Baseline Characteristics of Participants and Dropouts
Preliminary analyses (t-tests or chi-square tests) comparing participants in the Tai Chi group (n = 49) and the control group (n = 45) indicated that the two conditions did not differ significantly (p.09) at baseline on any of the demographic measures involving age, gender, income, and education. Additionally, there were no significant differences (p.13) by group on any of the HRQL measures at baseline. These results indicated no need for adjusting any demographic and/or baseline measures in the subsequent main analyses.
Canadian pharmacy no prescription viagra
Furthermore, two (completion status: drop vs. remain) by 2 (condition) analysis of variance (ANOVAs) on age, gender, income, and education revealed that there were no significant differences between those who dropped (n = 22) and those who remained (n = 72), nor were there any interactions between completion status and conditions (p=.25).

Change Over Time in HRQL Measures
The outcomes of the study were tested by comparing participants receiving Tai Chi with those on the waiting-list condition on change in HRQL measures from pretest to posttest. Scores from the various dimensions of SF-20 were entered into univariate repeated measures ANOVAs to test for differences in follow-up HRQL scores between individuals in the Tai Chi and control groups. Because preliminary analyses indicated that the participants who dropped out were similar to those who adhered to intervention on the demographic and outcome variables, we analyzed the change in HRQL measures with participants who completed the study (n = 72).

The results of ANOVAs indicated that, over the course of the study, individuals receiving Tai Chi had better outcomes had those on the waiting list on all the SF-20 subscales (significant Group by Time interaction): F(2,69) = 4.234, P < .02, for physical functioning; F(2,69) = 2.953, P < .05 for social functioning; F(2,69) = 5.369, P < .007, for role functioning; F(2,69) = 4.630, P < .05 for bodily pain; F(2,69) = 2.763, P < .01, for mental health; and F(2,69) = 6.653, P < .002, for health perceptions. Higher scores indicate an increase/improvement in HRQL measures over time. Examination of means for each condition showed that, in general, participants in the Tai Chi group reported a significant improvement (i.e., higher HRQL mean scores) over time in all domains of SF-20 measures. Identical results were obtained when these dropout individuals were also included in the analyses.

Using the five-point difference as the criterion (Ware et al., 1993), only changes (i.e., from baseline to posttest) in Physical Functioning and Health Perceptions reached clinical significance.
The effect sizes were: .69 (physical functioning), .92 for social functioning, .78 for role functioning, .5 for bodily pain, .56 for mental health, and .54 for change in health perceptions. Based on our criteria on effect size, these values were judged as meaningful treatment effects.