Source: General Psychology lecture notes, The Ohio State University
Difficulty: Introductory | Prerequisites: None
Tags: naturalistic observation, correlation, causation, illusory correlation, correlation coefficient, research methods, psychology, PSYCH 1100
This topic sits near the beginning of any introductory psychology course, inside the unit on research methods. Before you can evaluate any claim in psychology, you need to understand how researchers collect evidence and what kinds of conclusions each method allows. Naturalistic observation and correlational research are two of the most common non-experimental approaches, and both come with a critical limitation: they cannot, on their own, establish cause and effect. Understanding why that limitation exists, and what tool does allow causal claims (the experiment, covered in the companion notes), is one of the most heavily tested ideas in introductory psych.
Naturalistic observation means watching behaviour in the real world without interfering. Correlational studies measure the relationship between two variables using a statistic called the correlation coefficient (ranging from -1 to +1), but a correlation can never prove that one variable causes the other. People also fall for illusory correlations, where they "see" a relationship that is not really there because it confirms what they already believe.
Naturalistic observation
A research method in which behaviour is observed and recorded in its natural environment, without intervention or manipulation by the researcher.
Think of it as: just watching people (or animals) do their thing in the wild, and taking notes.
Reactivity
The tendency for participants to change their behaviour when they know they are being observed.
In simple terms, this means people act differently when they know someone is watching them.
Correlation
A statistical relationship between two variables, described by the direction (positive or negative) and strength of the association.
Think of it as: two things tend to move together, but that does not mean one is causing the other.
Correlation coefficient
A numerical value, ranging from -1.0 to +1.0, that expresses the strength and direction of a linear relationship between two variables. Values near +1 or -1 indicate strong relationships; values near 0 indicate weak or no linear relationship.
In simple terms, this means: the closer the number is to -1 or +1, the tighter the pattern. Zero means no pattern at all.
Illusory correlation
The perception of a relationship between two variables when no such relationship exists, or the relationship is much weaker than believed. Driven by confirmation bias: we notice and remember evidence that fits our expectations and overlook evidence that does not.
Think of it as: seeing a pattern that is not really there because you expected to find one.
Researchers watch and record behaviour in the setting where it would occur anyway, without manipulating anything.
Classic example: Jane Goodall observing chimpanzees in the wild over decades.
Advantages
High ecological validity: behaviour is genuine because participants are in their real environment.
Useful when it would be unethical or impractical to create the situation in a lab.
Disadvantages
Reactivity: if participants notice the observer, they may alter their behaviour, which contaminates the data.
No control over outside variables, so you cannot draw causal conclusions.
Replication can be difficult because natural settings vary.
A correlation describes the relationship between two different variables.
Correlations are especially useful when you cannot (or should not) manipulate a variable. For example, you cannot randomly assign people to smoke for 30 years to study lung cancer, but you can measure the correlation between smoking and cancer rates.
The correlation coefficient
Ranges from -1.0 to +1.0.
Positive values: as one variable increases, the other tends to increase (e.g., height and shoe size).
Negative values: as one variable increases, the other tends to decrease (e.g., hours of sleep and stress levels).
Values near 0: little or no linear relationship.
Critical rule: correlation does not imply causation.
Two variables may move together because of a third, unmeasured variable (a confound), or the causal direction may be reversed, or the link may be coincidental.
A cognitive bias in which people perceive a relationship between two things that is either non-existent or far weaker than they believe.
Driven by confirmation bias: we selectively notice and remember instances that fit our prior beliefs, and ignore or forget instances that contradict them.
Example: believing that a full moon causes strange behaviour. People remember the odd event on a full-moon night and forget the many uneventful full-moon nights.
Naturalistic observation is the backbone of field research in animal behaviour, child development, and ethnography. Primatologists, playground researchers, and anthropologists all rely on it.
Correlation is how epidemiologists first linked smoking to lung cancer. Because you cannot ethically assign people to smoke, the entire evidence base began with correlational data. Understanding why that evidence could not, alone, prove causation is the same reasoning you need for exam questions on research design.
