Research methods final exam

Experimental Study

Manipulates independent variable to determine causation; Example: Assigning therapy vs. control to test depression outcomes

Observational Study

No manipulation; examines associations; Example: Studying relationship between trauma and anxiety

Case-Control Design

Compares groups with vs. without an outcome; Example: PTSD group vs. non-PTSD group and comparing childhood trauma

Cohort Design

Follows one or more groups over time; Example: Tracking adolescents exposed to violence

Cross-Sectional Design

Measures all variables at one time; Example: Measuring stress and sleep today

Retrospective Design

Looks back at past variables; Example: Adults recalling childhood adversity

Case-Control Strength

Efficient for studying rare disorders; Example: Schizophrenia risk factors

Case-Control Weakness

Recall bias and no causation; Example: Misremembered childhood trauma

Cohort Strength

Establishes temporal order; Example: Trauma occurring before depression

Cohort Weakness

Attrition over time; Example: Participants dropping out of longitudinal study

Single-Group Cohort

One group followed over time; Example: Trauma survivors tracked longitudinally

Multi-Group Cohort

Compares exposed vs. non-exposed groups; Example: Violence-exposed vs. non-exposed youth

Accelerated Cohort Design

Multiple age groups studied simultaneously; Example: Ages 10, 15, 20 followed together

Birth Cohort

Group born at same time followed over lifespan; Example: 2000 birth cohort study

Construct Specification Issue

Difficulty defining variables clearly; Example: Defining “abuse”

Group Selection Issue

Groups differ on confounding variables; Example: SES differences

Causality in Observational Studies

Cannot establish causation, only associations; Example: Trauma linked to PTSD but not causal

Construct Validity

Measure reflects intended construct; Example: Depression scale measures depression not anxiety

Reliability

Consistency of measurement; Example: Same score over time

Validity

Accuracy of measurement; Example: Test measures what it claims

Content Validity

Measure covers full construct; Example: Depression scale includes mood and sleep

Criterion Validity

Measure relates to outcome; Example: Test predicts diagnosis

Concurrent Validity

Correlates with current measure; Example: New scale matches existing one

Predictive Validity

Predicts future outcome; Example: SAT predicting GPA

Test-Retest Reliability

Stability over time; Example: Same score after two weeks

Inter-Rater Reliability

Agreement between raters; Example: Two clinicians give same rating

Internal Consistency

Items correlate within test; Example: Cronbach’s alpha

Sensitivity

Detects small changes; Example: Tracking therapy improvement

Standardized Measure (Pro)

Validated and comparable across studies; Example: Beck Depression Inventory

Standardized Measure (Con)

May not fit all populations; Example: Cultural mismatch

Modify Existing Measure

Adapt measure for new population; Example: Adult scale modified for adolescents

Develop New Measure

Create when none exists; Example: New IPV beliefs scale

Global Rating

Overall judgment of functioning; Example: Clinician rates severity

Self-Report Inventory

Participant answers questions; Example: Depression questionnaire

Projective Technique

Interprets ambiguous stimuli; Example: Rorschach test

Direct Observation

Observes behavior directly; Example: Parent-child interaction

Psychobiological Measure

Biological indicators; Example: Cortisol levels

Computerized Assessment

Digital testing methods; Example: Online surveys

Obtrusive Measure

Participants aware of being measured; Example: Acting differently when observed

Reactivity Problem

Behavior changes due to observation; Example: Participant acts nicer

Solution to Reactivity

Use unobtrusive measures or disguise purpose; Example: Hidden observation

Unobtrusive Measure

Participants unaware of measurement; Example: Archival records

Simple Observation (Pro)

Behavior is natural; Example: Watching playground behavior

Simple Observation (Con)

Little control over variables

Contrived Observation (Pro)

High control; Example: Lab interaction task

Contrived Observation (Con)

Artificial environment

Archival Records (Pro)

Easy access to existing data; Example: Medical records

Archival Records (Con)

Data may be incomplete or biased

Physical Traces

Indirect evidence of behavior; Example: Wear patterns on floor

Converging Evidence

Multiple measures agree; Example: Self-report and observation both show improvement

Incremental Validity

New measure adds unique information

Practicality

Time and cost considerations

Inconsistent Results Reason

Measures assess different constructs or error; Example: Anxiety vs. depression measures differ

Manipulation Check

Tests whether IV worked; Example: Mood induction increases sadness

IV Works + DV Changes

Strong support for hypothesis

IV Works + No DV Change

Theory not supported

IV Fails + DV Changes

Confound present

IV Fails + No DV Change

Inconclusive results

Avoid Sensitization

Use subtle or indirect checks

Exclude Participants

Usually not recommended due to validity concerns

Pilot Testing

Test manipulation effectiveness before study

Clinical Significance

Real-world meaningful change; Example: No longer meets diagnosis

Compare to Norms

Compare to healthy population averages

Dysfunctional Comparison (Pro)

Clear benchmark for improvement

Dysfunctional Comparison (Con)

Limited generalizability

Diagnostic Change (Pro)

Clinically meaningful outcome

Diagnostic Change (Con)

May be too strict

Social Impact

Effect on daily functioning; Example: Teacher reports improvement

Scope/Breadth

Change across settings; Example: Home and school

Disseminability

Ease of spreading treatment to others

Cost-Benefit Analysis

Weighs effectiveness vs. cost

Acceptability

Whether treatment is acceptable to clients

Traditional Case Study

Generates hypotheses; Example: Rare disorder case

Single-Case Design Requirements

Repeated measures, baseline, and comparison

ABAB Design

Treatment introduced and withdrawn to test effects

Multiple Baseline Design

Treatment staggered across subjects or behaviors

Changing Criterion Design

Gradual stepwise behavior change; Example: Reducing smoking

Changing Criterion Problem

External factors may influence behavior

Evaluation of Data

Visual inspection of trend, level, variability

Qualitative Research Goal

Understand lived experiences

Ethnography

Study of culture; Example: Community norms

Phenomenology

Study of lived experience; Example: Trauma experience

Grounded Theory

Develop theory from data

Qualitative Characteristics

Flexible, narrative, small samples

Qualitative vs Anecdotal

Systematic and rigorous vs informal

Qual vs Quantitative

Depth vs numerical data

Qualitative Pros

Rich, detailed data

Qualitative Cons

Subjective interpretation

Descriptive Validity

Accuracy of observations

Interpretive Validity

Accuracy of meaning

Theoretical Validity

Fit with theory

Internal Validity (Qual)

Coherence of findings

External Validity (Qual)

Generalizability

Triangulation

Use of multiple data sources

Credibility

Believability of findings

Confirmability

Objectivity of findings

Transferability

Applicability to other contexts

Mixed Methods Research

Combines qualitative and quantitative approaches

Deception

Misleading participants about study purpose; Example: Hidden hypothesis

Debriefing

Explaining study after participation