An important concept in experiments of many scientific disciplines is the 'control'. This may involve a 'control group' of subjects, a 'control condition', or other controls depending on research method and topic.
At its core, the 'control' always serves as a comparator. Take as an example an experiment that seeks to use fMRI to determine which area of the human brain is responsible for processing visual motion.
In this experiment, the researcher will be able to present moving dots (the 'motion stimuli'). The brain activity will show that many areas of the visual system, both in the what and the where pathway, become active because of these motion stimuli. However, to conclude that all these areas are involved in the processing of motion, a 'control condition' is needed, in which 'control stimuli' are presented.
To do this, the researcher presents pictures of stationary dots (control stimuli) to the subject in the fMRI scanner. These images are thus very similar to the motion stimuli: the only difference between the motion stimuli and the control stimuli is the movement of the dots. Comparing the brain activity during seeing moving dots (‘experimental condition’) with the brain activity during seeing stationary dots (‘control condition’) reveals the area of the brain involved in processing motion information.
This method, in which the activity we are really interested in is compared with the activity in a control condition that looks like it but lacks the crucial aspect (here: visual motion), is called the contrast method. It forms the basis of a lot of research with almost all research methods (fMRI, EEG, behavioral, etc). Because you are basically 'subtracting' the activity in the control condition from the activity in the movement condition, to leave only the true movement activity, this is also called the 'subtraction method'.
Control conditions do not necessarily require control stimuli. For example, a researcher may be interested in the brain region involved in counting. The subject may count to one hundred during an fMRI measurement (experimental condition) or perform several other tasks such as reciting the alphabet (control condition 1), naming the same number over and over (control condition 2), or doing nothing at all (control condition 3).
Control groups are found, for example, in experiments that want to compare two types of subjects. For example, elderly people versus adults, or depressed patients versus healthy people. Suppose a researcher wants to know if blind people hear better than non-blind people. The hearing task will be the same for both groups, but the performance of blind subjects (‘experimental group’) will be compared with the performance of the non-blind (‘control group’).
But you'll also find control groups in experiments where it's just hard to measure the same people multiple times. For example, in research on a drug for depression. Everyone knows the phenomenon of the 'placebo effect'. To really know whether a new drug for depression works, a researcher divides a large group of depressed patients into two groups: an experimental group (they get the new drug), and a control group (they get the fake drug: the placebo). The difference between the two groups will reveal how well the drug works.
Author: Tom de Graaf (translated by Melanie Smekal)
Image: Marcel Loeffen