To decrease the risk of potential bias the therapist performing the treatment and the investigator taking the measurements remained blind to each other's activities throughout the study data collection period and the subjects were not informed of the true purpose of the study at the time of their involvement-that is, they were blind to which treatment was intended to have an effect. As the sample size was small, the estimate of the measurement error (ie, the SEM calculated from the reliability study) was also calculated. The SEM for both the posterior talar glide and dorsiflexion measures was small, and the differences from preapplication to postapplication of the treatment techniques were substantially larger than the SEM, providing some indication that the statistically significant differences were also larger than the error of the measurement.