Last month, the education directorate and researchers from the Australian National University went to war over school performance data at an assembly inquiry into standardised testing.

Education Minister Yvette Berry and Deb Efthymiades from the directorate attempted to hose down findings from a recent ANU paper which called for a public inquiry into "systemic academic underperformance" in ACT schools.

Rejecting the analysis as a "misinterpretation" of NAPLAN results, Ms Efthymiades instead pointed to a more in-depth but partially redacted report commissioned by the directorate and written by Professor Stephen Lamb at Victoria University in 2017.

That report had also found Canberra schools "on average achieve negative results on every measure" compared to schools of similar socioeconomic status, but had outlined the impact of an apparent class divide in the sector, as aspirational families moved their high-achieving children to popular schools outside their local area.

"[Professor Lamb found] the ACT performance issue is one of equity," Ms Efthymiades told the committee.

"Hence the really strong driver that came through from the minister, positioning equity at the forefront of the future of education [strategy].

"The VU report is the most robust report that could be used to inform our agenda."

But Professor Lamb told The Canberra Times an equity problem was no more to blame for poor academic results in Canberra than in any other school system across Australia, and he had included it in the report only to help sketch out the state of the local school system. His remit had been to assess the sector's performance, not find the reasons behind poor results, he said.

"Equity does impact on ACT performance, which is interesting given Canberra's relatively homogenous socioeconomic spread, but underperformance is not only due to that," Professor Lamb said.

"There's other things going on."

Policy analyst Andrew Macintosh, who co-authored the ANU report, told the inquiry a large-scale investigation was needed to find all the factors behind the problem and acknowledged that the ACT's uniquely thin socioeconomic spread could be partially to blame.

But if that was the whole story, he said the impact would likely be felt across the board. Instead, ACT students consistently perform better in reading than in other categories, with the worst results emerging in writing and numeracy.

Professor Lamb agreed policymakers needed to look at school performance at a classroom level as well as the impact of an equity divide.

Released in August, the ANU report revealed a "sea of red" where Canberra students were lagging up to 16 months behind their peers, particularly in high school. Ms Efthymiades, who has worked closely with the agency responsible for NAPLAN, ACARA, and acknowledges any comparative anlaysis of the data is tricky, said Professor Macintosh had not factored in appropriate margins of error as his data was limited to what was publicly available on the My School website.

If these had been built into the analysis, the results would have look much more like a sea of yellow, she said, on par with the national average.

Professor Macintosh hit back at the directorate's claims, saying even factoring in the full margins of error would not drag the results out of the red.

"Those numbers are miles out, I don't care what you do to them," he said. "We've explained our methodology clearly in the report.

"ACARA and the directorate deliberately withhold data to prevent external scrutiny so it's pretty rich to critique [us] for not following a particular method they know we can't follow."

The directorate also pointed out that in some cases academically selective schools in Victoria and NSW were being compared with ACT public schools, likely skewing the results, particularly for high schools.

Professor Macintosh said this would only affect about two or three schools enough to fudge the numbers.

In his own report, Professor Lamb excluded selective-entry schools but noted that decision might have in turn raised the performance of ACT schools.

The territory's new education strategy was also built on 18 months of consultation with school communities and recommendations from a 2017 ACT Auditor-General report.

That report again found underperformance across the sector and cast doubt over the territory's model of giving schools autonomy, though noted the directorate was already rolling out some initiatives focussed on lifting performance in literacy and numeracy.

"We are really on track in terms of delivering on the auditor-general's findings which came from the Lamb report and gives us a real essence of the...strategy," Ms Efthymiades said.

The strategy places holistic student-centred learning at its heart, promising to roll-out models appropriate to the individual school's context, in line with "international research". While the practical details are yet to be released, the strategy has also signalled plans to share effective practices across schools, and evaluate interventions in student learning.

As NAPLAN continues to come under fire in the sector and in the committee room, Professor Lamb said he had no problems with the test, but said it needed to be expanded beyond literacy and numeracy to include a broad range of capabilities and subject areas such as science, history and music.