TABLE II.

Statistics for experiments 3 and 4, using a mixed between/within ANOVA. Population consisted of musicians or non-musicians. Masker type was either speech-modulated buzzes or 2-voice maskers. The masker F0 was either fixed at 125 Hz, or roved pseudo-randomly across trials within a block among ten discrete values between 100 and 150 Hz. The ΔF0 was 0, −2, +2, or ±2 semitones in experiment 3; it was 0, −8, +8, or ±8 semitones in experiment 4. Mauchly's tests of sphericity were never significant in experiment 3 [χ2(5) < 5.2, p > 0.389]. In experiment 4, Mauchly's tests of sphericity were not significant [χ2(5) < 5.6, p > 0.344], except for the interaction masker type by ΔF0 [χ2(5) = 18.3, p = 0.003] whose degree of freedom was adjusted with Greenhouse-Geisser correction. *indicates p values less than 0.05.

Factors	Experiment 3	Experiment 4
population	F(1,30) = 6.8	F(1,30) = 11.4
	p = 0.014*	p = 0.002*
masker type	F(1,30) = 1747.5	F(1,30) = 1370.6
	p < 0.001*	p < 0.001*
masker F0	F(1,30) = 5.5	F(1,30) = 5.6
	0.026*	p = 0.025*
ΔF0	F(3,90) = 87.7	F(3,90) = 131.9
	p < 0.001*	p < 0.001*
population × masker type	F(1,30) = 4.2	F(1,30) = 4.6
	p = 0.050*	p = 0.041*
population × masker F0	F(1,30) = 1.4	F(1,30) = 1.3
	p = 0.253	p = 0.266
population × ΔF0	F(3,90) = 0.4	F(3,90) = 0.5
	p = 0.765	p = 0.657
masker type × masker F0	F(1,30) = 0.7	F(1,30) = 5.6
	p = 0.413	p = 0.025*
masker type × ΔF0	F(3,90) = 9.2	F(2.2,65.0) = 38.6
	p < 0.001*	p < 0.001*
masker F0 × ΔF0	F(3,90) = 4.3	F(3,90) = 0.4
	p = 0.007*	p = 0.778
population × masker type × masker F0	F(1,30) = 5.7	F(1,30) = 0.5
	p = 0.024*	p = 0.466
population × masker type × ΔF0	F(3,90) = 0.3	F(2.2,65.0) = 0.6
	p = 0.848	p = 0.578
population × masker F0 × ΔF0	F(3,90) = 1.1	F(3,90) = 2.2
	p = 0.359	p = 0.095
masker type × masker F0 × ΔF0	F(3,90) = 0.9	F(3,90) = 0.9
	p = 0.457	p = 0.461
4-way	F(3,90) = 0.9	F(3,90) = 0.8
	p = 0.462	p = 0.471