In 2009 a study by Priscilla Coleman and colleagues made a very big claim, that woman who had abortions were more likely to be diagnosed with mental health problems later in life. no only was this claim very big, but it was also very false. If the above statement is in fact true it is surely not proven by this study.
The flaws in Coleman’s study were brought to light when researchers Steinberg and Finer tried to replicate the study in 2012. When the failed to do so they contacted the original authors and looked more thoroughly into Coleman’s data.
Coleman at al. had taken into account diagnoses throughout the women’s lifetimes. They had no data regarding if such diagnoses had been given before or after the time of the abortion.
Pricilla tried to defend her study. Here’s what the New York Times reported:
[…] Priscilla Coleman of Bowling Green State University and her co-authors included all lifetime mental health disorders in their analysis, rather than only those instances occurring after the abortion took place. They were “hoping,” she says in a letter defending her methodology, “to capture as many cases of mental health problems as possible,” by including a longer period of time.
This argument is flawed. To say she was simply trying to capture as many diagnoses as possible is basically saying she did so to support her hypothesis and risked her validity as a result is experimenter bias. It in no way supports her methodology, it is simply an excuse.
To further defend herself, she now says that “majority of mental health problems likely occurred after the abortions” because most of the abortions studied occurred before age 21, but the data regarding diagnosis used in the study were not obtained until they were, on average, age 33. This is a valid point. However such facts should be included in the studies write up and to say ‘most’ does not mean all. It is a factor that could have been easily controlled and it seems Priscilla did not control it only to get the data she wanted. This goes against the idea of trying to prove the null hypothesis when researching.
The true numbers still remain unknown. The researchers only looked at the lifetime diagnoses with the knowledge of if that woman had an abortion at some point. Therefore the true outcome of the data could be that women who have mental health problems are more likely to have an abortion. Without the order of events known it is hard to make any definitive statements.
Julia Steinberg of the University of California at San Francisco and Lawrence Finer (2012) of the Guttmacher Institute found what they called, in a letter to the journal’s editors, “untrue statements about the nature of the dependent variables and associated false claims about the nature of the findings.”
“This is not a scholarly difference of opinion,” Dr. Steinberg said. “Their facts were flatly wrong. This was an abuse of the scientific process to reach conclusions that are not supported by the data.”
It is of course true that abortion causes emotional trauma. Researchers should try and better understand the psychological effects of such a procedure, but it should be done with honest, careful research.
libbyayres 
Many pieces of research into the psychological effects of abortion have been conducted, and none found that abortions increase the likelihood of developing mental health problems. Major et al. (2009) evaluated the evidence of some of the research conducted and concluded that mental health problems found in women who had previously had an abortion were probably due to pre-existing factors in their lives placing them at a greater risk, rather than the abortion being the cause. If mental health problems do occur after an abortion, they may be the cause of the way that emotions were dealt with, rather than a direct cause of the abortion.
The research that you describe by Coleman et al. (2009) cannot be backed up by many pieces of research, suggesting, as you have said, that experimenter bias played a massive role in the conclusions reached. The women who developed mental health issues after having an abortion may have been placed at a higher risk by other life factors and stressors, or perhaps didn’t have effective coping strategies, whether they were for emotions experienced after the abortion or other life events, as previous research suggests. Coming to such conclusions can make them dangerous for women considering abortions, as their pregnancy may be unwanted, but they feel they have to continue with their pregnancy so they are less likely to develop mental health issues. It may actually be that the emotional trauma of an abortion is not great enough to increase their risk, but is wrongly believed to be because of subjective research that lacks validity.
References:
Major, B., Appelbaum, M., Beckman, L., Dutton, M. A., Russo, N. F. & West, C. (2009). Abortion and mental health: Evaluating the evidence. American Psychologist, 64(9), 863-890. doi: 10.1037/a0017497
It is obvious that the results of the experiment were not very valid and the claim that ‘woman who had abortions were more likely to be diagnosed with mental health problems later in life’. This study was replicated and results were not consistant with the claim. The internal validity is very low in this experiment. Data cannot be found which means it may have been manipulated to support claims. It means that data found in the repeated experiments cannot be compared to it.
Pingback: Blogs I Liked. « Anthony's Research Methods Blog