"Research suggests that bilingual education helps children who are learning English." (Lisa Maree Williams/Getty Images)
Ms Scrymgour has repeatedly said there is no evidence to show that bilingual schools accelerate English literacy.
But two reports by her own department have surfaced. Both say schools that teach young indigenous children in their own language in initial years, and gradually introduce English, ultimately achieve better English literacy outcomes.
Ms Scrymgour last month angered the Territory's nine bilingual schools by announcing they must teach the first four hours of every day in English.
"The evidence is quite simple, there is no significant benefit to bilingualism as it is currently structured," she added this week.
"Indeed there are strong suggestions to the contrary that the nine bilingual schools fare fairly poorly in comparative terms to the acquisition of English language skills."
However, preliminary results from the Evaluation of Literacy Approach (ELA) report, leaked to AAP, found that for "active reading skills in English" students at bilingual schools achieve better results than non-bilingual schools by the time they reach Grade 5.
"Given international research and data, this result is expected to continue through to Years 6 and beyond," say the initial findings, which are due to be completed by April next year.
The Australian Council for Educational Research is compiling the data from 20 schools from four grades over four years for the NT Education Department.
The initial findings support the NT government's Indigenous Languages and Cultures 2005 report which found there was "irresistible evidence" that students achieve better levels of learning using home languages.
Under the teaching method, students are taught in their local language over the first few years of schooling and are gradually introduced to English. The method is based on the belief that literacy skills in a first language lead to greater success in the acquisition of English.
"Bilingual education programs achieve higher levels of outcomes, including literacy outcomes in the mainstream language, than non-bilingual programs in similar settings," the 2005 report said.
Ms Scrymgour's initiative is a response to appalling education results for indigenous students after more than 60 per cent failed the most recent national reading benchmark.
More than 70 per cent failing the writing benchmark.
But the 2005 report, which compared results from 10 bilingual schools and 10 English-only schools from 2001 to 2004, found that after three to four years of schooling, bilingual students had "built the bridge to English".
"Their MAP results surpass those students in the 'Like' (English-only) schools," it said.
"Research suggests that bilingual education helps children who are learning English."
The report also found its data replicated studies done internationally and within Australia.
Michael Duffy from the Council of Government School Organisations said he was not aware of any evidence to support Ms Scrymgour's claims that bilingual education was underperforming.
"I have no idea what research Marin is talking about and she won't show us because she can't," he said.
"This approach just flies in the face of all of that (existing research), it's not just shonky, it's shameful."
Mr Duffy said there were nine bilingual schools out of 150 schools in the Northern Territory.
"Thirty-five per cent of kids across the board in the Northern Territory don't reach benchmarks," he said.
"Now that suggests to me that there is a systemic failure, it's not the failure of the program."
Comment was being sought from the NT government.
