Limitations of Program Studies
It is important to temper some of the above conclusions by
mentioning the limitations of program studies.
These studies most commonly employed quasi-experimental
designs with only weak evidence to demonstrate causality; sample
sizes were frequently small; parents who agreed to participate
were different from those who did not, thus limiting the ability to
generalize beyond the sample; studies used different measures of
parent-child communication that were difficult to compare; most
studies did not measure the complexity of parent-child communication;
the reliability and validity of these measures were rarely assessed,
but when they were assessed, the results were not particularly
encouraging; few measured program impact upon potentially
important antecedents of sexual risk-taking; and few measured
impact upon actual adolescent sexual behavior.
Such important methodological limitations may have obscured
actual positive program impact.
