Abortion perils debated

As Stephen Genuis observes, “it is sometimes difficult to objectively determine what is factual and credible scientific information and what represents sexual and philosophical ideology.” Researcher bias clearly can affect the research process. Nowhere is this more obvious than in research on abortion. David Reardon has quite explicitly stated his intentions to use data such as those he reported in CMAJ1 to affect abortion-related legislation, bring litigation against physicians who perform abortions and reduce women's access to abortion.2

It is an error, however, to assume that because researcher neutrality is difficult to achieve, what passes for “evidence” on both sides of politically charged issues is likely to be equally valid and deserving of equal airing. Not all research is biased. It is possible to distinguish good science from bad. Good science is based on established scientific methods, eliminates confounders and uses appropriate control or comparison groups. The study by Reardon and his associates1 is not good science.3 It inappropriately used women who carried a (likely wanted, planned) pregnancy to term as a comparison group for women who aborted a (likely unwanted, unplanned) pregnancy. More appropriate comparison groups include women who carried a pregnancy to term and gave the child up for adoption, and women who wanted an abortion but who were denied one or did not obtain one because of external pressures or guilt, as Aaron Keshen points out in his letter.

Reardon and associates also failed to control adequately for demographic, social and psychological differences that likely existed at the time of the pregnancy between women who subsequently aborted versus those who carried their pregnancies to term. The inference that the abortion procedure itself caused postpregnancy differences observed between these 2 groups is faulty scientific reasoning and misleading. The studies referred to by Annie Banno, all of which were conducted by Reardon, are plagued by similar methodological problems.

Brenda Major Department of Psychology University of Santa Barbara Santa Barbara, Calif.