Most evaluations of sexual abuse prevention programs employ composite
knowledge scores that may mask whether individual skills are different
ially understood, The few studies examining separate skills have provi
ded no statistical justification for doing so. This study attempted to
validate empirically distinct prevention skills and whether children
comprehend these skills in the order taught by prevention programs: RE
COGNITION (of good and bad touch), SAY-NO (refuse the perpetrator), GO
(leave the situation), TELL-WHO (find an adult), and TELL-WHAT (accur
ately disclose the abuse). Subjects were 117 preschoolers who were ind
ividually administered the ''What If Situations Test'' (WIST). Analyse
s confirmed that the WIST was composed of six skills. The hypothesized
sequence of skills produced a high coefficient of reproducibility, bu
t an even higher coefficient was obtained when three common sequence r
eversals were added to the existing logical sequences. We discuss expl
anations for these cognitive sequence reversals and applications of th
is information for teaching preschoolers prevention concepts.