Psychology Research Methods

Overview

This lecture covers the pitfalls of intuition in psychology and outlines scientific research methods including case studies, observation, surveys, correlational studies, and experiments.

Problems with Intuition

• Intuition about behavior is often misleading due to biases.
• Hindsight bias makes us believe we “knew it all along” after an event occurs.
• Overconfidence leads people to think they’re right even when they’re not.
• Humans tend to see order in random events, creating false patterns.

Scientific Inquiry in Psychology

• Psychological research follows the scientific method: question, theory, hypothesis, test, and replication.
• Operationalizing questions means defining variables in measurable terms.
• Replication of results across studies increases reliability.

Research Methods

• Case studies provide in-depth information about an individual but cannot be generalized or replicated easily.
• Naturalistic observation involves watching behavior without interference, good for description not explanation.
• Surveys and interviews collect self-reported attitudes and behaviors, but are sensitive to wording and sampling biases.
• Random sampling is needed for survey results to represent a population accurately.

Correlation and Causation

• Correlation measures how traits or behaviors are related but does not prove causation.
• Correlations help predict possible relationships but may be influenced by other variables.
• Once a set of observations have been made about the behavior of interest, psychological researchers look for patterns in the data. This often means searching for correlations, which in essence means “co-relation.” If our research finds a correlation between two variables, we take that to indicate that as one variable changes, you can predict the value of the second variable with some level of accuracy (and that accuracy is better when correlations are stronger).
  • For example, if height and weight were correlated, we would be better at guessing someone’s weight if we knew their height than if we did not. Of course, height and weight are correlated, but so are many, many other features of humans, including things we study in psychology. This lesson prepares you for the great wealth of knowledge correlational studies have taught us.
    One important note of caution related to the concept of correlation is warranted, however.
• A correlation only suggests that two variables are linked - we cannot declare for certain that one variable is causing the other. This is because there are likely many other variables that we did NOT measure that could account for the observed connection. For instance, if we know there is a correlation between ice cream consumption and the number of murders committed such that both variables increase together (a positive correlation), we would not want to claim that eating ice cream causes murderous intent or that committing murder causes a craving for ice cream. Instead, we would look for an additional variable like heat which might be responsible for changes in both variables.
  Researchers usually only measure a handful of variables they think are important to understanding the behavior of interest. But in truth, human behavior is so complex that many hundreds of variables or factors are behind even the simplest of behaviors. A correlation, then, tells us that we might have identified a possible connection. But before we declare that two variables are causally linked, a different research methodology is needed: the experimental approach.

Experiments and Variables

• Experiments test cause-and-effect by manipulating independent variables and measuring dependent variables.
• Participants are randomly assigned to experimental or control groups to reduce bias.
• Control groups receive placebos; experimental groups receive the variable being tested.
• Double-blind procedures prevent bias from participants and researchers.
• Two of the most fundamental considerations in conducting psychological scientific research are the concepts of reliability and validity. Reliability refers to the repeatability of the observations that have been made. If a researcher is observing two children on a playground who appear to be arguing - she might mark the behavior down as being high in aggression. Another researcher may mark the same encounter as mildly aggressive. The lack of agreement makes both observations unreliable - which one should we trust? Consistency of observation is the key to reliability - either across multiple raters or across instances of the same behavior. Validity is the scientific concept that gets at how close to the actual reality an observation gets - is what you claim to be measuring really and truly captured by how you're measuring it? Take the scenario we just discussed. Perhaps the researchers were using a scale that measures the level of aggression exhibited by children. Both of them thought they were observing aggression, albeit to a different degree. When instead, what was really happening was play fighting - the kids were acting out a scene from their favorite movie! This is one of the first set of questions that you should ask yourself from now on whenever you read about findings from scientific research. Ask yourself: Was the measure that was used reliable, or consistent? And then, was that the true, correct measure of that behavior?
  
  

Example Experiment: Caffeine and Problem Solving

• Formulate a clear, testable hypothesis (“Caffeine increases maze completion speed in adults”).
• Independent variable: caffeine dosage; dependent variable: maze completion speed.
• Obtain informed consent and randomly assign participants to groups.
• Compare results to determine if caffeine significantly affects performance.
• Clear methodology allows for replication by other researchers.

Key Terms & Definitions

• Hindsight Bias — Belief after an event that it was predictable, the “I-knew-it-all-along” effect.
• Operationalization — Defining variables in practical, measurable terms.
• Replication — Repeating research to verify findings.
• Case Study — Detailed examination of one individual or group.
• Naturalistic Observation — Watching behavior in natural settings.
• Survey — Collecting self-reported data from participants.
• Random Sampling — Selecting participants so each has an equal chance.
• Correlation — Statistical relationship between two variables.
• Causation — One variable producing a change in another.
• Independent Variable — The factor manipulated in an experiment.
• Dependent Variable — The outcome measured in an experiment.
• Control Group — Group not exposed to the experimental treatment.
• Placebo — An inert substance or condition.
• Double-Blind Procedure — Participants and researchers do not know group assignments.

Action Items / Next Steps

• Review the major research methods and their strengths/limitations.
• Practice distinguishing between correlation and causation.
• Prepare to design a simple experiment using the scientific method.