Comparisons of the Outcomes between Different Time Points and between Groups
Self-efficacy for exercise (SEE): Results reveal that "SEE" had significantly changed over time in the experimental group (Table 2). Comparisons between different time points of study variables reveal that SEE had improved since the 4th week of the intervention (Table 3). A significant difference of "SEE" between groups was only found at the 4th week (Table 5). The SEE in the experimental group revealed large effect size (ES) at the 4th, 8th, and 12th week (ES= 1.186, 1.320, and 1.198, respectively) whereas medium to large. ES in the control group were found at the 8th and 12t1t week (ES = .679 and
.502, respectively). Medium to large ES (ES= .616) of "SEE", which is significant between groups at the 4th week, was demonstrated. In the experimental group, the largest power estimation for "SEE" was
1.00 at the 8th week and estimated power of "SEE" between groups was .76
Outcome expectations for exercise (OEE): With a ceiling effect of the OEE score, no statistical dif ference was found in both groups and between groups at every time point (Table 2 and 5).
Severity of knee symptoms (M-SEV):
M-SEV score in the experimental group had improved over time (Table 2). However, a significant difference was detected only between baseline and the 8th week (Table 3). Significant difference between groups was found only at the 4th week (Table 4). The experimental group expressed medium to large ES of M-SEV changes at the 4th, 81 and 12th week (ES
= .672, .762, and .687, respectively) whereas ES in the control group were small to medium (ES= .072 to .453). Medium to large ES of M-SEV difference between groups were examined at the 4th and 12th week (ES = .734 and .500, respectively). As same as the estimated power of the M-SEV difference at the 8th week in the experimental group, estimated power of the M-SEV difference between groups was .89.