Two participants from the experimental group
withdrew from the study due to work commitments
in another province. This study analyzed the data as
an intention-to-treat analysis of treatment responses
and included the two participants who withdrew
from the study. Participants with a posttreatment 9
Q score of below 7 were classed as responders. Data
were analyzed with STATA Version 8.2. Descriptive
statistics were used for all variables. Relative
risk was calculated using conditional logistic
regression to present a ratio of the proportion of
cases having a positive outcome (score on of below
7 the 9 Q) in the two groups. The variables used for conditional logistic regression analysis were the
same variables that were used for matching sample
assignment between the experimental group and the
control group, including age, gender, and score on
the 9 Q. Descriptive information about the patients'
perception of change from the group tape recordings
and patients' journal diaries was analyzed
based on patients' understanding of the Universal
Natural Laws (craving) using content analysis.