Behavioral Science Methods Notes
Public Observability and the Scientific Method
- Core idea: What's manipulated and measured in science must be publicly observable to verify results.
- Thoughts and feelings are not directly observable; verbal reports are observable proxies but don't guarantee true mental states.
- Historically, psychology used controlled environments to study observable behavior.
- Living organisms have "memory" (prior experiences influence future states), unlike atoms in physics/chemistry, requiring consideration of history in experiments.
- Three foundational methods: Descriptive, Correlational, Experimental.
Descriptive Methods
- Used for observing phenomena; don't establish causality.
- Case studies: Intense focus on a single subject for rare phenomena (less common now).
- Manualistic observation: Controlled observation in everyday settings.
- Naturalistic observation: Observing behavior in natural environments.
- Surveys: Snapshot of opinions/behaviors at one time. Limitations: momentary, sensitive to wording, susceptible to bias (e.g., social desirability).
Correlational Method
- Measures relationships between two or more variables when manipulation is unethical or impractical.
- Shows strength and direction of a relationship, but does not imply causation (due to third variables, reverse causation, or coincidence).
- Common pitfall: Inferring cause-and-effect from correlation.
Experimental Method (Overview and Context)
- Considered the most rigorous for establishing causality by manipulating independent variables, controlling factors, and measuring effects on dependent variables.
- Feasibility and ethics are key for its application.
Ethical, Philosophical, and Practical Implications
- Ethical constraints dictate study design (e.g., personality, drug use).
- Public observability ensures transparency and replicability.
- Memory in living systems requires longitudinal/sequential designs.
- Verbal reports are limited by accuracy and bias.
- Survey sensitivity highlights design importance.
- Distinguishing observed behavior from inferred internal states is crucial.
Key Takeaways
- Public observability is central: everything must be observable.
- Thoughts are not directly observable; verbal reports are cautious proxies.
- Descriptive methods provide rich data but limited causal inference.
- Correlational methods show relationships but not causation.
- Experimental methods aim for causality via variable control, when ethical.
- Behavioral sciences account for memory and history in living systems, differentiating them from non-biological sciences.
- There are approximately 8×109 people on the planet today, relevant for generalizability of descriptive studies.