An experimental group consists of a representative sample of a population under study that the researcher submits to the influence of a variable that is under his control. The purpose of the experiment is to determine the effect of this variable, called the independent variable , on one or more response variables called the dependent variables . Experimental groups are also called treatment groups, especially in the field of medicine and pharmacology.
On the other hand, the control group consists of a sample very similar to the experimental group, but which is not subject to the influence of the independent variable. The latter either remains constant in the control group (as is the case with variables such as temperature or pressure), or is a factor that does not apply at all (as in the case of a medication). Under these conditions, any change in the dependent variable in the control group cannot be attributed to the independent variable, but to other intervening variables.
controlled experiments
Not all experiments require the use of a control group. That depends on the intentions of the researcher, the nature of the experiment, and the complexity of the system being studied. An experiment in which a control group is used is called a “controlled” experiment .
Differences and similarities between the control group and the experimental group
differences similarities • The experimental group is subjected to the influence of the independent variable while the control group is not.
•The changes observed in the control group are directly attributed to variables other than the independent one, while, in the case of the experimental group, it must first be compared with the control to establish the cause-effect relationship.
•The experimental groups are essential to carry out an experiment, while the control groups are not always necessary.
•The experimental group gives meaning to the experiment while the control group gives reliability to the results. •Both depend on the experimental design and the hypothesis that the researcher wishes to test.
•Both are made up of subjects or study units from the same population.
• Both the control group and the experimental group must be representative of the population under study.
•Both are randomly selected to ensure the applicability of the statistical analysis of the results.
• Generally they are selected from the same initial sample, which is divided into two to give rise to both groups.
• Except for the independent variable, both groups are subjected to the same experimental conditions.
•It is assumed that both groups respond in the same way to any variation in the experimental conditions, whether this variation is intentional or not.
What are control groups used for?
Controlled experiments are carried out whenever the system under study is very complex and there are more variables than the researcher can control and keep fixed. Subjecting the experimental and control groups to the same conditions, except for the independent variable, ensures that any difference between the two groups is attributable to the independent variable. Thus, the cause-effect relationship can be established with greater certainty, which is the ultimate goal of all experiments.
Placebos and Control Groups
In some experiments, just being part of the control group or the experimental group can influence the response of the independent variable. This is the case of the placebo effect, which in clinical drug trials consists of an improvement that occurs in the body when taking an inert substance, but with the conviction that an effective drug is being received , when in reality it is not. So. To avoid the influence of this new variable (which is only relevant to us humans), in clinical studies members of the control group are given a “placebo” that looks, smells, and tastes the same as the real drug. , but without the active ingredient.
In these cases, none of the participants are told which group they belong to, so they take the drug or placebo “blindly,” which is why these studies are called ” blind” studies . In some cases, to avoid unintentional investigator bias, the investigator will also not know who received the placebo and who did not. Since neither the participants nor the investigator know who received the placebo, this kind of study is called “double-blind . “
Positive and Negative Controls
When an experiment has only two possible outcomes, control groups can be of two types:
positive control groups
They are those that, from experience, are known to give a positive result. They serve to prevent false negatives, since if the control group gives a negative result, knowing that it should be positive, instead of being attributed to the independent variable, it is attributed to an experimental error and the experiment is repeated.
Example:
If a new antibiotic is tested on a culture of bacteria and one known to be effective against the bacteria is used as a control, the results will only make sense if the control is positive (the bacteria do not grow on the control). If this does not happen, there may be a problem with the experiment (perhaps the researcher used the wrong bacteria).
Negative control groups
They are control groups in which the conditions ensure a negative result. As long as the result in the control group is negative, it is assumed that no variable is affecting the results, so a positive result in the experimental group can be considered a truly positive result.
Example:
The placebo group is an example of a negative control. The placebo is not supposed to have any effect on the disease (which is why it is a negative control) so if both the placebo and the experimental group show improvement, it is probably some other variable that is confounding the results and not of a true positive. Conversely, if the placebo is negative (as expected) and the experimental group shows improvement, then this is attributed to the study drug.
The selection of the control group and the experimental group
Proper selection of the control group and experimental group begins with the selection of a large random sample that is representative of the population. For example, if you want to study the effect of noise on the grades obtained by students in a test, the sample must be made up of students, and the chosen group must have, on average, the same characteristics as this population.
The next step is to divide this initial sample into two groups that are as similar as possible. It is always a question that any variable that is suspected of influencing the results (such as sex, age, ethnicity, educational level, etc.) is equally represented in both groups.
Then, it is attempted that both groups are subjected to the same experimental conditions. In the example of the students, it would be that all dedicate the same hours to the study of the subject, that they attend the same classes and that they receive the same guidance. At the time of the examination, both groups should receive exactly the same test, possibly at the same time and in similar rooms, but in one of the rooms (in the one of the experimental group) anything that produces a lot of noise is organized, while in the other, where the control group is located, does not.
Examples of control groups and experimental groups
Whenever you want to talk about specific examples of a control group and an experimental group, you must first describe the experiment in question and establish which are the dependent and independent variables. Let’s see the following example:
- Experiment: It is desired to determine the influence of the frequency of bathing on the shine of the coat of the Yorkshire Terrier breed of dogs.
- Independent variable: Bath frequency.
- Dependent variable: Yorkshire Terrier coat shine
Example of Experimental Group Example of a good Control Group They are not good control groups… ✔️ Group of 20 male and 20 female Yorkshire Terriers between 1 and 3 years of age who are bathed between 1 and 5 times a week for a period of one month. ✔️ Group of 10 male Yorkshire Terriers and 10 females between 1 and 3 years of age that are bathed only at the beginning of the experiment. ❌ Group of 20 male Yorkshire Terriers between 1 and 3 years of age who are bathed between 1 and 5 times a week for a period of one month.
❌ Group of 10 male Yorkshire Terriers and 10 female Golden Retrievers less than 1 year of age, bathed only at the beginning of the experiment.
❌ Group of 20 Persian cats between 1 and 3 years of age that are bathed only at the beginning of the experiment.
The three examples of poor control groups highlight the differences and similarities between the experimental group and the control. In the first case, both the experimental and control groups are subjected to the same variation of the independent variable (bathing frequency) and differ in other variables that should remain constant (sex).
The second example is not convenient either, since it introduces new variables (breed and age) and, furthermore, Golden Retrievers are not representative of the population to be studied, made up exclusively of Yorkshire Terriers. The same can be said of the last example, in which the group does not even consist of the same species of animals, despite the fact that the experimental conditions to which the group is subjected are adequate.
Sources
- Bailey, R.A. (2008). Design of Comparative Experiments . Cambridge University Press. ISBN 978-0-521-68357-9.
- Chaplin, S. (2006). “The placebo response: an important part of treatment”. Prescribe : 16–22. doi: 10.1002/psb.344
- Hinkelmann, Klaus; Kempthorne, Oscar (2008). Design and Analysis of Experiments, Volume I: Introduction to Experimental Design (2nd ed.). wiley. ISBN 978-0-471-72756-9.
