The notion of “evidence” is a vitally important one for decision-making in a democracy. People continually need to “weigh up” the evidence, explore it, assess it and make judgments based on it. Recent, well-publicised controversies such as “mad cow” disease have shown the need for all citizens to look critically at sources of evidence. The days of certainty, proof, and simple causality (the sacred cows of science) have long gone. Yet formal education has failed to recognise this post-modern shift.

The notion of evidence has particular relevance to the science national curriculum. The programmes of study talk of relating scientific ideas to their evidence, “empirical evidence” and “different ways of interpreting such evidence”’. The Sc1 component (experimental and investigative science) is based on the stages of obtaining evidence, analysing evidence and evaluating evidence. The use of evidence is also an element of the core skill of communication in that students are expected to construct an argument from data, to present data and to turn it into evidence. If teachers are to teach children these ideas then they themselves need a clear understanding of the term evidence (as opposed to, say, “data” or “results”).

Our own research into teachers’ views on the nature of science indicates that teachers may be able to evaluate evidence, analyse it, question its source, assess it and so on but these activities do not play a significant part in the science teacher’s repertoire of teaching and learning activities.

The demands of the national curriculum entail teachers presenting a very narrow scientistic view of evidence and their approach to practical work exhibits this view. In particular, research on teachers’ reactions to critical incidents in the classroom shows that their main concerns can be to produce results which conform to the standard scientific explanation and will even fabricate evidence to ensure this happens.

Turning data into evidence is seen as self-evident, which it is not. The current science curriculum still stresses ideas such as verification, proof, causal links and clear, indisputable evidence - the exact opposite of the “messy world” of scientific disputes. In fact, when the scientific explanation is in dispute then so is the evidence and resolving the dispute is about agreeing on the evidence as well as the explanation. If teachers are to help children to weigh evidence then present classroom practice will have to change.

Beyond formal education, it is also important that society in general understands the notion of evidence. Citizens can only make informed decisions if they have the skills and the ability to balance scepticism and trust needed to assess, weigh up, interpret, judge, evaluate and act on evidence.

Science is no longer (it ever was) able to prove definite links, or identify clear, causal chains. Decisions can no longer be made on the basis that X is, or is not, the cause of Y - whether X is car exhaust gas or radiation and Y is asthma or leukaemia.

The term “evidence” is now used extensively in any real world dispute relating to science and in particular in the media. Numerous examples could be quoted but a few will suffice to make the point.

The recent debate about whether proximity to power lines can “cause” cancer (reported in The Times on several occasions in February) has painted the picture as a dispute between “researchers” at Bristol University and the National Radiological Protection Board. The latter accuse Bristol University of having “speculative evidence”. The “researchers” accuse the NRPB as “failing to understand the science”.

In March a similar debate reached the media when a Government committee “concluded” that the Sellafield nuclear plant in Cumbria is “very unlikely” to have caused a leukaemia cluster among local children. The chair of the committee is reported as saying that “the evidence available at present does not convince us” (The Times, March 28).

But the story that has dominated newspapers, radio and television headlines is the dispute over the link between bovine spongiform encephalopathy (BSE, “mad cow” disease) and Creutzfeldt-Jacob diseas (CJD). In the media treatment of this debate (which in itself would make a valuable and extensive study) the following adjectives were used in conjunction with the word “evidence”: clearunclear; scientificunscientific; first-hand second-hand; hardsoft; woolly; objective; expert; independent.

The following verbs were used in connection with either doing something “with”, “to”, “at” or “on” evidence: assessing; weighing up; exploring; looking at; analysing; judging; sifting through; doctoring; interpreting.

Questions being asked of evidence in the media included: Whose evidence is it? Where did it come from? When was it foundpublished? Who paid for it? Is it biased?

Related terms used were: data, results, and risk, likelihood and probability.

It is interesting to note that at the height of the “hysteria”(as Prime Minister John Major called it) over BSE, the editorial in The Times on March 26 stated that “science is no more than an aggregate of the informed opinions of imperfect seekers after elusive truth”.

Formal science education has traditionally tried to deal in certainty. Science practical work, for example, has been used to demonstrate, to verify and to exhibit - pupils’ investigations are mainly about finding out the standard scientific explanation and have to be stage-managed to ensure that they can be done and assessed.

Presenting science in terms of the clean and unproblematic concepts of proof and causality has created false expectations of science which can never be met when citizens (and scientists) are faced with problems such as asthma, mad cow disease, and cancer. The closed world of the scientists’ bench cannot always account for the open system of the real world.

We are proposing that the time is right to create a shift in the way we conceptualise science education - to move it away from just the clean and certain notions to also embrace the messy ideas which characterise real problems (see table below).

The former set of concepts should not be abandoned but need to be seen as one face of science. The latter must come in not as a substitute but as another face of science and be given as much weight.

Students need to be empowered to deal with the messy concepts as well as the clean ones. This will require a new set of skills as a key aim in science education - the skills of weighing up evidence, of questioning where it comes from, of examining the status authority of evidence , of interpreting it and of acting on it.

Past projects and movements in science education such as the excellent SiSCON and SATIS materials and the Science in Society lobby have made inroads into traditional science teaching but few would deny that fundamentally little has changed. The national curriculum has only served to reinforce the status quo and inhibit experiment with teaching. We are advocating a radical change in the curriculum and in teaching strategies. We need a totally new view of practical work in science and the rationale behind it.

Teachers need to examine why they do the things they do in school labs. Practical investigations could be used to introduce students to the reality of messy data and to consider not whether it is right or wrong but how it might contribute to the body of evidence on which judgments are to be made. This may include dealing with fellow students’ evidence and subjecting it to critical scrutiny.

Students and teachers also need to learn how to use and examine second-hand evidence as well as the stage-managed, first-hand evidence in school practicals. This evidence should best come from a range of sources including books (old and new), magazines, newspapers and now multimedia and the Internet. Thus teaching strategies may involve practicals, data analysis, discussion and engineered debates. By using these sources and classroom activities, students may learn how to handle evidence and evaluate its worth.

We are interested in taking these ideas further by researching and developing teaching and learning strategies and curriculum materials. These would enable pupils and teachers to be explicit about judgment, interpretation, risk, probability, faith and belief in evidence. Please let us know if you are interested or have ideas which you have tried yourself, either by letter or e- mail to: j.wellington@sheffield.ac.uk Richard Gott and Phil Johnson are lecturers at the University of Durham; Mick Nott of Sheffield Hallam University and Jerry Wellington are lecturers at the University of Sheffield