Pupils learn poorly when using most computer programs

"In a pilot study, we examined the top 100 apps within math and Swedish, and barely half of them could be considered digital learning tools according to our standards, only 17% of which provided some sort of informative feedback. Some were so bad that we, as researchers, would never even consider to test them in class," says Björn Sjödén.

One example is the computer program to teach parts of speech, where illiterate 5 year olds do better than those who can read. A 5 year old who quickly guesses multiple times performs better than someone who tries to read and spell correctly.

"Probably more than 90% of the learning tools available online are simply test tools. They provide no explanatory information in addition to the correct answer. The pupils often compete against time, but not towards greater understanding," says cognitive scientist Björn Sjödén.

Björn Sjödén has a background in the computer games industry and is part of the interdisciplinary research group ETG (Educational Technology Group) at the universities of Lund and Linköping in Sweden. In his doctoral thesis, Björn Sjödén defines 'digital learning tools' as "subject-specific, interactive computer programs that provide feedback to achieve a specific learning objective."

In the last 15 years, Sweden has invested heavily in iPads and laptops for pupils, and, compared to other European countries, we are far ahead in terms of IT technology in schools. But the latest report from the Programme for International Student Assessment (Pisa) showed that the students who use the internet the most, both in and outside of school, also perform the worst on the PISA tests (standardized testing).

"However, digital learning tools can provide great educational benefits, as long as they do not become books on a screen, but use their digital advantages. This involves providing good feedback, showing that there are different ways of thinking to reach a goal, and presenting consequences that that cannot be demonstrated in a book," says Björn Sjödén.

For example, when calculating how long it will take you to get to the train station, a miscalculation of 13 minutes will result in the train leaving 13 minutes before you get there, or you having to wait x number of minutes. In chemistry, it is possible to show what happens if you combine different substances -- it may begin to bubble or explode.

Björn Sjödén has had two groups of pupils play a math game for eight weeks. Both groups were to help a computer character -- a digital pupil -- throughout the game. Then one group was to take a digital math test where the same character was featured. The other group took the same math test without their digital friend.

"The pupils that were helping their digital friend were more engaged. They wanted to solve more and harder math problems to help their digital character. Especially low-performing pupils became more motivated. This knowledge should be utilised in digital learning tools," says Björn Sjödén.

The research group ETG collaborates with Stanford University and others, to develop and study three digital learning tools -- two in math and one in history. The software is non-commercially developed, and will be free.

"However, researchers cannot be the only ones leading the way in the development of digital learning tools. If none of the large, well-established companies will, I hope that one of the new enterprises will succeed. The developer who makes the first real digital learning tool will have control of that entire market," says Björn Sjödén.