Rough Science
About the TV Series
Discover More
Site Map
Discover More

The Art of Science

Scientists need to develop a variety of skills. The following includes what we consider to be some of the more important.


The process of classification helps to make sense of an apparently disparate group of things by lumping them all together based on certain criteria. For example, rocks are usually classified according to their texture and appearance, plants and animals are grouped by species, stars by constellation and color, and chemical elements according to their reactions.

One of the important strengths of classification is that it can be used to predict previously unknown facts or observations; for instance, based on its supposed position in the periodic table, the properties of the element scandium were predicted long before the metal was actually isolated.

Formulating a Hypothesis

A hypothesis usually takes the form: 'I think something does this because...'. For example, suppose I see a piece of wood floating on water. From this I could propose the hypothesis: 'I think an object floats on a liquid because the material the object is made of is less dense than the liquid.' This is not a law; it is merely my suggestion (based on an observation) for why something behaves the way it does. To become a 'scientific law', a hypothesis requires rigorous testing, to confirm that it holds up in all possible situations and fits in with all the observations I make to test it. To test my hypothesis concerning floating wood, I would design an experiment in which I would try to float a range of materials that are more, or less, dense than water, and see whether my original hypothesis holds in all cases. I could then repeat the experiment, but this time using a range of liquids with different densities.

If all of my experimental observations still fit in with the original hypothesis it can stand, but if not, it will need to be revised. For example, if I take a model boat made of iron and float it on water, the boat still floats, even though the material that the boat is made of is denser than water. The problem is that when I put forward my original hypothesis I was only thinking about solid objects, but the iron boat floats because it is hollow. As a result of this new observation, I need to adjust my hypothesis to: 'I think a solid object floats on a liquid because the material the object is made of is less dense than the liquid.'

Mike Bullivant experimenting awayBy continually testing (by experiment) and modifying a hypothesis in this way, we refine it so that it has a wider applicability. Sometimes the experimental observations we make will contradict our hypothesis, so that no amount of tinkering with the words will resolve the problem. In such instances, the hypothesis must be completely rejected and replaced by a new, more universally applicable version.


This activity is at the heart of science. In trying to explain why things occur the way they do, we put forward a hypothesis that must be tested to destruction by experimentation. However, we must be sure that the experiment does in fact test what we think it tests. We must also be able to understand and interpret the outcome of the experiment.

There are various stages associated with experimentation:

It is important to design the experiment in a way that truly tests the hypothesis and does not give you a particular result because of some artefact. For example, the density of a material varies with the temperature, and the size of this variation depends upon the nature of the material. So, if I made no attempt to carry out the experiment at the temperature at which the densities were measured I might get some odd results.

Carrying out the Experiment
Once I have decided the details of the experimental procedure, it is important that it is carried out carefully. Many strange results have been obtained because of poor experimental technique!

Although it may seem obvious, observing what happens during an experiment is one of the most important scientific skills. Quite often, unexpected developments occur, and it is important to be able to recognize when this happens. Many significant scientific discoveries have been made by accident, and it is a great scientist who can recognise and capitalize on anomalous experimental observations.

It may seem easy to measure if a circular plate is 24 or 25cm wide, but what if I wanted to measure the value more precisely? Using a ruler, I could probably measure it to the nearest millimeter, but more accurate than that and it gets a bit tricky. There are two problems: i) the plate is circular, so working out the true diameter will be difficult because there's nowhere to measure between; and ii) to measure fractions of a millimeter requires special measuring devices. It's therefore important to recognize the limitations of the measuring device, and to ensure that my measurements are sufficiently precise for the needs of the experiment.

Recording the Results
It is all very well to carry out the experiment and observe and measure what happens, but unless I record my observations carefully, and at the time they're made, they will not be of much use. A practical notebook doesn't have to be neat, but it does need to be complete, readable and capable of being understood by other readers.

Assessing Uncertainties
When referring to measuring the diameter of the circular plate above, I mentioned that I could measure it to the nearest millimeter but after that I would need specialized instruments. Suppose I measured the diameter of the plate and it was between 24.3 and 24.4cm, but closer to 24.3cm. I would record the observed value as 24.3cm, but Id be aware that it could be 24.31, 24.32, 24.33 or even 24.34cm. Thus, there is an uncertainty in my measurements which I must bear in mind when I interpret my results. This becomes particularly important when I do calculations with my measurements, because the individual uncertainties in each of them can accumulate, to give completely the wrong answer. The danger is that I could end up with a result which I believe to be correct, but which in fact contains a lot of uncertainty. In such cases, I should treat my results with extreme caution!

Once I have carried out my experiment and recorded my observations and measurements, the next stage is to analyze the results. This could involve comparing the observations with those from previous experiments or doing a calculation. When I designed the experiment I had a particular hypothesis that I wanted to test, and the analysis should involve testing whether or not the hypothesis has held up to scrutiny.

Critical Interpretation
Once I have a result, the next stage is to question its validity. This may seem a very negative move, but it is crucial that I re-examine everything I did, just in case my results are invalidated by something I overlooked. This is possibly the most difficult task for many scientists because there's a natural tendency to become emotionally involved with what they are doing and to start looking for a particular result. The critical nature of scientific study, by which one's results are rigorously evaluated by one's peers, ensures that if the individual scientist doesn't recognise her/his weakness then at least their colleagues will!

Individual scientists are part of a wider community and it is important that I am able to communicate my results to other scientists. In this way, they will have the opportunity to comment on and perhaps benefit from my work. Scientists have their own way of communicating that often appears very dense and impersonal, but you get used to it and it does help to ensure that results are communicated dispassionately and accurately. Scientists also have a responsibility to communicate what they do to society at large. In recent years, an inability to do this has led to some understandable distrust of some parts of the scientific community.

Further Reading

Here are some more general books and articles that you may want to try and get hold of:

Barrow J. D., The Artful Universe, Oxford University Press, 1995 ISBN 0 1985 3996 7. A quite remarkable book that will change the way you view the world. Extremely accessible.

Burton et al., Chemical Storylines, G. Heinemann Educational Publishers, 1994 ISBN 0 435 63106 3. Part of the Salters Advanced Chemistry course, which explores the frontiers of research and the applications of contemporary chemistry. For A level and other science courses aimed at 16 to 19-year olds.

Fraser A. and Gilchrist I., Starting Science (Book 1), Oxford University Press, 1998 ISBN 0 19 914235 1. Part of an integrated science course for the [U.K.] National Curriculum Key Stage 3 and Scottish Environmental Studies (science) for S1 and S2.

Northedge A. et al., The Sciences Good Study Guide, The Open University, 1997 ISBN 0 7492 3411 3. Indispensable for students of science, technology, mathematics and engineering. Packed with practical exercises and activities, all aimed at making studying more enjoyable and rewarding. Lots of hints and tips for those returning to study.

Selinger B., Chemistry in the Marketplace, 5th edn., Harcourt Brace, 1998 ISBN 0 7295 3300 X. An excellent and informative reference source for all kinds of real-life applications of chemistry. Explores the world of chemistry that surrounds us in our daily lives, explained in terms that everyone can understand. 'Makes chemistry come alive.'

PS547 Chemistry for Science Teachers course materials, The Open University, 1992 A course designed for use by science teachers from a wide variety of backgrounds, with varying experience of teaching science. A familiarity with some basic science (perhaps physics or biology) is assumed, but little understanding of chemistry is required. The mathematical understanding needed for the course is not great.