Spotting Extraneous and Confounding Variables, Explained

UPI Study › Courses › Research Methods in Psychology › Free lessons › Lesson 10.2 Free with a free account

Spotting Extraneous and Confounding Variables

Research Methods in Psychology · Chapter 10: Internal Validity and Its ThreatsSection 10.2 · lesson 59 of 9410 min · free with a free account

← What Internal Validity Means in ResearchNext lesson: Physical and Researcher Influences on Internal Validity →

Earn 3 college credits for Research Methods in PsychologyRecommended for credit by NCCRS · Psychology 111 · self-paced · one payment, lifetime access
Enroll in Research Methods in Psychology · $250 →Download the syllabus (PDF)

In this lesson

Recognizing the variables that quietly compete with the independent variable for credit.

  • Define extraneous variables and why every study has them
  • Distinguish a confound from a mere extraneous variable
  • Apply the three conditions that create a confound
  • Use the third-variable test on any causal claim

Free study notes

Get this lesson's notes by email

The key points and 4 practice questions with answers, in your inbox. One email, free.

One email. Unsubscribe in one click.

Transcript

Everything said in the video, section by section (1,432 words). Read along, or search it.

Internal Validity and Its Threats

Spotting Extraneous and Confounding Variables

Students who attend tutoring get better grades. So tutoring works. That inference is wrong, and this section explains exactly why. The culprit is a variable nobody put in the study at all. You will define extraneous variables, and see why every study contains them. You will distinguish a genuine confound from a merely extraneous variable. You will apply the three conditions that turn one into the other. And you will learn the single most useful diagnostic question in research methods. Once you have it, you will not be able to read a headline the same way again.

Two Concepts

A Distinction Worth Getting Exactly Right

The previous section identified the central challenge of internal validity. Eliminating plausible alternative explanations. To pursue that systematically, researchers use two related but distinct concepts. Extraneous variables, and confounding variables. These terms get used loosely in conversation, even by working researchers. But the conceptual distinction between them is essential. Mistaking one for the other is a common source of design error. It leads people to control the harmless things, and ignore the dangerous ones. So take them one at a time, and be precise about the difference.

The Background

Extraneous Variables

An extraneous variable is any variable other than the independent variable. One that could, in principle, influence the dependent variable. The key word in that definition is could. Take a study of caffeine and reaction time. Its extraneous variables include hours of sleep the night before. Age, baseline reaction speed, room temperature, and time of day. And dozens of other factors nobody has bothered to list. None of them is the focus of the study. Yet each could in principle move the dependent variable up or down. And crucially, they exist whether or not the researcher attends to them. They are part of the background of any study ever conducted.

Noise or Bias

Most of Them Are Harmless

Here is the good news, and it is more important than it sounds. Most extraneous variables are harmless. They add random variability to the data. That is often called noise, or error variance. But notice what they do not do. They do not systematically favor one experimental condition over another. So they do not bias the conclusion at all. They make the picture fuzzier without moving it. Which means the researcher's task is not to eliminate every extraneous variable. It is to identify them in advance, and decide which ones need controlling. And which can safely be tolerated as noise. Controlling all of them is impossible anyway.

The Foreground

Confounding Variables

Now the dangerous kind. A confounding variable is an extraneous variable with one extra property. It is systematically associated with the independent variable. In such a way that its effect cannot be separated from the effect of interest. Where extraneous variables sit quietly in the background, confounds occupy the foreground. They actively compete with the independent variable for credit. Go back to the caffeine study, and add one detail. Suppose the caffeine drinkers happen to be older than the non-drinkers. Age is no longer just an extraneous variable. It has become a confound. Because any difference in reaction time between the groups is now ambiguous. It could plausibly be due to the caffeine, or to the age, and nothing in the data separates them.

The Test

Three Conditions, All at Once

Now an important conceptual point that students often miss. Confounding is not an inherent property of a variable. There is no such thing as a confounding variable in the abstract. It is a property of the relationship between that variable and the rest of the study. The same variable that is completely harmless in one design is catastrophic in another. Three conditions must hold simultaneously for an extraneous variable to act as a confound. It must be related to the independent variable, differing across conditions. It must be related to the dependent variable, plausibly influencing the outcome. And it must remain uncontrolled, either unmeasured or unequal. All three at once. Break any one of them and the threat dissolves.

Conditions One and Two

It Must Vary, and It Must Matter

Take the first two conditions together. First, the variable must vary along with the independent variable across conditions. If treatment-group participants tend to be older, or more motivated, or wealthier. Then those characteristics covary with treatment status. And notice the negative case, which is genuinely useful. A variable equally distributed across conditions cannot confound the independent variable. It provides no differential explanation for any group difference you observe. Second, the variable must plausibly influence the outcome. Something with nothing to do with the dependent variable cannot account for changes in it. However unbalanced its distribution happens to be. Suppose the treatment group wears blue shirts more often than the control group. That is unbalanced, and it still will not confound a study of memory. Because shirt color does not plausibly influence memory performance.

