What is the purpose of control in a experiment?

A control experiment is a very important aspect of the scientific method.

Control experiments are used to minimize influences such as bias, genetic and biological variations, environmental changes, and random skewing.

In a controlled experiment one set of samples have been (or is believed to be) modified and the other set of samples are either expected to show no change (negative control) or expected to show a definite change (positive control).

Control experiments are especially useful in testing the effectiveness of new pharmaceuticals.

For example:

Let's say a company is developing an anti-smoking aid - which we'll call 'Cigstop' - and it has reached the human testing stage. To judge the effectiveness of the drug a control experiment would be administered to two groups.

The first group would receive Cigstop in its prescribed format, while the second group (the control) would receive a placebo believing it to be Cigstop.

Both groups are then instructed to attempt to quit smoking over a set period of time.

Once the experiment has concluded one can compare the two groups and determine the success rate of the drug. If the group that actually received Cigstop had a higher percentage of quitters than the control, then the drug shows an ability to assist someone in quitting smoking. Multiple control experiments showing the same results would verify the statistical probability of the drugs effectiveness.

Experiments such as the one above conducted without a control group could lead to unverified claims, and may be regarded as psuedo-science.