Second, we were able to use our qualitative data to strengthen our
quantitative design. From our observations and interviews, we were in a
better situation to select comparison groups for each garden. We
understood which social processes and demographic variables were
important (and which were not) when we were seeking specific comparison
groups. For example, we leamed during interviews that gardeners used the
program to transmit their cultural heritage to their young, and that this was
particularly important to the Hmong clientele. We then knew that we had to
match our comparison group participants by ethnicity.