How does he know that?" I wonder how many of our pupils asked themselves that question when Scotland's chief medical officer was quoted as saying that the recent outbreak of E.coli 0157 could be traced to a specific butcher's shop. Supposing they had pondered on the matter, how many would have been able to put forward a reasonable explanation based on their experience of investigational work at school? Would their studies enable them to suggest anything other than a fair test, involving the manipulation of variables; clearly appropriate in this case?
Many schools, both primary and secondary, ask their pupils to carry out investigations which are mostly of a similar type. Typical of the genre are the following: "Which shoe has the best grip?" "How can we make grass grow faster?" "What factors affect the rate of dissolving?" "What is the relationship between temperature and the number of seeds that germinate?"
All these investigations share a common structure. They require pupils to change one variable (for example, the kind of shoe) and measure the effect (for example, the force needed to pull the shoe over a surface) while keeping other variables the same. To do this, pupils have to identify the important variables, decide on an observation or measuremen t strategy and collect, interpret and evaluate data in order to identify the best shoe, or to identify the effect of one variable on another.
This kind of investigation is only one of a range of strategies that scientists use. Others, that are less typical in schools, are identification investigations, descriptive investigations, investigations which involve survey data and investigations which test scientific models.
Identification investigations are quite different in structure from "fair tests". Pupils are asked to identify something unknown, such as a chemical,a bird, or a faulty component in an electrical circuit. Students have to use their knowledge and understanding of science in order to be able to select appropriate strategies. For example, in determining the fault in an electrical circuit, pupils have to decide what they should check and set about their task in an orderly way. Their list might include checking if the battery is flat, that all the connections are making contact, and whether or not the filament in the bulb is broken or a motor jammed .
If pupils can apply these strategies and explain why they are useful, they will be demonstrating an understanding of the need for a complete circuit and that some materials are better electrical conductors than others.
In this kind of investigations pupils must select procedures which discriminate well and can generate a unique set of results in order to identify the chemical, bird or faulty component.
They must then apply the tests and interpret them in the light of their scientific knowledge and understanding.
A third example of investigations uncommon in the science class is of those that involve looking at trends and patterns in a variety of different kinds of data, involving survey data. This sort of work tends to crop up when pupils are learning about humans as organisms and often involves secondary evidence.
Pupils might suggest differences in the risk factors related to heart disease and check their predictions against recent figures, or trace an outbreak of cholera to its source, given maps and other information relating to the outbreak of the disease.
This last example might stand them in good stead when pondering on the chief medical officer's statement mentioned at the start of this article. These investigations would give pupils an indication of the way that medical scientists work when studying epidemics.
How often do you ask pupils to consider evidence and how can it be used to evaluate different scientific models? Here is an example which considers copper burning. Copper turns black on heating in air. Suppose you then asked pupils "What do you think happened to the copper when it burnt?" They might reply: "The copper changes to carbon. I think the carbon weighs more because it is now a different material."
If pressed to say what part, if any, air played in the process, they might respond by saying: "You need air for burning. Some of it gets used up and that's why the volume of air gets less."
The scientific model is, of course, quite different: copper combines with oxygen from the air to form copper oxide, which weighs more because of the added weight of the oxygen. The volume of air is less because oxygen has combined with the copper.
How could these two models be evaluated? Evaluation of scientific models involves identifying what counts as evidence, collecting evidence (including the use of secondary sources), critically evaluating the evidence that has been collected, and examining different inferences that may be drawn from the data in the light of the different models. The models can also be used to make predictions and evidence collected to test them.
While this list of different types of investigative work is clearly not exhaustive, it offers a greater variety than is often seen in schools. What should be the balance between these different types? What is the influence of the national curriculum in determining the range of investigative approaches used in schools? How can we teach pupils to select and use appropriate strategies and become more effective investigators in a range of investigative contexts?
These and many other questions will be the ones that we will be tackling in a new three-year project considering investigative work in key stages 2 and 3. The project is a joint venture between the Association for Science Education and King's College London, funded by the Wellcome Trust, called the ASE-King's Science Investigations in Schools (AKSIS) Project. We are looking forward to investigating!
For further details contact: Dr R J Watson, AKSIS Project director, School of Education, King's College, London, Cornwall House, Waterloo Road,London SE1 8WA. Fax: 0181 872 3184. E-mail: firstname.lastname@example.org
Anne Goldsworthy and Rod Watson are members of the AKSIS Project team, with Valerie Wood-Robinson