Extraneous variables
Extraneous variables are uncontrollable variables that affect the study results. So it is any variable other than the independent variable that affects the outcome of the experiment. Extraneuos variables in research can take the form of characteristics of a person, such as gender, age or environmental factor such as noise, lighting, etc. It is not necessary that the extraneous variables be part of the study.
Types of extraneous variables
The following paragraphs discuss the various types of extraneous variables.
1. Situational variables
Such types of variables are environmental factors affecting the way a subject is behaving in an experiment. For example, lighting, noise, temperature, etc.
2. Participant variables
In participant variables, the personal feelings of participants influence the study results. For example, physical and mental capacities of the individuals, etc.
3. Experimenter variables
Here, the researcher’s behavior influences the study results. For example, consider the biased attitude of the experimenter or researcher.
4. Demand Characteristic Variables
When clues regarding the outcomes of the study are conveyed to the participant, or he or she comes to know about the results of the study, it is called the demand characteristic variable. For example, the items of the questionnaire or the behavior of the researcher can affect the results of the study.
 Extraneous variable examples
The following paragraphs present the examples of extraneous variables.
1. Age
Age is one of the common extraneous variables that may affect the research studies. For example, if a research study examines the effects of a particular diet on the cholesterol level, age may be an extraneous variable that needs to be controlled. There is a chance that the diet will have different effects on different age levels.
2. Gender
Another type of extraneous variable, for example, is gender, which may have an effect on the results of many studies. For example, if the researcher wants to explore the effect of self-efficacy on the academic achievement of students, gender may be an extraneous variable that needs to be controlled. There is a change in gender that may affect the results.
3. Timing
The next extraneous variable example is the timing of the day. For example, if a study intends to explore the effect of a particular drug on alertness, the timing of the day may be an extraneous variable that needs to be controlled.
4. Lighting
A further extraneous variable example is lighting. Bright light has a different effect on visual activities as compared to low light. This may have an effect on the results of the study as an extraneous variable that needs to be controlled.
5. Experimental setting
The fifth type of extraneous variable in research is the experimental setting. It may have an effect on the results of the experimental study. So it will play its role as an extraneous variable. In case the researcher wnats accurate results, he or she will control the said variable.
Purpose of controlling extraneous variables
The extraneous variables are controlled for the purpose of increasing the internal validity of the research study. Hence, it is ensured by the researcher that the results are due to the effect of the independent variable and not the extraneous variable. At the same time, it increases the reliability of the results of the study as well. So if they are not controlled, it may lead to incorrect results and, ultimately, conclusions from the study, which may have negative consequences for the applications of the results.
TECHNIQUES TO CONTROL EXTRANEOUS VARIABLES
Several techniques for controlling for extraneous variables are presented, including:
-
Control Groups
The use of a control group is among the experimental designs’ most crucial procedures. In certain designs, the control group is a distinct group that receives the same therapy as the experimental group, or a different treatment altogether, but is otherwise identical to it. The design of the study ensures that uncontrollable factors will have an equal impact on both groups. Making participants act as their own controls by monitoring them at least twice before and after the therapy is an alternative to using control groups.
-
Random Assignment of Individuals to Treatments
In experimental research, the technique called randomization is employed to account for any participant variations. Researchers anticipate that several variations in the experiences of the participants will be similarly dispersed among the groups by randomly allocating them to treatments.
-
Matching
The researcher ensures that unrelated variables are equally characterized in both experimental and control groups by using a control approach known as matching. For example, if a researcher thinks that gender might have an impact on the dependent variable, she would employ pair-wise matching of participants
Comparing Homogeneous Subgroups
Researchers frequently design studies in which they compare homogeneous subgroups to facilitate the process of selection while still controlling for unimportant characteristics. For instance, a researcher would suggest creating many IQ groups, each with a range of IQ values, in a study where IQ was the irrelevant variable.
Pretesting of Participants
Randomly assigning entire groups to treatments is a common practice in experimental studies, known as quasi-experimental studies. Researchers frequently give both groups a pretest to make sure the groups are comparable to one another. A pretest evaluates the initial parity between the experimental and control groups. It is a test to see if the groups’ talents or other qualities already differ from one another.
One would not be able to conclude that differences observed at the end of the trial are related to the treatment used if there were preexisting differences. A pretest would assist the researcher in figuring out whether the two groups were equal before the study began.
Keeping Extraneous Variables Constant
By keeping certain auxiliary variables constant between the experimental and control groups, they can be managed. The kinds of variables that are kept constant in experimental research in education frequently have to do with the environment or the instructors. You could control for the time of day, the temperature in the classroom, or the ideas addressed by keeping them constant for the treatment and control groups if you were comparing two distinct teaching strategies.
If the two approaches were taught by two different teachers, you may try to control for the teaching quality as an auxiliary variable by mandating that both teachers have the same number of years of classroom practice. The study’s lack of control over teaching quality could be criticized and should be discussed as a study limitation.
Factorial Designs
A factorial design consists of a number of independent variables. A researcher may opt to include a possibly superfluous variable as an additional independent variable in the study after identifying it during the literature evaluation.
Statistical Control of extraneous variables
Measuring and statistically controlling extraneous variables is a popular strategy for dealing with them. Several statistical tests are employed for this aim. Analysis of covariance, often known as ANCOVA, is one statistical technique used as a control mechanism. By using ANCOVA, individuals’ posttest results are statistically corrected for variations in their pretest results. It certainly seems like “massaging of the data,” doesn’t it? That is, in a way, what ANCOVA is.
Assume you’ve chosen to carry out research on the impact of two reading programs—one that emphasizes literature and the other that emphasizes skills—in comparison to one another. You requested permission to use Holly Elementary School from a school district.The researcher uses lessons.
You randomly allocate one class to the literature-based instruction and the other to the skills-based training since random participant selection or assignment is not permitted. But being the wise researcher that you are, you worry that the kids’ reading abilities might differ at first. If the participants’ reading abilities were different at the start of the study, then the experimental treatment could not uncover the differences.
Before the research begins, you decide to test each group, and you discover that one class has some pupils reading at a second-grade level, whereas the other class is essentially reading at a first-grade level. You would employ an ANCOVA instead of giving up on your study due to the variations between the groups! In order to compare the two classes evenly, the ANCOVA approach would statistically eliminate whatever advantage the first class may have had.