Psychology Research Methods

The Scientific Method in Psychology

Steps of Scientific Investigation

1. Observation: Identify phenomenon of interest
2. Question: Formulate research question
3. Literature Review: Examine existing research
4. Hypothesis: Develop testable prediction
5. Method Design: Plan data collection
6. Data Collection: Gather information systematically
7. Analysis: Examine and interpret data
8. Conclusion: Accept or reject hypothesis
9. Replication: Repeat for verification
10. Theory Building: Integrate findings

Key Scientific Principles

• Empiricism: Knowledge through observation
• Objectivity: Minimize bias and subjectivity
• Replicability: Results must be reproducible
• Falsifiability: Hypotheses must be testable
• Parsimony: Prefer simpler explanations
• Systematic: Organized and methodical approach

Types of Research Questions

• Descriptive: What is happening?
• Correlational: How are variables related?
• Causal: Does X cause Y?
• Comparative: How do groups differ?
• Developmental: How do things change over time?

Research Designs

Between-Subjects Design

Different participants in each condition:

• Advantages: No order effects, shorter sessions
• Disadvantages: Requires more participants, individual differences
• Control: Random assignment essential
• Analysis: Independent samples t-test, ANOVA

Within-Subjects Design

Same participants in all conditions:

• Advantages: Controls individual differences, fewer participants
• Disadvantages: Order effects, carryover effects
• Control: Counterbalancing required
• Analysis: Paired samples t-test, repeated measures ANOVA

Mixed Design

Combination of between and within factors:

• Example: Treatment type (between) × time (within)
• Advantages: Flexible, comprehensive
• Complexity: More complex analysis required
• Analysis: Mixed ANOVA

Factorial Design

Multiple independent variables:

• Main Effects: Effect of each variable independently
• Interactions: Combined effects of variables
• Notation: 2×2, 2×3, etc.
• Power: Efficient testing of multiple hypotheses

Experimental Methods

True Experiments

Gold standard for establishing causation:

• Manipulation: Independent variable controlled
• Random Assignment: Participants randomly allocated
• Control: Confounding variables minimized
• Measurement: Dependent variable observed

Variables in Experiments

• Independent Variable (IV): Manipulated factor
• Dependent Variable (DV): Measured outcome
• Confounding Variables: Unwanted influences
• Moderator Variables: Affect relationship strength
• Mediator Variables: Explain relationship mechanism

Experimental Control

• Random Assignment: Equalizes groups
• Control Groups: Baseline comparison
• Placebos: Control for expectations
• Blinding: Single or double-blind procedures
• Standardization: Consistent procedures
• Counterbalancing: Controls order effects

Quasi-Experiments

When random assignment isn't possible:

• Natural Groups: Pre-existing differences
• Examples: Age, gender, clinical diagnosis
• Limitation: Cannot infer causation definitively
• Designs: Nonequivalent groups, interrupted time series

Field Experiments

Experiments in natural settings:

• Ecological Validity: Real-world relevance
• Less Control: More confounding variables
• Examples: Bystander intervention studies
• Ethics: Informed consent challenges

Descriptive Research Methods

Observational Studies

Naturalistic Observation

• Setting: Natural environment
• Intervention: No manipulation
• Advantages: High ecological validity
• Challenges: Observer effects, lack of control
• Example: Playground aggression studies

Structured Observation

• Setting: Controlled environment
• Standardization: Specific situations created
• Advantages: More control, replicable
• Example: Strange Situation attachment study

Participant Observation

• Role: Researcher as participant
• Insight: Insider perspective
• Challenges: Objectivity, dual roles
• Example: Ethnographic studies

Survey Research

• Purpose: Gather self-report data
• Formats: Questionnaires, interviews, online
• Scales: Likert, semantic differential, ranking
• Advantages: Efficient, large samples
• Limitations: Response bias, social desirability

Case Studies

• Focus: In-depth individual or group analysis
• Data: Multiple sources and methods
• Advantages: Rich detail, rare phenomena
• Limitations: Generalizability issues
• Famous Examples: Phineas Gage, H.M.

