Residents of a remote Arnhem land community have destroyed about 50 same-sex marriage surveys in a fire, because they misunderstood the question.
Australian Bureau of Statistics deputy statistician Jonathan Palmer told the ABC a group of people in Ramingining, 560 kilometres east of Darwin, had believed the survey was asking whether or not a man should be "compelled" to marry another man.
In reaction to the misunderstanding, Mr Palmer said the residents burned their forms, and the incident was brought to the attention of the ABS by a member of the community.
Mr Palmer said ABS staff then contacted the residents to explain what the survey was asking, and successfully re-issued the forms.
Many people in remote Aboriginal communities do not have English as a first language, and the ABS has deployed teams of field officers to about 200 communities to assist with the returning of surveys.
"Our people are out there for between one to three days, and I think there's plenty of time for them to have the length of conversation required," Mr Palmer said.
ABS staff do not have interpreters with them when they visit remote regions, instead relying on audio recordings of survey explanations in about 14 Aboriginal languages.
"I'm very satisfied that we've got a pretty comprehensive program to get out and give as many people as possible a say in this matter," Mr Palmer said.
"I think our estimate of 9.2 million forms received as of last Friday shows there's been very strong community interest in this.
Results won't shed light on areas with poor participation
When the results of the same-sex marriage survey are released, Australians will not know exactly which remote parts of the country did not participate.
The smallest breakdown of the survey's results will be by federal electorate.
For the seat of Lingiari, which takes up the entire Northern Territory except for Darwin, the data won't show which remote communities had poor participation.
Mr Palmer said the survey results would be detailed enough for most people.
"What you'll see for every electorate is not only the results for that electorate, but you'll also get how many people participated and the breakdown of those by both their age group and their gender," Mr Palmer said.
"That information will give people the ability to have their own informed view on what reliance they can place on the statistics."
