When attempting to create a relationship between two variables, it is best to discover which of the variables affects the other. If by the end of an experiment you have discovered which is the dependent variable and which is the independent variable, you will have created a much more valid study than one which simply finds a connection, as you can then start investigating to what extent the IV affects the DV.
Yet what if you have indeed found a connection, yet the methods you used to imply that connection ended being the reason it occured? In 2006, Fillmore et al. conducted a meta-analysis of 54 studies looking at moderate alcohol use and if it had an effect on a person’s health. The studies all seemed to indicate that a moderate use of alcohol could give a person a healthier heart, yet Fillmore found that many of the studies (47 of them) hadn’t randomly divided the participants into groups of drinkers and non-drinkers. Instead, it was a comparison between people who drank regularly and people who couldn’t drink because they were either a) old or b) dying. Now we know why the drinkers had healthier hearts. Not because of drinking, but because they were not ill or too old. So by not using the correct method, the studies found a connection that was in fact not there. Random assignment would have shown that this connection did not exist, any other assignment could have left the bias intact.
Random assignment ensures that participants in a cause and effect study are unbiased as it prevents people’s history from causing an extraneous variable within the experiment. Only for ethical reasons should it be changed; many of the studies could not have used this as it means they would have had to convince non-drinkers to drink. Many of the teetotallers had their own reasons for not drinking alcohol, meaning that the scientists would have had to either force them to drink (highly unethical) or drop them from the study, leaving them with just drinkers who they would have had to convince not to drink, this dictation of a way of life could again be highly unethical. So we can see just how difficult it is to use random assignment in some cases, yet in others experiments wherein the participants past cannot make a large impact, I consider it to be the best assignment type available.
Interesting alcohol related fact: A brewery tank ruptured in a London Parish in 1814, releasing 3,500 barrels worth of beer, destroying two houses and killing nine people.
Posted in Year One Research Methods | 5 Comments
In a psychology experiment, the experimental group (or experimental condition) refers to the group of participants who are exposed to the independent variable. These participants receive or are exposed to the treatment variable. The data that is collected is then compared to the data from the control group, which did not receive the experimental treatment.
By doing this, researchers are able to see if the independent variable had any impact on the behavior of the participants.
A Closer Look at Experimental Groups
Imagine that you want to do an experiment to determine if listening to music while working out can lead to greater weight loss. After getting together a group of participants, you randomly assign them to one of three groups. One group listens to upbeat music while working out, one group listens to relaxing music, and the third group listens to no music at all. All of the participants work out for the same amount of time and the same number of days each week.
In this experiment, the group of participants listening to no music while working out is the control group. They serve as a baseline with which to compare the performance of the other two groups. The other two groups in the experiment are the experimental groups. They each receive some level of the independent variable, which in this case is listening to music while working out.
In this experiment, you find that the participants who listened to upbeat music experienced the greatest weight loss result, largely because those who listened to this type of music exercised with greater intensity than those in the other two groups.
By comparing the results from your experimental groups with the results of the control group, you can more clearly see the impact of the independent variable.
Some Things to Know
In order to determine the impact of an independent variable, it is important to have at least two different treatment conditions.
This usually involves using a control group that receives no treatment against an experimental group that receives the treatment. However, there can also be a number of different experimental groups in the same experiment.
So how do researchers determine who is in the control group and who is in the experimental group? In an ideal situation, the researchers would use random assignment to place participants in groups. In random assignment, each individual stands an equal shot at being assigned to either group. Participants might be randomly assigned using methods such as a coin flip or a number draw. By using random assignment, researchers can help ensure that the groups are not unfairly stacked with people who share characteristics that might unfairly skew the results.
A Word From Verywell
Experiments play an important role in the research process and allow psychologists to investigate cause and effect relationships between different variables. Have one or more experimental groups allows researchers to vary different levels of an experimental variable (or variables) and then compare the effects of these changes against a control group. The goal of this experimental manipulation is to gain a better understanding of the different factors that may have an impact on how people think, feel, and act.
Myers, A. & Hansen, C. Experimental psychology. Belmont, CA: Cengage Learning; 2012.
Robbins, P. R. Understanding psychology. Portland, Maine: Walch Publishers; 2003.