The Experiment Variables
Our experiment involves taking
sheets of glass and coating them with sunscreen lotion of different SPF
ratings, then placing the coated sheets of glass under bright sunlight and
taking a UV reading behind the coated glass.
In our example, you would set up several
identical experiments (called "controls") with only the SPF rating
changed. The other variables for the experiment, such as the brand of the
sunscreen lotion, the quantity of sunscreen lotion used for each experiment,
the type of materials and the type of UV meter used for each experiment, and the
temperature, lighting conditions and humidity of the place where the experiment
is conducted, must be kept the same for every experiment. You have to perform
the experiment within a small window of time, under the hot sun and when there
are no clouds. You should also make sure to purchase all of the lotion from the
same pharmacy, and to purchase the lotion at the same time (just in case
different batches of the same brand of lotion are manufactured with different
specifications). The same UV meter should be used for each control experiment,
as different meters (even though they are of the same brand/model) may produce
different results due to many possible reasons (for example, the batteries may
have run out in one UV meter, but not the other).
All of these variables are called control
variables. Notice that when you're designing the procedure for your project,
you must include steps for measuring the results of all of the experiments.
Watch out!
All scientific theory must be substantiated
by reproducible test results. These tests must be reproducible by other people.
That is why it is so important for you to keep good, systematic records of your
experiment!
As you perform your experiment, you need to
watch out for the following types of errors or mistakes:
1. Errors in the measuring instruments that you are using.
Because this type of error will always produce a measurement that is higher or lower
than the "true" value, it is called random error.
2. Systematic errors (also known as non-random errors), which are due to factors which bias
the results of your experiment in one direction.
As you design your experiment
procedure, you need to identify the independent variable, and how to measure
how the changes in the independent variable affects the dependent
variable. Bear in mind that there can
only be one independent variable. All other variables must remain constant.
This is the ensure that any changes in
the dependent variable can be properly attributed to changes in the independent
variable. In our example, we need to
ascertain how changes in SPF ratings affect the level of protection. If you
don't keep all other factors constant during the experiment, other than the SPF
level (for example, if one control experiment was done under bright sunlight
whilst another is done late in the evening during sunset), you won't be in the
position to say that the results which you observed were caused by the changes
in SPF levels. Your procedure should also specify how many trials are to be
performed. At a minimum, you should repeat your experiment three times,
although there's nothing to prevent you from repeating more. For experiments
that involve surveys, there's no sense in asking the same person the same
question more than once - however, for this type of experiment, you should
ensure that you survey a large number of participants for more reliable
results.
