Academics win access to data goldmine

MELBOURNE

Australia's vice-chancellors have negotiated a unique agreement with the Australian Bureau of Statistics that will give academic researchers electronic access to a vast quantity of the bureau's data.

The agreement will enhance significantly the capacity of university researchers to assess and report on Australia's social and economic trends. It is the first time the ABS has agreed to a deal with the universities on a sector-wide basis.

Under the agreement, researchers will be able to access the bureau's Confidentialised Unit Record Files - or CURFS. These contain unidentifiable record data from most of the key ABS social and labour household surveys.

The Australian Vice-chancellors' Committee negotiated the deal with the ABS to help universities minimise charges they pay for access to the CURFS. AVCC deputy executive director, John Mullarvey, who has charge of the agreement, described it as a major breakthrough for scholarship.

"In time, this will be a great boon not just to university research but also to social and economic planners," Mr Mullarvey said. "The best data will give us the best opportunity for the highest quality assessments. Better assessments, in turn, should give us better planning."

Mr Mullarvey said the joint action by universities under the agreement should significantly lower the financial disincentive researchers had in accessing the CURFS independently. He said that while the ABS files did not contain information that could be used to identify any individual, release of the data was still restricted to academic teaching and research purposes.

Academics misusing the data could face a two-year jail sentence and a Aus $5,000 (Pounds 2,000) fine. Each application will be personally vetted by the chief statistician.

Some of the data sets available under the agreement include surveys of national literacy, families, housing, child care and care for the disabled and the elderly, household expenditure, income distribution, labour mobility, national health and nutrition, and education and training experiences.

A bureau spokesman said the files were in digital format and the content was raw "confidentialised" data. In most instances, the only way sense could be made of the content was in the hands of a skilled researcher using analytical tools such as SAS or SPSS.

"It is the output from their analysis which can be summarily printed or graphed," the spokesman said.

The bureau checks each unit record file to ensure that individuals cannot be identified.