Menu Close

On the Hunt for credible information [citation needed]

Research is hard, but it’s worth doing properly. ulisse albiati

A valid argument about a scientific issue requires support using robust, objective, peer-reviewed scientific evidence. This notion is drilled into university students from the beginning of their tertiary education.

As a lecturer, the first thing I tell my students when they are undertaking research for an assignment is to never, under any circumstances, use Wikipedia as a credible scientific source.

Wikipedia is an open-source encyclopaedia where anyone can update the content at any time and it’s up to others to correct any mistakes. It’s ripe for misrepresentation, personal bias and the occasional joker.

Wikipedia might technically be a form of peer-review, but it’s not rigorously undertaken and there are no guarantees that those updating the page are experts in their field.

In response to a statement from a UN official that the bushfires currently ravaging NSW could be linked to climate change, environment minister Greg Hunt said on Thursday that he

looked up what Wikipedia said […] and it opens up with the fact that bushfires in Australia are frequently occurring events […] and that’s the Australian experience.

He was widely lambasted.

But just how accurate is Wikipedia for scientific information?

I asked my students exactly this question this year. I gave them an assignment: evaluate the content of nominated Wikipedia pages for scientific accuracy. The pages covered topics on climate science and the students compared these Wikipedia pages to the peer-reviewed scientific literature.

Johann Dréo

My students found that the Wikipedia pages missed important nuances in the science that were fundamental for understanding. Most contained information from sources that were either non-authoritative or non-verifiable. Some also contained subtle but significant inaccuracies.

The assignment I set my students was a simple analysis of certain content. It was not a rigorous and in-depth study of the general accuracy of Wikipedia in reporting climate science or bushfires. However, the assignment did highlight that using Wikipedia as a source of scientific information is fraught with problems.

Mr Hunt correctly quoted Wikipedia as saying, at time of writing,

Bushfires in Australia are frequently occurring events during the hotter months of the year due to Australia’s mostly hot, dry climate. Large areas of land are ravaged every year by bushfires.

This is true.

However, the fires currently surrounding Sydney started in mid-October, which is far earlier than the normal “Australian experience”. Moreover, the fires have occurred following months of very dry and record-breaking hot conditions.

The Intergovernmental Panel on Climate Change, CSIRO and a wealth of other peer-reviewed scientific literature describe an increasing prevalence of such fire weather conditions because of human-induced climate change both now and into the future.

Surely Mr Hunt should be following and quoting evidence provided by these objective expert organisations rather from than website whose slogan is “the Free Encyclopedia anyone can edit”. (Yes, that is a link to Wikipedia for a definition of Wikipedia.)

Policy decisions should have a firm basis in evidence built on reliable science. Politicians have access to a wealth of scientific knowledge on climate science and bushfires. They can use the Bureau of Meteorology, CSIRO, university-based researchers, the Chief Scientist and other experts, including the Bushfire Cooperative Research Centre (now called the Bushfire & Natural Hazards Cooperative Research Centre).

Using Wikipedia to establish and defend a scientific argument is like using it to diagnose an illness in lieu of a doctor’s judgement. You might do it if you had no access to a doctor, but when there are so many doctors around, it makes no sense.

If Mr Hunt were one of my students, I’m afraid I would have to fail him.

Want to write?

Write an article and join a growing community of more than 184,400 academics and researchers from 4,972 institutions.

Register now