Condition Three

Failure of Control

The third condition is the one you control. The variable must remain uncontrolled in the design. If the researcher has measured it and statistically adjusted for it, the threat is gone. The same applies if it has been held constant by design. It can no longer compete with the independent variable for credit. So control transforms a potential confound back into a benign extraneous variable. That is a genuinely useful way to think about experimental design. You are not eliminating variables. You cannot, and you do not need to. You are severing the link between them and the independent variable. The variable can still vary. It just cannot vary in step with your treatment.

The Diagnostic

The Third-Variable Test

Which brings us to the most useful question in this entire chapter. Whenever you encounter a claim that X causes Y, ask yourself the following. What third variable could be related to both X and Y? Related in such a way that it explains the observed association. Without any direct causal connection between X and Y at all. If you can identify such a variable, the original causal claim is in trouble. This habit of mind is sometimes called thinking like a Devil's advocate. And it is the single most useful diagnostic skill for evaluating internal validity. The good news is that it is a skill, which means it improves with practice. Practice it on news headlines, on advertising claims, and on published studies. It sharpens surprisingly quickly.

In the Wild

Four Confounds Wearing Causal Clothing

Four real examples show how ordinary this failure is. Students who attend tutoring sessions earn higher grades, so tutoring improves grades. The confound is motivation. Students who choose to attend tutoring are likely more motivated to begin with. Motivation is associated with attendance, and independently raises grades. Children who watch more educational television score higher on vocabulary tests. The confound is parental involvement. Parents who select educational programming may also read with their children. The vocabulary advantage may flow from those behaviors, not from the television. Employees who use a new software tool complete projects faster. The confound is self-selection. Early adopters are already more skilled and motivated. Their advantage may reflect who they were before adopting the tool. And a university with a higher graduation rate has a better learning environment. The confound is admission selectivity. Admitting better-prepared students raises graduation, whatever the environment does.

Prevention

Catching Confounds While You Can Still Redesign

Confounds are far easier to prevent than to remedy. Four diagnostic habits help you spot them while the study can still be redesigned. First, before data collection, enumerate every plausible third variable. Every one that could differ across the conditions you care about. Second, examine your recruitment and assignment procedures. Look specifically for sources of self-selection. Third, ask what else changes when the study moves from one condition to another. The location, the time of day, the experimenter, the instructions, the materials. Fourth, request critical review from a colleague unfamiliar with the project. They will often see candidate confounds the investigator has stopped noticing. And take particular care with quasi-experimental designs. When groups already differ on the independent variable, they likely differ on much else too. Comparing smokers with non-smokers, or a city before and after a policy change.

The Response

Four Ways to Sever the Link

Once a candidate confound is identified, you have four options. You can hold the variable constant, so it cannot vary across conditions. You can match it across conditions, so the groups become equivalent on it. You can measure it and enter it as a covariate in your statistical analysis. That partials out its contribution, though only for what you measured. Or you can neutralize it by random assignment. Which equalizes groups in expectation on every variable simultaneously. Both the observed ones, and the ones nobody thought of. A later section returns to these techniques in detail. The conceptual point here is simply what control means. It means severing the link between the independent variable and a specific extraneous one. So the latter can no longer compete for credit.

About Research Methods in Psychology at UPI Study

Research Methods in Psychology (Psychology 111) is a self-paced online course of 94 video lessons in 16 chapters, from the scientific method and research ethics to designs, data collection, sampling, measurement, validity, experiments, descriptive statistics and hypothesis testing. It carries a NCCRS credit recommendation for 3 semester credits.

Every lesson transcript is free on pages like this one, and the videos are free with a free UPI Study account. Enrolling adds a practice quiz after each lesson, the graded work and final exam that earn the credit, lifetime access, and an official transcript on completion. UPI Study has helped over 130,000 students with college credits. ACE and NCCRS have over 2,100 cooperating colleges that will consider accepting credits; the final decision is with the universities.

Enroll now → Course details Download the syllabus All Psychology and social science courses Every course with PRO →

Compare: Research Methods vs Statistics

Why take it here

Why students take Research Methods in Psychology with UPI Study

How psychologists know what they knowResearch Methods in Psychology teaches how psychological studies are designed, run and judged: research ethics, surveys and observation, correlational and experimental designs, sampling, reliability and validity, and the statistics used to test hypotheses. It is a core course in psychology and many social-science degrees.
Credit at a fraction of the priceThe same 3 credits can cost $750 or more at a US college. Here it is one payment, self-paced, with lifetime access.
Finish your degree soonerTransfer credits reduce the courses you still owe, so you can graduate earlier or lighten a heavy semester.
Strengthen an applicationA completed college-level course on an official transcript shows admissions and scholarship committees you can do the work.
Study on your own scheduleNo start dates, no deadlines. Watch the lessons, practice, and take the final when you are ready.
130,000+students helped with college creditsACE and NCCRS have over 2,100 cooperating colleges that will consider accepting credits; the final decision is with the universities.

