The design of the current study included several features to minimize bias, thus enhancing the internal validity of results obtained and strengthening the conclusions that can be drawn from this study. Participants were allocated to 1 of 2 treatment groups using a predetermined random and concealed allocation process. An exercise approach of proven efficacy was implemented as the control treatment.22,23 Passive mobilization therapy was isolated to the shoulder region joints in order to evaluate the contribution of this specific component of the common group of treatment modalities used by physical therapists to treat people with shoulder pain. Participants in both groups were provided with the same mean number of treatments and thus the same amount of participant-therapist contact time. Short-, medium-, and longer-term reassessments were conducted. The number of participants unavailable for reassessment was low at all follow-up periods, thus preserving the successful randomization achieved at the commencement of the clinical trial. The sample size was large enough to provide adequate power to detect significant differences, and responsive, reliable outcome measures were used. With the exception of duration of symptoms, which was taken into account in the analysis, baseline demographics were similar at the commencement of the trial, and baseline outcome measurements were well matched. Finally, an intention-to-treat analysis was implemented.