Students often think that a strong correlation (close to -1 or +1) means one variable causes the other. It does not. Strength tells you how tightly the variables move together, not why.
A negative correlation does not mean "no relationship." It means the variables move in opposite directions. A correlation of -0.85 is a very strong relationship.
Students sometimes confuse reactivity with the observer simply being present. The issue is not the observer's existence but whether the participant is aware of being watched and changes behaviour as a result.
Illusory correlation is often confused with simple confirmation bias. They are closely related, but illusory correlation is specifically about perceiving a statistical relationship that is not there (or is weaker than believed), whereas confirmation bias is the broader tendency to seek out supporting evidence.
⚠️ "Correlation does not imply causation" is one of the most frequently tested statements in introductory psychology. Expect it on the exam in multiple forms.
⚠️ You will likely be asked to identify the advantages and disadvantages of naturalistic observation. Reactivity is the key disadvantage to know.
⚠️ Be ready to interpret a correlation coefficient: know what the sign (positive/negative) and the magnitude (close to 0 vs close to 1) each tell you.
⚠️ Illusory correlation often appears in questions about why people hold false beliefs or stereotypes. Connect it to confirmation bias.
True or false: A correlation coefficient of -0.90 indicates a weak relationship between two variables.
Fill in the blank: The main disadvantage of naturalistic observation is __________, which occurs when participants change their behaviour because they know they are being watched.
True or false: If two variables are strongly correlated, we can conclude that one causes the other.
Fill in the blank: An __________ correlation is when we perceive a relationship between two variables that does not truly exist.
True or false: Naturalistic observation is useful when it would be unethical to manipulate the variable of interest.
Answers: 1. False (it is very strong, the sign only indicates direction). 2. Reactivity. 3. False. 4. Illusory. 5. True.
Q: A researcher finds a correlation of +0.72 between hours spent studying and exam scores. Can the researcher conclude that studying more causes higher exam scores? Why or why not?
A: No. Correlation does not imply causation. A third variable (e.g., motivation or prior knowledge) could explain both the increased study time and the higher scores. To establish causation, an experiment with random assignment and manipulation of the independent variable would be needed.
Q: Jane Goodall spent years observing chimpanzees in Tanzania. What research method is this, and what is one key limitation?
A: Naturalistic observation. A key limitation is reactivity: the chimpanzees may have changed their behaviour because of Goodall's presence, although this tends to decrease over time as subjects habituate to the observer.
Q: A student claims that because ice cream sales and drowning rates are positively correlated, eating ice cream must cause drowning. What error is the student making, and what is a more likely explanation?
A: The student is assuming that correlation implies causation. A more likely explanation is that a third variable, hot weather, independently increases both ice cream sales and the number of people swimming (and therefore the number of drownings).
Q: What is the range of possible values for a correlation coefficient, and what does a value of 0 indicate?
A: The range is -1.0 to +1.0. A value of 0 indicates no linear relationship between the two variables.
Q: Explain illusory correlation and give an example.
A: Illusory correlation is the perception of a relationship between two variables where none exists (or it is weaker than believed). Example: believing that full moons cause emergency-room visits to spike. People remember the busy nights that coincide with a full moon and forget the busy nights that do not, creating a false sense of association.
This material connects directly to experimental design (covered in the companion notes). The whole point of learning what correlation cannot do is to understand why experiments, with their random assignment and manipulation, are the gold standard for causal claims.
Illusory correlation ties into the broader topic of cognitive biases and heuristics, which you will encounter in the social psychology and thinking/reasoning units later in the course.
Reactivity in naturalistic observation is related to the Hawthorne effect, which comes up in industrial-organisational psychology and in discussions of demand characteristics in experiments.
Naturalistic observation, field observation, observational research, Jane Goodall, correlation, correlation coefficient, Pearson r, positive correlation, negative correlation, zero correlation, causation, third variable problem, confounding variable, illusory correlation, confirmation bias, reactivity, observer effect, Hawthorne effect, ecological validity, research methods in psychology, PSYCH 1100, General Psychology, non-experimental methods