Correlational Studies

• Purpose: Examine relationships between variables
• Correlation Coefficient: -1 to +1
• Direction: Positive or negative
• Strength: Weak, moderate, strong
• Limitation: Correlation ≠ causation

Longitudinal Studies

• Design: Same participants over time
• Advantages: Developmental changes, within-person
• Challenges: Attrition, time, cost
• Example: Grant Study of Adult Development

Cross-Sectional Studies

• Design: Different groups at one time
• Advantages: Quick, economical
• Limitation: Cohort effects
• Use: Age differences research

Data Collection Methods

Self-Report Measures

Questionnaires

• Closed-ended: Multiple choice, rating scales
• Open-ended: Written responses
• Design: Clear, unbiased questions
• Administration: Paper, online, mobile

Interviews

• Structured: Standardized questions
• Semi-structured: Flexible follow-up
• Unstructured: Open conversation
• Skills: Rapport, probing, recording

Behavioral Measures

• Direct Observation: Recording specific behaviors
• Coding Systems: Systematic categorization
• Frequency: Count occurrences
• Duration: Time measurements
• Intensity: Severity ratings

Physiological Measures

• Brain Imaging: fMRI, EEG, PET
• Autonomic: Heart rate, skin conductance
• Hormonal: Cortisol, testosterone
• Eye Tracking: Attention and processing
• Reaction Time: Processing speed

Psychological Tests

• Intelligence: IQ tests (WAIS, Stanford-Binet)
• Personality: Big Five, MMPI
• Clinical: Depression scales, anxiety inventories
• Neuropsychological: Memory, attention tests
• Projective: Rorschach, TAT

Archival Research

• Sources: Records, databases, documents
• Advantages: No participant burden, historical data
• Challenges: Data quality, missing information
• Examples: Census data, medical records

Sampling Techniques

Probability Sampling

Simple Random Sampling

• Every member has equal chance
• Random number generation
• Unbiased but may not represent subgroups

Stratified Random Sampling

• Population divided into strata
• Random sampling within strata
• Ensures subgroup representation

Cluster Sampling

• Groups selected randomly
• All members within clusters studied
• Efficient for geographic distribution

Systematic Sampling

• Every nth member selected
• Starting point random
• Simple but may have periodicity issues

Non-Probability Sampling

Convenience Sampling

• Easily accessible participants
• Quick and economical
• Limited generalizability
• Common in pilot studies

Purposive Sampling

• Specific characteristics sought
• Expert judgment used
• Good for specific populations

Snowball Sampling

• Participants recruit others
• Useful for hidden populations
• Network bias possible

Quota Sampling

• Predetermined quotas filled
• Non-random within quotas
• Ensures demographic representation

Sample Size Determination

• Power Analysis: Statistical power calculation
• Effect Size: Expected magnitude of effect
• Alpha Level: Significance threshold (usually .05)
• Power: Probability of detecting effect (usually .80)
• Resources: G*Power software

Statistical Analysis

Descriptive Statistics

Central Tendency

• Mean: Average score
• Median: Middle value
• Mode: Most frequent value

Variability

• Range: Maximum - minimum
• Variance: Average squared deviation
• Standard Deviation: Square root of variance
• Interquartile Range: 75th - 25th percentile

Distribution Shape

• Skewness: Asymmetry of distribution
• Kurtosis: Peakedness of distribution
• Normal Distribution: Bell curve

Inferential Statistics

Hypothesis Testing

• Null Hypothesis (H₀): No effect/difference
• Alternative Hypothesis (H₁): Effect exists
• Type I Error: False positive (α)
• Type II Error: False negative (β)
• p-value: Probability of results if H₀ true

Common Tests

• t-tests: Compare two means
• ANOVA: Compare multiple means
• Chi-square: Categorical associations
• Correlation: Relationship strength
• Regression: Prediction models

Effect Size

• Cohen's d: Standardized mean difference
• r: Correlation coefficient
• η²: Eta squared for ANOVA
• Interpretation: Small, medium, large

Confidence Intervals

• Definition: Range of plausible values
• Common: 95% CI
• Interpretation: Precision of estimate
• Relationship: To hypothesis testing

Qualitative Research Methods

Approaches

Phenomenology

• Lived experience exploration
• Essence of phenomena
• In-depth interviews
• Bracketing researcher assumptions

