Source: General Psychology, The Ohio State University
Difficulty: Introductory | Prerequisites: Part 1 (Scientific Method and Descriptive Research)
Tags: correlation, correlational study, correlation coefficient, positive correlation, negative correlation, illusory correlation, experiment, independent variable, dependent variable, confounding variable, control group, experimental group, random assignment, placebo effect, double-blind experiment, causal inference
Part 1 of this unit covered how psychologists describe behaviour. This section moves to the next two levels: measuring relationships between variables (correlational research) and testing cause-and-effect (experiments). The distinction between correlation and causation is one of the most important ideas in the entire course, and exam questions return to it repeatedly. If you can cleanly explain why a correlation does not prove causation, and what an experiment adds that a correlational study cannot, you are in strong shape for the rest of the unit.
Correlational studies measure whether two variables move together (and in which direction), but they cannot prove that one causes the other. Experiments are the only method that can establish causation, because they involve manipulating one variable and measuring the effect on another while controlling everything else. Key safeguards in experiments include random assignment, control groups, and double-blind procedures.
Correlational study
A research method that looks for the relationship between two events, measures, or variables. Example: hours studied vs. grade achieved.
In simple terms, you are asking "do these two things tend to go together?" without trying to change either one.
Correlation coefficient
A numerical value that represents the strength and direction of the relationship between two variables. It ranges from -1.0 to +1.0.
In simple terms, the number tells you how closely two variables move in step. The sign (+ or -) tells you the direction; the size tells you the strength.
Positive correlation
A relationship where both variables move in the same direction: as one increases, the other increases (or as one decreases, the other decreases).
In simple terms, they rise and fall together. More hours studied, higher grades.
Negative correlation
A relationship where the variables move in opposite directions: as one increases, the other decreases.
In simple terms, one goes up while the other goes down. More absences, lower test scores.
Zero correlation
No systematic relationship between the two variables.
In simple terms, knowing one variable tells you nothing about the other.
Illusory correlation
Perceived relationships where none exist. People see patterns in data that are not there, often driven by expectations or stereotypes.
In simple terms, your brain is a pattern-finding machine, and sometimes it finds patterns that are not real.
Causal inference
A conclusion that one variable directly causes a change in another. Only experiments can support causal inferences.
In simple terms, this is the "A causes B" claim, and you need an experiment to make it.
Experiment
A research method in which the researcher directly varies a condition (the independent variable) and measures the resulting changes (in the dependent variable) while controlling other factors.
In simple terms, you change one thing on purpose and watch what happens, keeping everything else the same.
Independent variable (IV)
The condition altered by the experimenter. This is the suspected cause.
In simple terms, it is the thing you deliberately change.
Dependent variable (DV)
The condition measured by the experimenter. This is the suspected effect.
In simple terms, it is the outcome you measure to see whether your change had an impact.
Confounding variable
A variable outside your research design that alters the results, making it unclear whether the IV truly caused the change in the DV.
In simple terms, it is a hidden third factor that muddies your conclusions.
Experimental group
The group that receives the independent variable (the treatment or manipulation).
In simple terms, this is the group that gets the "thing you are testing."
Control group
The group that does not receive the independent variable. It serves as the baseline for comparison.
In simple terms, this is the "nothing changed" group you compare against.
Random assignment
Every participant has an equal chance of being placed in either the experimental or control group. This helps ensure the groups are comparable at the start.
In simple terms, you flip a coin (or equivalent) to decide who goes where, so the groups start off as similar as possible.
Representative sample
A small group that reflects the characteristics of the larger population you want to generalise to.
In simple terms, if your population is all university students, your sample should not be entirely first-year psychology majors.
Placebo effect
Changes in behaviour that occur because of a participant's belief that they are receiving a treatment, rather than from the treatment itself.
In simple terms, thinking you took the real pill can make you feel better, even if it was a sugar tablet.
Double-blind experiment
An experimental design in which neither the participant nor the experimenter knows who received the treatment and who received the placebo, until after data collection is complete.
In simple terms, nobody in the room knows who is in which group, which prevents both participant expectations and researcher bias from contaminating the results.
A correlational study measures the direction and strength of a relationship between two variables.
The correlation coefficient is a number between -1.0 and +1.0:
+1.0 = perfect positive correlation
-1.0 = perfect negative correlation
0 = no correlation
The closer the number is to +1.0 or -1.0, the stronger the relationship.
A correlation, no matter how strong, does not demonstrate causation. There may be a third variable driving both, or the direction of influence may be reversed.
Illusory correlations are common in everyday thinking. People tend to notice and remember instances that confirm their expectations while ignoring instances that contradict them.
Experiments are the only research method that can establish a cause-and-effect relationship.
The basic structure:
The researcher manipulates the independent variable (IV).
The researcher measures the dependent variable (DV).
All other variables are held constant or controlled.
Participants are split into an experimental group (receives the IV) and a control group (does not receive the IV).
Random assignment is used to place participants in groups, minimising pre-existing differences.
