The Scientific Method in Psychology

A. There are 3 types of studies:


-1. Descriptive Studies (naturalistic observations, case studies).

Example: Individuals observed in their environment.


-2. Correlational studies

Establishing correlations between non-manipulated variables. Relationships are not necessarily causal.


-3. Experiments.

Manipulating variables and testing cause and effect relationships.



B. Theory Construction and a Philosophy of Science.

Theory: System of interelated ideas used to explain a set of observations


Hypothesis: specific prediction derived from theory


Designing experiments to test hypotheses and theory


The best way to advance knowledge is to attempt disproving a theory or hypothesis.

Alternative: design single experiment to test several contradictory hypotheses.



C. The Scientific Method.


-1. Identify a problem and formulate hypothesis


-2. Design experiment using dependent and independent variables


-3. Perform experiment using experimental and control groups


-4. Evaluate hypothesis (maybe reformulate corrected hypothesis and replicate data)


-5. Communicate results



D. Identifying a Problem


-1. Having a good idea


-2. The hypothesis


-3. The theory (one theory, several testable hypotheses)


-4. The Dogma


-5. Using naturalistic observations



E. Designing an Experiment


-1. Definitions



Dependent vs independent variable



Experimental vs control group



Validity (do the variables really reflect what they are supposed to manipulate or measure?)


-2. Independent variables (beware of confounding variables), and counterbalancing



F. Performing an Experiment


-1. Reliability of measurements (are the measurements repeatable?)


-2. Selecting the subjects (random assignment)


-3. Circumvent subject’s and experimenter’s expectations (placebo, double-blind studies)



G. Correlational Studies.


Correlations do not necessarily indicate a cause and effect relationship



H. Case Studies.


Not necessarily representative of entire population



I. Surveys.


Sample has to be representative of population



J. Generality


Generalize from a sample.