But the Australian Curriculum, Assessment and Reporting Authority (ACARA) said the results do show some significant improvements in the areas of numeracy, reading, and language conventions across the year levels, compared to the base year of 2008.
Australian students in Years 5 and 9 numeracy, Years 3 and 5 reading, Years 3 and 5 spelling, and Years 3 and 7 grammar were significantly above the 2008 average.
But the writing tests results in Years 5, 7 and 9 were all well below those observed in the 2011 base year.
ACARA chief executive Robert Randall said this is a point of concern.
"I think what we'd like to always see is improvement in results," he said.
"We need to have a look at the teaching of writing, and see whether there are various ways that we can improve the testing of it."
"But what's tested in the writing is what's taught in the curriculum - the fundamentals of how to write a persuasive argument or a narrative argument," he said.
This year about one-fifth of students from Years 3, 5, 7 and 9 completed the NAPLAN online.
This is part of a wider plan to transition the test to digital by 2020.
The Australian Education Union (AEU) said the validity of this year's results must be questioned.
AEU President Correna Haythorpe said the conflicting methods will provide unfair results.
"The Naplan data this year is essentially useless," she said.
"Effectively, you've got 200,000 students who have sat this online trial and you cannot compare their results with the 800,000 who've done their test by pen and paper."
The criticism comes from the different questions asked - students sitting the test with pen and paper all answer the same questions, while those completing the test online get questions that evolve, based on their answers.
Author of a report commissioned by the New South Wales Teachers Federation, Professor Les Perelman from the Massachusetts Institute of Technology, said comparing this year's data with previous years taken in the traditional way is like comparing apples and oranges.
"They're really not comparable," he said. "And even if they were comparable, it would really be impossible to equate scores."
"I know of no other large-scale test that ever did something like this where they gave part of the population pen and paper, and part of the population a computer-adapted test," he said.
But ACARA's Robert Randall said there is no cause for concern.
"We've worked with state and territory people to give us the confidence to say that data can be put on the same scale," he said.
"So mums and dads can be confident that what they're getting back when they get the individual student report is an accurate, reliable indicator of what their son or daughter was able to do when they sat down and did the NAPLAN test," he said.
Individual student reports will be distributed to homes from now through to mid-September.
The final report will be published in December.