Worked example from the course (Assignment 5.3):
Research question: Do flashcards improve spelling test performance?
Control group: random half of the class that will not use flashcards
Experimental group: random half of the class that will use flashcards to study
Independent variable: flashcards (the thing being manipulated)
Dependent variable: grades on the weekly spelling test (the outcome being measured)
Confounding variables: amount of time students study, difficulty of the spelling test
Placebo effect: Participants may improve simply because they believe they are receiving treatment. A placebo condition (e.g., a sugar pill) helps detect this.
Double-blind design: When neither the participants nor the experimenters know who is in which group, both participant expectations and experimenter bias are controlled for. This is the gold standard for clinical trials.
Experiment structure (from the course notes):
Subjects in Experiment
/ \
Experimental Group Control Group
(receives IV) (does not receive IV)
| |
Measure DV Measure DV
| |
\--- Compare results --/
The IV is the possible cause. The DV is the possible effect. Any difference in the DV between the two groups is attributed to the IV, provided confounding variables have been controlled.
Correlational studies are used constantly in public health (e.g., smoking and lung cancer were first linked through correlation, long before experimental evidence was available). Experiments with double-blind, placebo-controlled designs are the foundation of pharmaceutical testing. Every drug that reaches the market has gone through this kind of trial. Understanding confounding variables is critical in fields from education policy to economics, where controlled experiments are often impractical and researchers must be cautious about causal claims.
Students frequently claim that a strong correlation proves causation. It does not. "Correlation does not imply causation" is the single most important sentence in this section.
Students mix up the independent and dependent variables. A useful mnemonic: the IV is what I (the experimenter) vary; the DV depends on what I did.
Students sometimes forget that confounding variables can exist even in well-designed experiments. Random assignment reduces their impact but does not eliminate all possible confounds.
The placebo effect is often described as "fake" or "not real." The effect on the participant is real; what is not real is the treatment they believe they received.
⚠️ "Correlation does not imply causation" is the most frequently tested principle in this section. Be able to explain why, using a third-variable or reverse-causation example.
⚠️ Know how to identify the IV, DV, and confounding variables in a described scenario. Exam questions love giving you a study description and asking you to label each.
⚠️ Understand why random assignment matters and what problem it solves (pre-existing differences between groups).
⚠️ Be able to explain the purpose of a double-blind experiment and how it improves on a single-blind design.
⚠️ Know the difference between the experimental group and the control group, and why both are necessary.
True or False: A correlation coefficient of -0.85 indicates a weak relationship.
Fill in the blank: The variable the experimenter manipulates is called the __________ variable.
True or False: Random assignment guarantees that there will be no confounding variables.
Fill in the blank: A __________ experiment is one in which neither the participant nor the researcher knows who received the treatment.
True or False: A positive correlation means that as one variable increases, the other decreases.
Answers: 1. False (it is a strong negative relationship; strength is determined by distance from zero, not by the sign). 2. Independent. 3. False (it reduces the likelihood of confounds but does not guarantee their absence). 4. Double-blind. 5. False (that describes a negative correlation; in a positive correlation, both variables move in the same direction).
Q: A study finds that students who sleep more hours tend to earn higher grades. Can we conclude that more sleep causes better grades? Why or why not?
A: No. This is a correlational finding, and correlation does not establish causation. A third variable (e.g., better time management) might cause both more sleep and higher grades, or the direction could be reversed (less academic stress from good grades leads to better sleep).
Q: In an experiment testing whether background music improves reading comprehension, identify the IV, DV, and one possible confounding variable.
A: The IV is background music (present or absent). The DV is reading comprehension scores. A possible confounding variable is the difficulty of the reading passages, if they differ between groups.
Q: Why is a control group necessary in an experiment?
A: The control group provides a baseline. Without it, you have no way of knowing whether changes in the DV were caused by the IV or would have occurred anyway.
Q: What is the difference between the placebo effect and a double-blind experiment?
A: The placebo effect is a change in behaviour caused by a participant's belief that they are receiving a treatment. A double-blind experiment is a design that controls for this (and for experimenter bias) by ensuring neither participants nor experimenters know who is in which condition.
Q: A correlation coefficient of +0.92 is found between ice cream sales and drowning rates. Does ice cream cause drowning?
A: No. A third variable, hot weather, likely drives both: people buy more ice cream and swim more often when it is hot. This is a classic example of why correlation does not imply causation.
This section connects back to Part 1 (descriptive research answers "what," correlation answers "how much do these relate," and experiments answer "does A cause B"). It also connects forward to the statistics section of Unit 2, where you will learn how researchers determine whether experimental results are meaningful. The logic of independent and dependent variables reappears in nearly every research-based unit in the course, from learning and memory to social psychology.
correlational study, correlation coefficient, positive correlation, negative correlation, zero correlation, illusory correlation, causation vs correlation, experiment, independent variable, dependent variable, confounding variable, control group, experimental group, random assignment, representative sample, placebo effect, double-blind experiment, causal inference, PSY 1100, general psychology research methods, Ohio State psychology, third variable problem