Notes on Research Methods in Psychology
Scientific Method
Observation
Define problem
Propose hypothesis
Gather evidence
Test hypothesis
Reject hypothesis
Retain hypothesis
Publish results
Theory building
Hypothesis
A tentative explanation of an event or relationship
Testable educated guess
It is the rewording of a question into a statement
Operational Definitions: Turn a theoretical question into an empirical one
Conceptual Level vs Concrete Level
Conceptual Level
Hypothesized relationship
Concepts
Frustration
Aggression
Concrete Level
Operational definitions
Interrupted finishing a puzzle
Aggression
Number of times interrupter insults interrupter
Observed relationship
Frustration-Aggression Example (from Conceptual/Concrete Levels)
Conceptual Level: Frustration relates to Aggression as a hypothesized relationship between concepts
Concrete Level: Measured with operational definitions such as the number of times an interrupter insults the interruptee while finishing a puzzle; observed relationship between interruption and aggression
Critical Thinking and Scientific Research
Questions to ask when reading or hearing about research:
What am I being asked to believe or accept?
What evidence is available to support the assertion?
Are there alternative ways of interpreting the evidence?
What additional evidence would help to evaluate the alternatives?
What conclusions are most reasonable?
Examples to illustrate
Theories and Secondary Sources
What is a theory?
What are the sources of information?
Primary versus secondary
Research Methods used in Psychology
Method: Naturalistic Observation
Features: Observations of human or animal behavior in the environment it typically occurs
Strengths: Provides descriptive data about behavior presumably uncontaminated by outside influences
Pitfalls: Observer bias and subject self-consciousness can distort results and can produce anthropomorphizing of animal behavior
Method: Case studies
Features: Intensive examination of the behavior and mental processes associated with a specific person or situation
Strengths: Provide detailed descriptive analyses of new, complex or rare phenomena
Pitfalls: May not provide representative picture of phenomena
Method: Surveys
Features: Standard sets of questions asked of a large number of subjects
Strengths: Gathers large amounts of descriptive data relatively quickly and inexpensively
Pitfalls: Sampling errors, poorly phrased questions, and response biases can distort results
Method: Correlational studies
Features: Non-experimental study designed to measure the degree of relationship (if any) between two or more events, measures, or variables
Strengths: Demonstrates existence of relationships, can lead to predictions from one variable to the other, can be used in a variety of settings
Pitfalls: Relationships may be coincidental, can not determine cause-effect relationships
Method: Experiments
Features: Manipulation of an independent variable and measurement of its effects on a dependent variable
Strengths: Can establish a cause-effect relationship between independent and dependent variables
Pitfalls: Confounding variables may prevent valid conclusions (but if done well, provides the strongest evidence)
Correlation: Understanding Relationships
The concept of correlation as a non-experimental measure of the degree of relationship between variables
Correlation scale (r values):
Perfect negative relationship: r = -1.00
Very large negative: r = -0.75
Large negative: r = -0.50
Moderate negative: r = -0.30
No relationship: r = 0.00
Small positive: r = +0.10
Moderate positive: r = +0.30
Large positive: r = +0.50
Very large positive: r = +0.75
Perfect positive: r = +1.00
Note on interpretation: Correlation does not imply causation; relationships may be coincidental
General interpretation framework: strength and direction of association
Basic formula (for reference): r = \frac{\mathrm{cov}(X,Y)}{\sigmaX \; \sigmaY}
Experiments
Purpose: establish a cause-effect relationship by actively manipulating an independent variable and observing effects on a dependent variable
Experimental design features:
Possible subjects
Experimental group
Random assignment (controls for subject differences)
Study and testing conditions
Identical conditions to control extraneous variables
Music included (in the example)
Control group
Independent variable (Cause)
Dependent variable (Effect)
Question: Is there a difference?
Example labeling: Behavior (test scores) as the dependent variable, Independent variable such as presence/absence of music
Quasi-experiments
Features: Measurement of dependent variables when independent variables were not entirely under the experimenter’s control
Strengths: Can provide strong evidence suggesting cause-effect relationships
Pitfalls: Lack of full control may weaken conclusions