Where the credit goes

Universities that work with NCCRS

Research Methods in Psychology carries a NCCRS credit recommendation. These universities work with NCCRS; some are UPI Study's signed partners. ACE and NCCRS have over 2,100 cooperating colleges that will consider accepting credits; the final decision is with the universities.

Thomas Edison State UniversityThomas Edison State UniversitySigned partnerUniversity of the PeopleUniversity of the PeopleSigned partnerCharter Oak State CollegeCharter Oak State CollegeSigned partnerStrayer UniversityStrayer UniversitySigned partnerColumbia Southern UniversityColumbia Southern UniversitySigned partnerIndiana TechIndiana TechSigned partnerPenn StatePenn StateWorks with NCCRSSouthern New Hampshire UniversitySouthern New Hampshire UniversityWorks with NCCRSSUNYSUNYWorks with NCCRS +many more cooperating universitiesFind my college →

Find my college → Not listed? Ask Athena in the corner of this page, or book a 15-minute call.

FAQ

Questions students ask us

About Spotting Extraneous and Confounding Variables

Do extraneous variables always bias a study's results?
No, most extraneous variables do not bias results; they add random variability, often called noise or error variance, without systematically favoring one condition. In a caffeine and reaction-time study, sleep, age, or room temperature blur the picture without shifting it, as long as they are not tied to the conditions. Bias arises when an extraneous variable varies systematically with the independent variable and becomes a confound. Researchers therefore decide which ones need control rather than trying to control them all.
When does an extraneous variable become a confound?
An extraneous variable becomes a confound only when three conditions hold at once: it differs across the conditions of the independent variable, it plausibly influences the dependent variable, and it remains uncontrolled. Break any one and the threat disappears. A variable spread equally across groups cannot confound the comparison, and an unbalanced variable that cannot affect the outcome, such as shirt color in a memory study, cannot either.
What is an example of a confounding variable?
A classic example: students who attend tutoring earn higher grades, but students who choose tutoring are likely more motivated to begin with, and motivation independently raises grades. Similarly, children who watch more educational television may have larger vocabularies because involved parents both choose that programming and read with them. In each case a third variable related to both the presumed cause and the outcome could explain the association.
How can you spot confounding variables before collecting data?
The most useful habit is asking what third variable could be related to both the presumed cause and the outcome. Before data collection, list every plausible variable that could differ across conditions, check recruitment and assignment for self-selection, ask what else changes between conditions, such as room, time of day, experimenter, or materials, and have a colleague unfamiliar with the project review the design. Confounds are far easier to prevent than to remedy.

About the Research Methods in Psychology course

How many college credits is Research Methods in Psychology?
Research Methods in Psychology (Psychology 111) carries a NCCRS credit recommendation for 3 semester credits. ACE and NCCRS have over 2,100 cooperating colleges that will consider accepting credits; the final decision is with the universities.
How much does Research Methods in Psychology cost and what is included?
$250, paid once, with lifetime access: all 94 video lessons, a practice quiz after each, the graded work and final exam, and an official transcript when you finish. Students who want several courses choose UPI Study PRO instead.
Which universities accept UPI Study credits?
UPI Study has signed partner agreements with universities such as Thomas Edison State University, University of the People and Charter Oak State College, and works with universities such as Penn State, SNHU and SUNY. ACE and NCCRS have over 2,100 cooperating colleges that will consider accepting credits; the final decision is with the universities. Use "Find my college" on this page to check yours.
How long does Research Methods in Psychology take?
It is self-paced with no deadlines. The 94 lessons total about 16 hours of video; most students add reading and practice time and finish in a few weeks to a few months, on their own schedule.
How are the exams done?
Quizzes and assignments are online in the course. The final exam is taken online with proctoring; a person on the UPI Study team checks assignments, usually within two days.
How do my credits reach my college?
When the course shows 100% complete, request the official transcript from your dashboard. It is sent to the college you name, and you can send it to more than one. Check your college first with "Find my college" on this page, or ask Athena.

Ready for the full course? Earn 3 credits in Research Methods in Psychology.

94 video lessons, a quiz after each one, and the graded work that earns the credit. Recommended for credit by NCCRS.

Enroll in Research Methods in Psychology →Download the syllabusEvery course with PRO →
Research Methods in Psychology
3 credits · NCCRSEnroll →