For correct response times, the main effect of Tonal Relatedness was significant (F(1,33) = 4.77; MSE = 3360.34; p < .05): related targets were judged significantly faster than less-related targets. The main effect of Target Timbre was also significant (F(1,33) = 92.83; MSE = 10652.97; p < .001): response times were faster for Timbre B than Timbre A. The interaction between Tonal Relatedness and Target Timbre was not significant (p = .24).

Participants were separated into 12 moderately experienced participants (instrumental instruction = 3 to 15 years, mean = 7.0  3.6 years, median = 6 years) and 22 less-experienced participants (instrumental instruction = 0 to 2 years, mean = 0.23  0.61, median = 0 year). For correct responses and response times, ANOVAs with Musical Expertise added as a between-participants factor did not reveal any additional significant effects.

Figure 5. Correct response times in Experiment 2 presented as a function of tonal relatedness (related/less-related) and target timbre (timbre A, timbre B). Error bars represent between-participants standard errors.