Discussion: Spurious Correlations and Extraneous Variables
Correlational research focuses on relationships between variables, but does not indicate that one variable is responsible for another variable.
Statistically significant correlation coefficients simply indicate that there is some relationship between a predictor variable, and an outcome variable.
Correlational research is when there is a statistically significant relationship coefficient between two variables. You want to confirm that this relation exists.
This is difficult because you may not be able to study other variables, which could complicate the study or the interpretation of the results.
Sometimes, a spurious relationship may be found in which one common causal factor, also known as a “third variable”, is responsible for the observed relation between predictor and outcome variables.
Imagine seeing a news report about the results of a study showing that children younger than 17 years old who saw R-rated movies were more likely to develop a smoking habit.
Permissive parenting could be a third factor that can explain both the outcome variables and the predictors.
Permissive parents could allow their children to watch R-rated films if they’re under 17.
Permissive parents may allow children to see R-rated movies if they are younger than 17.
The movie watching does not necessarily cause the children to smoke.
It is possible that the permissive parenting led to children’s movie-watching and their subsequent smoking habit.
Other variables may be identified that might affect the outcome variable. But, unlike the spurious correlation described above, these variables are not related to or influence any predictor variables.
This Discussion focuses primarily on spurious relations or extraneous variables.
After studying the examples provided in the course text you will be able to find your own examples from the media and explain how these might affect the relationships among the variables.
Spurious correlation is a relationship between two variables, which is not due to any relationship between the variables but its relation with additional variables.
However, extraneous variable is a variable that has an impact on the outcome of an experimental other than the dependent or independent variable (Nugent 2017,).
Extraneous variable are variables that can lead to an error in an experiment (T Wilson & Shuttleworth 2017, 2017).
One student had been to a party the weekend before and, as he sat in the bar looking at his friends, noticed that those who had the most fun dancing were also most likely not to get up at the end of the night. (“Spurious correlation explained with examples”, 2017).
In this example you can see the independent, dependent, extraneous, and spurious variables.
The students are not prevented from dancing if they get sick.
The extraneous variable that alcohol consumption has is called “extraneous”.
Consuming alcohol leads to throwing up and dancing.
If students have fun dancing, it must also be associated with being sick.
This implies that students can’t expect to be sick after having fun at a party (independent variable) or being sick only after having had fun at a party.
Another factor that may allow one to draw inferences from the associated variable relation (stangor (2015) is the temporal relationship).
Imagine that having fun happens before you get sick. Then, it is possible to conclude that having fun causes being sick.
But if you are sick before you have fun dancing at a dance party, then it is not possible to create sickness in a student.
It is not possible for a student to get sick if they only dance at the party.
Predictable variables are often used to test the outcomes of dependent variables.
The dependent variable on the other side is variable that’s put to experimental testing.
The influence of the predictor variable is measured and recorded as the researcher alters it.
In the example above, the researcher wants to know if students’ dancing at parties has an impact on their health.
The researcher will observe the students who had the most fun dancing.
This is the predictor variable.
This is the outcome variable. It’s how students feel about it.
Some extra variables are involved in the relationship between predictor variable (and outcome variable) and student.
The outcome of getting sick is caused by an extra variable.
Extraneous variables include: 1. Students may attend a party, and then become ill from being exposed to cold for a long time. 2. Students could get flue or other symptoms.
Assume that the student has health problems before they attend the party.
Third, let’s say the researcher selected students who are healthy to be at the party. This will manipulate the outcome.
Let’s say that the student had consumed too much alcohol during the party, resulting in a hangover the next day.
There is a connection between spurious variable and predictor variables. First, consider that the student attended the party, did not drink alcohol, but danced all night. This could have a different result.
Second, students who were present at the party and consumed alcohol of different types, but did not dance, could make it worse.
The student could also get sick if they have had a medical condition, such an ulcer, or if they have ingested alcohol that contains methanol.
A student might have been sickened by drinking alcohol, using other drugs, and dancing at the party.
What is EXTRANEOUS VOARIABLE?
With examples, spurious correlation is explained.
T Wilson, L., Shuttleworth M. (2017).
Confounding Variable/Third Varable. Retrieved November 1, 2017 from Explorable.com.
The fifth edition of research methods in behavioral sciences.