“…Furthermore, given the low number of participants and the strength of the repeated-measures ANCOVAs, we felt it would be better to proceed with the parametric analyses without eliminating those few, more extreme values. Parametric analyses were applied to the STAI, even though the data came from a Likert scale, given the simulations documenting the strength of this type of analysis when used with ordinal data (Velleman and Wilkinson, 1993;Davidson and Sharma, 1994;Stiger et al, 1998;Carifio and Perla, 2008;Norman, 2010). The homogeneity postulates for the regression slopes were respected [F(1.28) = 1.03 for heart rate, 1.23 for the STAI, all ns].…”