A researcher wants to prove that structural family therapy is the
most effective modality. She conducted a study a year ago using
a signifi cance level of .05. Several colleagues felt her signifi -
cance level needed to come down. She thus ran the same basic
experiment again with new people using a signifi cance level of
.01. Her chances of making a Type I error or so-called alpha
error went down. Now assume you compare her new research
with her old research. What could you say about the possibility
that her results will indicate that structural family therapy was
not signifi cantly different when in reality it truly is signifi cant?
