Mia is a psychologist who is interested in developing a program for first generation college students. In order to gain baseline data that can be used to create her program, Mia decides to conduct a study. Mia wants to know if a relationship exists between first generation college students' attendance and grade point average. More specifically, she wants to test her hypothesis that attendance can be used to predict grade point average. In this example, attendance is the predictor variable.
A predictor variable is a variable that is being used to predict some other variable or outcome. In the example we just used now, Mia is using attendance as a means to predict another variable, grade point average.
Predictor variable is often confused with independent variable. Though these terms are often used interchangeably, they actually refer to two different concepts. Predictor variable and independent variable are both similar in that they are used to observe how they affect some other variable or outcome. The main difference is that independent variables can be used to determine if one variable is the cause of changes in another, whereas predictor variables cannot.
Independent variables are manipulated by the researcher in such a way that allows the researcher to conclude the independent variable either causes or does not cause the changes in the other variable or outcome. Predictor variables are generally not manipulated by the researcher. The differences and the changes in the predictor variable across subjects are usually naturally occurring. Independent variables are used in true experimental designs; predictor variables are used in nonexperimental research.
Let's look at Mia's study. She is looking at the attendance of the students as it is. She is not going to make students alter their attendance pattern for the sake of her study. Could you imagine how students would respond if Mia were to say I need you to miss thirty days of school, and then report your grade point average to me? This would be highly unethical and would rub many students the wrong way. Mia is going to look at their attendance as it naturally occurs, and then see how their attendance influences their grade point average.
Let's say Mia conducted her study and found that students who attended more days of school had the highest grades. Does this mean attending school causes your grade point average to increase? No, it does not. It simply means that a positive relationship exists between attendance and grade point average and that you can predict someone's grade point average based on their attendance.
Let's say Mia is also interested in adding sports and music to her program. She conducts a study to determine whether participation in sports or music can predict grade point average. Mia asks all 350 first generation college students whether or not they play sports or are involved in music, and then records their grade point average over four semesters. Two predictor variables are illustrated in this example: participation in sports and participation in music. Both variables are being used to predict a single outcome: grade point average.
Mia's colleague Jim is working on an independent project. He wants to know if the level of education can predict income. He asks 300 participants how many years of education they have and their current income. Jim found those with high school diplomas made on average twenty-thousand dollars a year less than those with a bachelor's degree and forty-thousand less than those with a graduate degree. Jim concludes that level of education does indeed predict income. The predictor variable in this example is education level.
Predictor variables are variables that are being used to predict some other variable or outcome. Predictor variables are often confused with independent variables, which are manipulated by the researcher in an experiment. They're often confused, though, because they're both used to see how much (or little) they influence the other dependent variable. Unlike independent variables, predictor variables are generally not manipulated by the researcher, do not indicate that one variable causes another, and are used in nonexperimental designs. For example, you probably wouldn't want to tell someone to deliberately take a pay cut and report their happiness to you; it can only be observed from afar and without interference.
To unlock this lesson you must be a Study.com Member.
Create your account
As you read in the lesson, predictor variables do not show cause-and-effect but show a relationship between variables. For example, wearing short sleeves is a predictor variable for buying ice cream. Does wearing short sleeves cause a person to buy ice cream? No. There is at least one other variable at work. List three reasons why there may be a correlation between the predictor variable of short sleeves and buying ice cream.
It is often the case that the media will portray a predictor variable as an independent variable, creating the impression that a research study shows cause-and-effect when it actually only shows correlation. Watch the news for a few days and look for instances wherein the journalists seem to be confusing independent variables (indicating causality) with predictor variables (indicating correlation). Write a paragraph describing an instance where you have found this confusion.
Imagine that you are interested in discovering the effect of smoking during pregnancy. Why must smoking be a predictor variable rather than an independent variable? What would be the problem with conducting a true experiment regarding smoking during pregnancy?
In a paragraph, discuss why age and sex must be predictor variables and can never be true independent variables. Next, design a simple study using age or sex as a predictor variable. What variable of interest might be correlated to age or sex? Write a two to three paragraph description of your design.
Already a member? Log In
Back