Prof Sugata Mitra has shown that groups of eight-year-olds can answer GCSE and A-level questions using the Internet, and also retain this knowledge several months later.Writing in the Journal of Education and Human Development, Newcastle University’s Prof Mitra and Emma Crawley, a teacher at St Aidan’s Church of England Primary School in Gateshead, UK, where the research took place, also discovered that the harder the question, the better the results.
“From these small sample studies we can conclude that groups of children, using the Internet, can achieve acceptable test scores using material seven years ahead of their time,” says Prof Mitra.
The aim was to measure the effectiveness of self-organised learning environments (SOLEs) for children, building upon Prof Mitra’s hole-in-the-wall experiments which began in 1999 to see if this method could translate effectively into a classroom setting.
They set up an environment that simulated the conditions in the ‘hole in the wall’ experiments to create a SOLE in a Year 4 classroom.
A SOLE usually involves one computer connected to the Internet for every four children. They are then given a question and asked to research the answer. They will need to form groups to carry out the work, but are not told to do so. Moving around, changing groups and looking at what others are doing is all allowed, and there is no adult intervention. At the end of 45 minutes, each group is asked to make a brief presentation of their results.
“After their initial disbelief that we were going to let them do ‘anything they like’, the children reacted with great enthusiasm and declared the SOLE to be a great way to learn,” says Prof Mitra. “One child even said afterwards that they weren’t ‘working’ as that was when the teacher told them things and they have to write them down.”
Four experiments were carried out over a three-year period from 2009 to 2011, presenting the children with material from their age level (eight-years-old), through to GCSE (14-16-years-old) and up to A level (16-18-years-old).
The children were left to research increasingly difficult topics in groups using the Internet, tested immediately afterwards as a group, and then tested several months later individually without the Internet.
In the first experiment, 80% of the students got the question right three months later, which was actually higher than they did initially in their groups (76.8%). The children explained that they had researched further in their own time and some had discussed it with their parents or classmates.
“This is quite contrary to the belief that learners ‘forget everything’ after an examination and would do badly in the same examination at a later date unless they ‘prepared’ again for it,” says Prof Mitra. “We called this ‘anomalous expansion of understanding’, which is maybe characteristic of this age group’s response to technology.”
The individual recall of the test answers was consistently good over time, which seems to indicate that children in a group all learn quite uniformly even when different children do different things. One child operates a computer, another takes notes, a third directs the other two, while a fourth entertains, and often disturbs, the other three. They take turns in each of these roles.
‘Flocking’ behaviour was also observed, when almost all the children will go over to another group to look at something important. Then this ‘flock’ would disband and return to their groups. The fact that they organise their own interchangeable groups is very important to the children.
Most children continue their discussions after a SOLE session at home, in the playground or on their computers and phones.
“This type of learning seems to work because of the ability of groups of children to read and understand reading material at a higher comprehension level than each individual in the group,” says Prof Mitra. “This intellectual ‘amplification’ may lead us to an explanation of how the learning process in a SOLE works.”
Teacher Emma Crawley adds: “Neither of us entered into this project with any expectations other than having fun and learning with the children. The significance of what we were seeing unfolded quite steadily and over the course of many months. The diary I kept was a vital piece of evidence for how SOLE developed and to this day, I refer to what was learnt from SOLE in all aspects of my teaching and learning career.
“It's also something I am extremely proud of. I'm pleased we had no expectations because that is a crucial element of how SOLE works.”?
A school-based SOLE can either be timetabled every week with a general Big Question such as ‘why is blood red?’; built into the curriculum so the question reflects a school-leaving examination, such as GCSE or SAT in the UK; used as an aspirational process where the children listen to a short lecture on the Internet (such as TED talks) and then present their findings about this particular issue; or open for free use before and after school hours.
Next week (March 2 2015) School in the Cloud, which is part of SOLE Central and joins together SOLEs from across the globe, is joining forces with Skype in the Classroom to get more children exploring Big Questions as part of a six-week literacy campaign.
Paper: Journal of Education and Human Development September 2014, Vol. 3, No. 3, pp. 79-88 ISSN: 2334-296X (Print), 2334-2978 140124%20Arthur%20D.%20Little%20%26%20UITP_Future%20of%20Urban%20Mobil
Background: How the Gateshead SOLE experiment was conducted
Four experiments were carried out over a three-year period from 2009 to 2011, using the following questions:
- Can groups of children, using the Internet, answer test questions ahead of their time and obtain acceptable scores?
- What difficulty levels (i.e. how far ahead of their time) can they obtain acceptable test scores in this environment?
- Do they retain individually what they’ve learnt in groups?
- Can children read and understand better in groups than individually?
For the initial experiment, Year 4 (eight year olds) were given five GCSE questions about how animals adapt to their environments to answer within 45 minutes using the SOLE method. When the groups handed in their answer papers for grading, the same score was assigned to each member of the group. Three months later, the children were tested individually on the same questions without the use of the Internet. The children did not know they would be tested again.
80% of the students got the question right, which was actually higher than they did initially in their groups (76.8%). The children explained that they had researched further in their own time and some had discussed it with their parents or classmates.
Trying more ‘difficult’ subjects, three tests were devised from GCSE Physics and Biology subjects and the same method as the first experiment was repeated with nine-year-olds. The results varied from a slight to a significant increase over time, with one test showing an 80.4% recall individually compared to 26.1% in a group.
The third experiment was based on A Level questions: molecular structures, radiation and geography. This time they also checked whether the children had any previous knowledge of these subjects which might bias the results. In all apart from Geography (17.4%/30%), the answers were better in groups with the Internet than individually without, but still showed good quality retention of learning across the board.
The results of the first three experiments, along with earlier experiments in India, indicated that groups of children could, with the help of the Internet and without supervision, understand topics traditionally considered many years ahead of their age level. So the final test looked at whether children are capable of understanding text at levels ahead of their individual capabilities if they can read in groups.
Dividing a class into groups and individuals, the children were given a text suitable for eight-year-olds (their average age). The groups were given one copy of the text to share, where the individuals were given one copy each. They then swapped over so each child experienced the text as both a group and an individual, and were given a text meant for 10-year-olds.
While the children read better in groups for both texts, the difference was significantly greater for groups with the more difficult one. The average reading comprehension of the 10-year-old level text was 77.8% as a group/50% as an individual, compared to 66.1%/61.3% with the eight-year-old level text.
published on: 24 February 2015