In April 2020, Flinders University researchers published an online survey entitled ‘Australian Perceptions of Abortion Study’. The survey appeared to have bias and, concerned that it may be referred to in abortion debates in state parliament in coming months or years, I applied under Freedom of Information laws (FOI) for the survey’s supporting data.

Though many of the documents sought were withheld and others substantially redacted (ie ‘blacked out’ for secrecy), from what I have received there are clear indications that the survey’s findings should be dismissed as inherently flawed.

The problems start right at the beginning, with the aim of the project that launched the survey being in fact, ‘Perceptions, Experiences and Impacts of Abortion Stigma in Australia’ (emphasis added). The project was biased from the beginning, assuming that there is abortion stigma in Australia and seeking to build the case to back this up. This aim directed all the questions of the survey.

Many people also picked up and commented on a biased assumption that this abortion stigma came from people of faith. I commented in May 2020:

“A quick scan of the survey will show you that it has a heavy pro-abortion bias. There is also a concerted attempt to paint anyone who is pro-life as being motivated by religious views. That will allow the researchers to discount those views as being ‘merely religiously based’”.

We know from standard best practice in conducting research, things which make up a solid, scientific survey include asking unbiased questions. A biased aim, which led to biased questions and assumed hostility towards people of faith should therefore render this survey’s results as unscientific and unreliable.

Best practice in research also indicates that a survey must define the target population to be researched and then devise ways to get a representative sample of that target. This abortion perceptions survey claimed to be a survey of Australian perceptions of abortion. But the FOI reveals that the survey was likely only advertised on the Flinders Uni Facebook page. Are those who follow the University Facebook page representative of the general Australian population? Facebook’s demographics of the audience of a SA university page would likely indicate that this audience, for instance, would have a disproportionate number of people with a university degree or who are employed at the university. According to the 2016 census 75% of Australians have not attended a university. What about their views?

The unscientific methods of gathering data then worsened when it was suggested to a researcher that the survey be posted on Twitter as “our audience there does respond well to anything related to research and academic studies.” The researcher responded “No worries, sounds great.” How did a researcher think that the Twitterverse represented the views of the people of Victor Harbor, Port Adelaide, Mackay, Kalgoolie or Katherine? What due diligence was done to ensure that the views of farmers, mechanics, truckies and baristas were represented?

The population surveyed was clearly not representative of the general Australian population. A more accurate survey title would have been ‘Flinders University Students and Alumni Perceptions about Abortion Survey’.

After doing the survey myself and noting the bias in the questions and its title, I notified ACL supporters and encouraged them to participate in the survey. Many did. Interestingly, the FOI reveals from an email (15 May 2020):

“This study has been shared on fb (over 100s of times from an ACL post) and on twitter (sic) by various groups. This morning I woke up to several thousand responses each from women’s rights groups and the ACL, so it’s something of a twitter (sic) war/who’s go the biggest following apparently (I’m not on twitter (sic)). I never responded to the group who got in touch offering to share it’s happened organically anyway.”

On Monday, 18 May the survey was closed.

It is interesting to note from the comment above that:

  • Someone at the University considered that one large group of responses were “from the ACL”. It didn’t appear to matter to them that the responses were from individuals, with committed views. They were lumped in as one group, however this may not have been a conclusion of all the researchers (the redacted documents do not allow a conclusive determination in that regard).
  • A group that was not ACL had contacted the University and offered to share the survey. The University did not respond.
  • The survey received 1,000’s of responses from people who the University identified as being from “the ACL” or “women’s groups” - but the general populace has not properly been surveyed. Therefore the survey’s answers reveal nothing about ‘Australian Perceptions about Abortion’.

The seeds of this survey’s destruction were in its biased aim, questions and likely unscientific method of data gathering. Identifying and grouping the sources of the responses compounded this problem.

The survey has no credibility. It should be abandoned and dismissed the moment anyone seeks to draw conclusions from it.