Scientific method

Science aims to develop a good explanation of natural events (phenomena) that are observed.

Scientists develop explanations using the scientific method. This follows a specific way of working:

  • propose a hypothesis, a tentative explanation of an observation
  • make a prediction based on the hypothesis, suggesting how changing one factor will affect the outcome
  • plan an experiment - this should test the prediction and the hypothesis on which it is based

For example, a student may be considering the solar panels on a house and may hypothesise that "the voltage obtained from a solar panel is affected by the size of the solar panel."

Based on this hypothesis, the student may make the prediction: “as the size of a solar panel increases the size of the voltage produced will also increase.”

A prediction can be improved by adding the scientific reasoning behind the prediction. In this case it could be "as the size of a solar panel increases the size of the voltage produced will also increase because they will be able to absorb more energy from the light source."

Planning experiments

Always think carefully before starting an investigation. It is usual to write a plan, which should follow a logical sequence of steps. The plan should be capable of being used by another scientist using similar apparatus to produce results that are reproducible. It should specify the apparatus used, the experimental technique the range and number of readings.

It is important to choose appropriate apparatus, materials and techniques.

Solar panels used on houses are made from a number of individual photovoltaic (PV) cells.

In the laboratory, it is unrealistic to have a selection of different sized solar panels.

A PV cell is connected to a voltmeter. There is a ruler next to the cell and a light source above it.



  • Measure the voltage from the photovoltaic cell when it is a set distance from the light source.
  • Repeat the experiment with different sized PV cells.
  • Choose a range of different sized PV cells that is sufficient to give a valid conclusion. If the measurements are too close together, it is difficult to have confidence in any trend shown.
  • Plot a graph of voltage against area of PV cell.


When planning a practical investigation to test a prediction it is important that the equipment that is chosen will lead to results that can be considered valid, precise and accurate.


A measurement is valid if it measures what it is supposed to be measuring, and if the measurements made have been due to the cause claimed.

So going back to the solar panel experiment example, if the student used an ammeter to measure the voltage – the results would be invalid. This is because the student measured the wrong thing, and this shows why it is important to use the appropriate apparatus and techniques during an investigation.


A student is investigating if the size of a solar panel increases the size of the voltage produced. The students measure how much voltage is produced by a 50 mm × 25 mm photovoltaic cell. The student then measures how much voltage is produced by a larger cell the following day, when the Sun is much brighter.

Why is any conclusion from this investigation invalid?

The conclusion would be invalid because the Sun is much brighter on the second day. The increase in voltage could be a result of the better weather conditions and not just the increased size of the solar panel.

Results are only valid if the measurements made have been affected by only a single independent variable. Results are invalid if other variables have not been controlled.

Precision and accuracy

Measurements are precise if the results are very close in value.

The accuracy of a numerical result is how close it is to the true value.

Results are precise if they are close in value, but they must also be close to the true value to be accurate.
Four targets each have an arrangement of dots to represent 1, accurate and precise, 2, accurate but imprecise, 3, inaccurate but precise, 4, inaccurate and imprecise.

It is usual to repeat measurements. This is to improve the reliability of an experiment. If three or four repeated readings are precise, then the average of those readings is reliable.

The choice of measuring apparatus can affect the precision of data collected.

For example, the output from a small PV cell may be less than 0.5 V, so using a 0-1 V voltmeter will produce more precise readings than using a 0-10 V voltmeter.