Overall effects of the experimental manipulation on the frequency of chills show a marginally significant difference in distribution of chills depending on listening conditions ( x 2 = 3.85, p = 0.049, φ = 0.23); this is reflective of a consistent nominal but non-significant pattern found for each piece, with all original conditions resulting in a higher frequency of chills responses across listeners.
To explore whether chills were more moving or intense in the original stimuli, ratings were compared across experiences of chills in original and manipulated conditions; a difference approaching significance between conditions was found for being moved and emotional intensity ratings in chills experiences [ F (2, 48) = 2.67, p = 0.07]; notably, chills were rated as more intense in the original conditions [ F (2, 48) = 10.22, p = 0.0004, partial eta 2 = 0.11], but not more moving.