The effect of repeated attempts and test-retest reliability in children’s singing accuracy

The purpose of this study was to examine the effect of repeated attempts at singing accuracy tasks. Test-retest reliability of singing accuracy was examined in a second administration of the test. A secondary purpose was to analyze individual variability in children’s singing accuracy. Test stimuli were designed using five attempts each at a single pitch, interval, and four-note pattern, and song singing. Children aged 6–11 were given the one-on-one singing accuracy test, and an identical form of the test was administered again within 1–6 weeks. Pitch matching items were scored by measuring the deviation in Hertz from the stimuli. The song singing item was scored by singing teachers using an 8-point scale with acceptable inter-rater reliability (r =.86). Participants’ individual best attempt out of five was equally distributed, with overall performance increasing across subsequent attempts measured in signed cent deviation. A repeated-measures ANOVA with the task type (single, interval, and pattern) and attempt (1, 2, 3, 4, and 5) as the within-subjects variables indicated no main effect for task type (p =.129), but a significant main effect for attempt (p <.001, (Formula presented.) =.087). Test-retest reliability was considered high (r =.69).