3. Research design and methods
A group randomization scheme was used in the current study.
That is, all individuals in the SB group (families having a child
with and without special needs) was randomly assigned to
receive the home fire safety DVD intervention via group
setting, while the individuals in the WR group was assigned to
receive the home fire safety DVD intervention in the
traditional face-to-face manner in one to two parent–child
cohorts. The DVD examines kitchen, bathroom, electrical burn
prevention, smoke alarm use, home exit plans and practice,
and burn first aid and takes five to six minutes to watch.
Outcomes were measured at baseline and post-intervention to
measure recall and retention of information provided. We
then tested for differences between groups and over time
using Repeated Measures ANOVA techniques; subsequent to a
descriptive analysis. In addition, we evaluated whether having
a child with special needs impacted outcomes. From the
anticipated sample size (n = 80) the study had 82% power to
detect a 10% main effect of each treatment using a RM ANOVA
model. In addition, this study had 80% power to detect a smallto-
medium increase in proportions over time, if/when
samples were pooled. The analytic sample size was n = 87,
so the study was adequately powered.