contamination between groups was
lessened by a distance of over 10 kilometers
between the experimental and the three control
settings. The use of no intervention at all might
have eliminated these limitations. However, studying
in a community setting could not account for the
attention and socialization paid to both groups.
Participants in the control group might perform some
types of exercise provided for all community
members. In addition, there were many women with
knee pain related to osteoarthritis of the knee in the
community to whom the researcher could not reach.
Those women might not be elderly club members,
or might still be full time workers, or be too busy.
Therefore, generalization was limited to the special
group of women in elderly club. The other possible
factor contributing to the control groupûs improve-
ments might include the learning experiences
during the isokinetic strength testing. All partici-
pants had to perform the same tests four times
during the study period. Interaction with the
machine and testing protocol may have been
sufficient to produce some training effects. As van
Barr and colleagues21 indicated that maturation may
occur from several repeated measures.