Guest Columnist Bad science: holes in the study claiming 'most' women don't regret their abortion

Jul 23, 2015

In the midst of the recent postings of controversial videos discussing the use of fetal remains for medical research, I cannot get away from thinking about abortion. To me, this is a relief. I never want to be able to shake the reality of abortion from my thoughts and prayers.

Ever read something that made you question the validity of your beliefs? For the pro-life person, reading "Decision Rightness and Emotional Responses to Abortion in the United States: A Longitudinal Study", by Rocca et al (2015), might be one of those times.

Published this month, the article has received some significant press based on its findings. It concludes that the 667 post-abortive women studied "experienced decreasing emotional intensity over time, and the overwhelming majority of women felt that termination was the right decision for them over three years." This single study has been cited in numerous articles, hoping to dispel the notion that women suffer from abortion. The titles say it all: "Overwhelming majority of U.S. Women don't regret abortion", "Like 95% of women, I don't regret my abortion – it was the happiest day of my life", "Less than 1% of women regret their decision to have an abortion", "Women rarely regret their abortions. Why don't we believe them?", "Hardly Any Women Regret Having an Abortion, a new Study Finds"… and the list goes on and on.

If these conclusions are true, the Pro-life movement has lost the wind beneath its wings. How can you argue that you are pro-woman when women are reportedly fine with the procedure you do not want them to have? After all, 95% of women is a lot of women.

I have four concerns when I read the Rocca et al study and the conclusions drawn from media sources as a result.

1. 667 women who have had an abortion may not be generalizable to all American women who have had an abortion.

Why? In 2008, the year that the women were first studied, 825,564 women aborted children, according the Center for Disease Control. The findings of this sample only represent .00081 of the women who received an abortion that year. A good scientist knows that this is not necessarily problematic, insofar as the sample is a good randomized sample. At the same time, an honest scientist must admit that there are limitations to the conclusions drawn in any study, but especially a study in which so little information is given regarding the sampling process. We do not know what we do not know. Not knowing the sampling method, it would be a grave overstatement to propose that the attitudes of 95% of women in a study that captures only .00081 of the represented group of women can be generalized to all women who have gotten an abortion. That is just bad science. If only media sources considered that before grossly overstating the findings of the Rocca et al study.

2. When a study is funded by organizations that openly fund numerous pro-choice organizations, the assertion of uniquely being unbiased is less believable.

The study was funded by the Wallace Alexander Gerbode Foundation, a research grant from an anonymous foundation, and an institutional grant from the David and Lucile Packard Foundation. The first foundation openly offers grants to a multitude of pro-abortion organizations, including Catholics for Choice, Medical Student for Choice, NARAL Pro-Choice America, Nursing Students for Choice, etc. The second foundation listed has noteworthy motivations for this research as well. The David and Lucile Packard Foundation openly describes its three fundamental beliefs as "1. Rapid population growth puts pressure on the environment and hinders development; 2. Access to family planning enhances health and quality of life for women and their families; and 3. A woman's right to safe abortion is an essential component of reproductive health services and should be protected." It has a "longstanding commitment to comprehensive sexuality education, voluntary family planning and safe abortion." To be clear, this does not invalidate the study's findings. But it may color the   conclusions drawn from the findings, a point worth considering before believing that the results are purely scientific and untainted by biases. Again, popular media sources are naive, at best, if they do not acknowledge the potential for bias when a study is funded by staunchly pro-abortion organizations.

3. Many people do not regret decisions they make, even if these decisions have consequences that they are unaware of.

Further, people are not always willing to admit their regrets, and those who have regrets may be much less likely to participate in a study asking about their regrets over time. This is known as selection bias, acknowledged by Rocca et al as well. Good science treads carefully to make generalizable conclusions when selection bias is implicated.

After a year of working in addiction, I started to see how self-awareness is not a give-in, especially when people have been wounded. We are not always the expert on our own experience. A man suffering from alcoholism who sits next to you at the bar may tell you he is the happiest he has ever been. From what we know about human nature and addiction, we can conclude his self-report to be less helpful than that of the sober man suffering from alcoholism. The drinking man's report captures his experience in the moment, but the more important question is whether it is helpful in making conclusions about others in his shoes. Is he truly the happiest he has ever been? Is any alcoholic, or even many alcoholics sitting at the bar, the happiest they have ever been? Similarly, is the day a woman gets an abortion truly the happiest day of her life, for most American women? A news reporter may say so, but a single story cannot so quickly profess to be the majority story. And 667 stories cannot so quickly profess to be every story.

4. Always look at the strengths and limitations of a study before drawing conclusions.

Research provides us snap-chats of experience. It can never provide the whole picture, nor is it meant to. That is what makes it so important to highlight what we can and cannot say from any given study. In the limitations section of the Rocca et al study, the authors are hesitant to highlight any major limitations of their study. They do admit, however, that "Because no formal measures of abortion emotions exist, the scales we used may not have validly captured women's emotions."That is a pretty important limitation to keep in mind. Further, "the relatively low participation rate might raise concerns about selection bias", as stated above. All the more reason for this research method to be repeated, to ensure that the findings are capturing the real experience of the majority of post-abortive women, and not the political aims of pro-abortion groups.

There are two sides to every story. For every journal article that "disproves" something, there is likely a journal article that "proves" that very thing. Ironically, no good journal article should claim to "prove" much of anything, especially if it involves a research design and measures that have not been retested for validity or reliability. Good research can capture experiences, it can provide insights, but it must never overstate its power. More importantly, we must never overstate its power. And even if we take the latest study at its word, for every 633 women who are truly fine after abortion, there are likely 633 testimonials of women who are saying they are not fine. Let us not forget them either.

One more thing. How we think we are doing, or what is legal at any given time in history, may not be the best measure of our wellbeing after all.

"Never forget that everything that Hitler did in Germany was legal." – Martin Luther King Jr.

It was legal in Germany to kill people because of their ethnicity and religious belief. Hitler, if asked in 1936, "Do you regret your decision to blame the Jews for Germany's problems?", would have probably said a resounding "No." Slavery was legal in our country until 1807. Slave-owners, if asked three years after purchasing their first slave, "Do you regret your decision?", would probably have said no, more than 95% of the time. Should we ground the goodness of laws or choices based on the report of those who have made misguided choices? The legality of an act may impact greatly our experience of the goodness or morality of it. Nevertheless, legality, or the report of individual persons, do not determine the inherent goodness or morality of any act. We must not forget that. If we do, history will continue to repeat itself. In such relativistic waters, the will of those in power is free to run riot, dictating and destroying the dignity of the powerless.