Starting early on statistics

The CensusAtSchool project is introducing kids to data collection while creating an international learning resource, report Neville Davies and Doreen Connor.

Civilisations throughout the ages have recognised the need to obtain information about their most valuable asset, their people, and on April 29 the population of the United Kingdom will take part in just such a survey - the national census.

One adult in each household is responsible for filling in the census form, and each person aged 16 and over is required to answer up to 32 questions. But an adult answers on behalf of children under 16. So while information is gathered about children under 16, they do not have an opportunity to contribute directly.

The aim of the CensusAtSchool project is to involve thousands of schoolchildren in a version of the census and to help them appreciate the effort that goes into gathering, handling and analysing data on a national scale. CensusAtSchool comprises 18 questions divided into three categories - about the children themselves, their households and their schools. Some questions are identical to the ones on the adult census.

The exercise has much educational potential. Data handling is a key component of the UK national curriculum for many school subjects at Key Stages 2-4 and at A level, and CensusAtSchool will, it is hoped, help promote the improvement of standards of data handling among children across the curriculum.

The project is managed and run by the Royal Statistical Society Centre for Statistical Education at Nottingham Trent University and it is working closely with the Office for National Statistics, which is responsible for the adult census. The aim is for each school class to be used as an illustrative population for a census. The same principles apply to the class as to the general UK population. As they are collected, data are being added to a national database.

The initial findings from the first 54,000 responses show:

• 59 per cent of secondary pupils own a mobile phone

• 85 per cent of seven to 16-year-olds have access to a computer

• Children are most likely to walk to school in England, travel by bus in Wales and by car in Northern Ireland.

The information collected about the heights of schoolchildren has enabled comparison with data collected more than 160 years ago. Boys and girls today are 15-20cm taller than those of 1837. And whereas boys and girls in 1837 were consistently of similar heights, in 2000, boys aged 14 and over were significantly taller than girls.

The focus of the CensusAtSchool activities is the website www.censusatschool.ntu.ac.uk .
The distribution of questionnaires, the receipt of summary spreadsheets of the returns, worksheets for use in the classroom and the national database of pupils' responses are administered via the internet, although the project can be run using traditional means in countries where web access is less widespread.

Several other countries have expressed an interest in the project,and with a few minor revisions - such as definitions of types of households - the UK model can be applied just about anywhere.

Also, the exchange of information among schoolchildren in different countries will be a unique way to assess global social changes. For example, resources in homes and schools such as electronic methods of learning and teaching can be assessed on an international scale.

The Centre for Statistical Education plans to create a worldwide database so that teachers and pupils can have access to, and create learning and teaching material from, data to which they have contributed. This could enhance statistical literacy for many years to come.

The New Zealand and Australian equivalents of the Office for National Statistics are organising school census projects. The Queensland Treasury Department, along with Queensland University of Technology, is sponsoring the project in Queensland schools this summer, while Statistics New Zealand, in conjunction with Auckland University of Technology, will run a pilot census with the intention of carrying out a full census in 2002. Talks are also under way with Norway and Canada.

But it is the South African government that has been most enthusiastic about implementing CensusAtSchool. At the invitation of Statistics South Africa, three members of the Centre for Statistical Education spent nine days in Pretoria, Durban and Johannesburg discussing the project.

There was overwhelming support from everyone, including the director-general of the education department and teachers at universities. The idea of grounding a statistically literate society from school level up was universally applauded.

The South African Curriculum 2005 is a forward-thinking and ambitious venture in which the aims and objectives of CensusAtSchool fit beautifully. In particular, three out of seven critical outcomes for the curriculum can be illuminated by the pupils' involvement in CensusAtSchool, when it asks the children to collect data, incorporate the results in a spreadsheet and organise and evaluate findings.

Pupils can compare data with regional, national and international information: data on the heights of South African schoolchildren can be compared with the heights of UK pupils. The height difference between boys and girls might occur at a different age. The reasons for such a difference might be the result of diet and nutrition, and this issue could form the basis for classroom discussion.

Among the many problems facing South Africa is the lack of computer networks and internet access in schools. The CensusAtSchool project will, therefore, have to be distributed on paper. But the Centre for Statistical Education will put the information into a database, analyse it, distribute it on a website and create worksheets for use by South African teachers.

As more pupils and teachers become computer literate, and as access to the internet grows, the census data could provide a way of helping teachers and pupils see the web as an educational tool.

Neville Davies is professor and director and Doreen Connor is teaching fellow and CensusAtSchool coordinator at the Royal Statistical Society Centre for Statistical Education, department of mathematics, statistics and operational research, Nottingham Trent University.