Fair testing is something that science teachers often talk
about. While writing a method they ask, “how will you ensure it is a fair
test”? In the conclusion, you must state
if it was a fair test and if not why.
But what does it really mean to have a fair test and is it actually
possible?
Variables
When completing a science experiment, there are three types
of variable we need to consider:
- The independent variable – this is the thing we are changing or testing.
- The dependant variable – this is what we measure to see what changing the independent variable does
- The controlled variables – these are any other factors which could change during the experiment but need to be prevented from changing.
It is the controlled variables which are the most important
when trying to ensure a test is fair.
Example:
In an experiment to investigate the effects of sunlight on
coloured paper (video here), the independent variable was the amount of sunlight the paper was
exposed to. The dependant variable was how much the colour changed or faded.
The controlled variables were; using the same type of paper, using the same
size of paper and leaving the paper for the same amount of time before making
observations.
If any of the controlled variables were not kept the same
for each trial it would not have been a fair test. For example, if the paper that had been left
in the sunlight remained out for a week, but the paper that was kept inside was
only checked after an hour it would not be fair. There is a chance that the
paper inside would have changed somewhat, if it had been left for the same
amount of time.
Sometimes, especially when doing experiments at home, it can
be very difficult to keep the controlled variables the same. Have a look at this video and see if you can figure out why it
is not a fair test.
Errors
No matter how carefully you conduct your experiment, there
will always be a small amount of error.
There are three different types of error:
- Random Error
- Systematic Error
- Human Error
A random error is just that, random. It is something that cannot be predicted or
controlled. For example, when conducting
the insulation experiment the room could have heated up between the trials. This
would have made the last material appear to be a better insulator than it was,
as the temperature change would not have been as great.
A systematic error is one which will make all your results
wrong by the same amount. For example, if the thermometer was not calibrated
correctly then all the readings in the insulation investigation could have been
off.
Reading the meniscus, Wikipedia Commons, accessed 21/04/2020
A human error is one caused by human misreading or
miscalculating. In the insulation investigation this could have been when
measuring the volume of water for example – was it always up to the same line? Not measuring at eye level and measuring at
the meniscus is a common human error.
However, just because a small amount of error cannot be
completely avoided, it is vital to try to minimise the errors as much as
possible. Although a test may never
truly be a fair test – we still have to give it a fair go!