Twenty-one research projects largely funded by the federal government breached integrity standards in the past two years, figures from the Australian Research Council reveal.

The federal government body, which provides tens of millions of dollars in taxpayer-funded research grants, stopped funding three grants, recovered funding for one grant, and placed restrictions on funding for one researcher.

Misleading data, ethical breaches and other "minor errors" involving referencing were reported to the council. A total of more than $3,367,700 in government money was linked to six problematic projects in the past two years.

The council would not reveal which universities or academics were behind the research, while several universities approached by Fairfax Media would not disclose data on misconduct investigations.

It comes as research integrity campaigners criticise the country's leading research bodies for proposing to allow universities to internally investigate misconduct.

Walter and Eliza Hall Institute of Medical Research's deputy director Professor David Vaux took aim at the new draft national code for researchers, authored by the Australian Research Council, National Health and Medical Research Council (NHMRC) and Universities Australia.

Professor Vaux said the removal of a requirement that universities establish an independent external body to investigate serious misconduct was worrying.

"If it's all done internally, then there are serious conflicts of interest," he said.

In an attempt to streamline the code, the authors proposed scrapping the term "misconduct" and instead allow universities to create their own definitions.

Professor Vaux said this allowed too much "wriggle room".

"Universities could say no misconduct has occurred, even if there was a major breach."

A years-long call for a watchdog to police research misconduct in Australia has been ignored in the draft, Professor Vaux said, despite the US, the UK and many European countries having a similar regulatory body.

Emeritus professor of obstetrics and gynaecology Alastair McLennan said misconduct at the Group of Eight universities was rare, but "second-rate" institutions offering profitable pseudo-scientific courses were "not up to self-police".

The Friends of Science in Medicine's vice-president said he feared a less prescriptive code would create a "variable standard across universities" on research integrity.

In explanatory material, the code's authors said they were moving away from a "rigid definition of research misconduct" as it was incompatible with enterprise agreements.

University of Canberra's assistant professor, Dr Wendy Bonython, who sits on several ethics committees, said a less prescriptive code could allow for more instances of misconduct to be prosecuted.

Last year, former University of Queensland professor Bruce Murdoch pleaded guilty to 17 fraud-related charges relating to Parkinson's disease research.

In 2015, former researcher at Baker IDI Heart and Diabetes Institute Dr Anna Ahimastos admitted to fabricating research on blood-pressure medications.

The explosion of fake academic journals continues to threaten scientific research, with academics being paid to cite products in their papers. Last week, leading academic journal publisher Springer retracted 107 papers approved by fake peer reviews.

University of Sydney's children's cancer expert, Professor Jennifer Byrne, who wrote a paper exposing "systematic fraud" in China's cancer research, called for more protections for whistleblowers.

An NHMRC spokesperson said it was considering submissions to the draft code, which was aimed at being "relevant across the broad range of research institutions".