The politics of research

It would be hypocritical of me to stand on my professional dignity and claim that survey research from an engaged researcher should be devalued. I am a professional researcher and a well known activist, so am very aware of the contradictions in these roles, consequently have always been very committed to professional and ethical practice when I publish or publicise the results of any survey I am involved in. I make all aspects of my survey methodology public so those who may disagree with my personal position can judge whether it has affected the research data collected. We may then agree or disagree on the interpretation, but, hopefully, not on the legitimacy of the data collection.

So in writing this article, I am applying what I consider these appropriate ethical and professional standards to a set of survey results that were released recently by the Southern Cross Bioethics Institute in Adelaide. These were claimed to be a legitimate record of Australian attitudes to abortion but, by their failing to release their data collection instrument, I would claim the legitimacy of their results becomes questionable.

Social survey research is often political: when a survey is done, who funds it, what questions are asked and what is done with the findings. The funder and or researchers may be responsible for what they have done, insofar as they make the decisions on what questions are to be asked, and who will be asked. In selecting the wording, the coding, the sample and interpreting the results, researchers inevitably stamp the product with their own interests, ideas and values, and or those of the funding body.

If the research is part of a wider project that may be involved in some advocacy or change exercise, it’s more likely to be questioned than something that may be primarily an academic exercise. That is why there are conventions that govern the ethics of good quantitative research which include making one’s methodology available for public scrutiny. I teach this to my research methods students and in my lectures jokingly state that I keep an image of Fred Nile in my head to ensure that my questions are not leading or designed to get particular answers I want.

So when I saw the media reports of a survey by the Southern Cross Bioethics Institute called Give women choice: Australia speaks on Abortion (pdf file 1.76 MB) I downloaded the summary on their website and read it with some interest. There were some useful findings but some of their reporting of results worried me. I wanted to see what questions they had asked so I could read the data and see whether their interpretations were legitimate. I rang a polite young man who said he’d check with the researchers and get back to me. I was then rung back by the PR person who told me the questions were not available, as they had been informed this could contaminate future stages of the project.

I was taken aback because this was such a weird excuse as it was highly unlikely that anyone who would read the questions could be involved in a future stage. I argued it was poor professional ethics not to provide the questions and pointed out there were big differences in interpreting responses to pre-coded or open ended questions. This had no effect and she was adamant that the questions were not available and there was no one else available with whom to discuss this.

I was surprised that a so-called bioethical organisation was prepared to breach a basic code of good ethical research practice and told her so. Releasing such a controversial issue without being open about the research instrument is not professional good practice as some of the interpretations seemed to reflect researcher views and their legitimacy could not be judged without the original question wording. This was disturbing because these results were distributed widely to media and politicians.

I became more curious about its genesis and the secrecy. This was no modest search for knowledge but presumably part of political campaign in disguise. What was this institute and its connections? The website claims its non sectarian but it’s owned by a Catholic health group. The principal researchers are both public anti-choice advocates which is not a problem were they not also pretending to be objective. I am now also curious about the funding for the research, as this was not a cheap exercise. I estimate the fieldwork would have cost about $25,000 and it is claimed to be one part of a four stage project! Who is paying the piper and how much does that influence the tune?

They have published some facts: the sample is properly drawn, the CATI system for completing interviews is legitimate, consent is fine and their quoted margins of errors are legitimate. But the questions themselves still elude me apart from some indications in the results. Their major claim that attitudes are “soft” is based on a form of words that uses the term “abortion on demand no matter what the reason” which is double barrelled and sounds very authoritarian. So a 62 per cent agreement is not surprisingly lower than other surveys. Asking if women should have a right to choose whether to have an abortion is less aggressive and responses tend to be different. This explains why there was an 82 per cent response to that type of question in the 2003 ANU survey.

Did they really offer the term “necessary evil” as a response? It appears in the results and I want to know what other responses were offered before taking those answers seriously. What do they mean by the “morality of abortion”? How do I interpret that one in four think abortion is “morally justified”? Morality is a term that creates ambivalence regardless of its context so what was the question they put? Many people do not personally approve of abortion but still think it should not be illegal.

There could be some interesting findings in the survey, but should they be believed? I don’t think so.