Expressive language sampling as a source of outcome measures for treatment studies in fragile X syndrome: feasibility, practice effects, test-retest reliability, and construct validity.

Although the sample sizes were small for this youngest group (18 for conversation and 17 for narration), it is worth noting that none of the 10 t test comparisons of the first and retest scores approached significance, with (uncorrected) p values ranging from .18 to .71 for conversation and .15 to .94 for narration.

also in 8,237 other papers

In a few instances, however, an individual parametric test that was significant at p < .05 failed to reach significance in the nonparametric equivalent and vice versa.

also in 4,113 other papers

There were minimal practice effects, with only a nonsignificant trend (after correction for multiple tests) for an increase in dysfluency from the first to the retest administration in narration for the two older age groups.

also in 5,406 other papers