The mass center method seems to be more sensitive to the number of trial configuration repetitions than the method of edge-division.
The level of small, local detail between these repetitions is progressively increased.
Repetitions that did not correct a child's error or revise the child's well-formed speech were coded as non-corrective repetitions.
Figure 3 gives the averaged results for the three repetitions of the three-syllable minimal pair.
The frequency of corrective repetitions is questionable and may undermine usefulness of the feedback.
In addition, the frozen noise repetitions are shown to produce unexpected pitch jumps with a potentially useful musical structure.
Minimal confidence intervals might imply a high number of repeat tests, but the findings indicate that there is little benefit in exceeding 4 repetitions.
To control for run-down of the response, the lowest and highest intensity stimuli were bracketed by two repetitions of the intermediate intensity.
