On the basis of earlier hippotherapy
research,10 –12,16,17,25 a sample size of
20 participants was determined with
a repeated-measures design and a
power-level analysis, calculated at
0.8 power under the assumptions of
mean differences of 2 and a pooled
standard error of 2.45. As a control
for alpha inflation, the power analysis
included a Bonferroni post hoc
pair-wise correction (0.0167).