Grounded Theory

• Theory development from data
• Constant comparison method
• Theoretical sampling
• Saturation point

Ethnography

• Cultural group study
• Participant observation
• Extended fieldwork
• Thick description

Case Study

• Bounded system investigation
• Multiple data sources
• Within-case and cross-case analysis
• Instrumental or intrinsic

Data Analysis

Thematic Analysis

• Identifying patterns/themes
• Coding process
• Theme development
• Interpretation

Content Analysis

• Systematic categorization
• Quantifying qualitative data
• Manifest and latent content
• Inter-rater reliability

Narrative Analysis

• Story structure examination
• Temporal sequencing
• Personal meaning-making
• Identity construction

Quality Criteria

• Credibility: Truthfulness of findings
• Transferability: Applicability to other contexts
• Dependability: Consistency of findings
• Confirmability: Neutrality of findings
• Reflexivity: Researcher self-awareness

Research Ethics

Ethical Principles

• Beneficence: Maximize benefits
• Nonmaleficence: Do no harm
• Autonomy: Respect for persons
• Justice: Fair distribution
• Fidelity: Trust and integrity

Informed Consent

• Information: Study purpose, procedures, risks
• Comprehension: Understanding verified
• Voluntariness: Free choice to participate
• Documentation: Written consent forms
• Ongoing: Right to withdraw

Special Populations

• Children: Assent and parental consent
• Vulnerable Groups: Extra protections
• Deception: Justification and debriefing
• Anonymous: No identifying information
• Confidential: Protected identity

Institutional Review Board (IRB)

• Purpose: Protect human subjects
• Review Types: Exempt, expedited, full
• Submission: Protocol, consent forms
• Monitoring: Ongoing oversight
• Reporting: Adverse events

Data Management

• Storage: Secure and encrypted
• Access: Limited to research team
• Retention: Time limits specified
• Destruction: Secure disposal
• Sharing: De-identified datasets

Reporting Research

APA Format Structure

• Title Page: Title, authors, affiliations
• Abstract: 150-250 word summary
• Introduction: Background, rationale, hypotheses
• Method: Participants, materials, procedure
• Results: Statistical analyses, findings
• Discussion: Interpretation, implications
• References: APA style citations

Writing Guidelines

• Clarity: Clear, concise language
• Objectivity: Avoid bias
• Precision: Specific details
• Organization: Logical flow
• Voice: Active preferred
• Tense: Past for method/results

Tables and Figures

• Purpose: Supplement text
• Clarity: Stand-alone understanding
• Format: APA guidelines
• Numbering: Sequential
• Captions: Descriptive titles

Statistical Reporting

• Test Statistics: Include all values
• Effect Sizes: Always report
• Confidence Intervals: When appropriate
• Format: F(2, 147) = 8.92, p < .001, η² = .11
• Rounding: Two decimal places

Reliability and Validity

No research method in psychology is useful unless its measurements are both consistent and meaningful. These two qualities are captured by reliability and validity, and together they determine whether a study's conclusions can be trusted. A measure can be highly reliable yet still invalid: a poorly designed questionnaire may produce the same biased score every time it is administered, demonstrating consistency without actually capturing the construct it claims to measure.

Reliability refers to the consistency of a measure. Common forms include test-retest reliability (the same participants score similarly when retested), internal consistency (items on a scale correlate with one another, often quantified with Cronbach's alpha), and inter-rater reliability (independent observers reach the same judgments). High reliability reduces measurement error and is a prerequisite for valid conclusions.

Validity refers to whether a study actually measures or tests what it intends to. Internal validity concerns whether observed effects are genuinely caused by the independent variable rather than confounds, and it is strengthened by random assignment and experimental control. External validity concerns how well findings generalize to other people, settings, and times. Construct validity asks whether the operational definitions truly capture the underlying psychological concept, while ecological validity reflects how closely a study mirrors real-world conditions. Strong research designs deliberately balance these forms of validity, since maximizing tight laboratory control can sometimes come at the expense of real-world applicability. These standards are equally central to psychological testing and tools such as IQ testing, where a test must be both reliable and valid before its scores can be meaningfully interpreted.