Overview
This lecture covers how psychological research uses the scientific method to overcome intuition, bias, and error, outlining key methods and experimental design.
Intuition and Bias in Psychology
- Intuition is not always reliable and can be exactly wrong.
- Hindsight bias ("I-knew-it-all-along") leads us to believe outcomes were predictable after they happen.
- Overconfidence can cause us to falsely believe we're correct about people or outcomes.
- We naturally perceive order in random events, leading to mistaken assumptions.
Scientific Method in Psychological Research
- Research begins by turning general questions into measurable, testable propositions (operationalizing questions).
- A scientific theory explains and organizes observations and predicts outcomes.
- A hypothesis is a specific, testable prediction derived from a theory.
- Clear definitions and language allow for experiment replication, which is essential for validation.
Research Methods
- Case studies: In-depth studies of one individual; useful for description, but limited generalizability and replication.
- Naturalistic observation: Observing subjects in their natural environments; great for describing behavior, not for explaining it.
- Surveys and interviews: Collect self-reported data; results depend on question phrasing and on obtaining a representative, random sample.
Correlation and Causation
- Correlation measures the relationship between two variables but does not prove causation.
- Other factors (confounding variables) may explain a observed relationship.
Experimental Design
- Experiments isolate effects by manipulating the independent variable while keeping others constant.
- Participants are randomly assigned to experimental or control groups to minimize confounding variables.
- Placebos and double-blind procedures prevent bias from participants and researchers.
- Informed consent is required from all participants.
Example Experiment: Caffeine and Problem-Solving
- Hypothesis: Adults given caffeine will navigate a maze faster than those not given caffeine.
- Independent variable: Caffeine dosage.
- Dependent variable: Maze completion time.
- Groups: Control (decaf), low dose (100 mg), high dose (500 mg).
- Results are compared to test the hypothesis and can be replicated by others.
Key Terms & Definitions
- Hindsight bias — the tendency to see events as predictable after they have happened.
- Operationalizing — defining concepts so they can be measured or tested.
- Theory — an explanation that organizes observations and predicts outcomes.
- Hypothesis — a testable prediction derived from a theory.
- Correlation — a relationship between two variables.
- Causation — when one variable directly affects another.
- Confounding variable — an outside factor that might influence results.
- Placebo — an inert substance used to compare against an experimental treatment.
- Double-blind procedure — both researchers and participants do not know who is in the experimental or control group.
Action Items / Next Steps
- Review the definitions and examples of research methods.
- Practice identifying independent and dependent variables in sample experiments.
- Read more about survey design and sources of bias in psychological research.