Notes on Chapter 4: Research Methods in Psychopathology
Learning Objectives
L04.1 Describe the goals of research as they relate to psychopathology.
L04.2 Compare the uses of case studies, correlational research, and experimental research as research methods used by psychologists to understand mental disorders.
L04.3 Distinguish between cross-sectional and longitudinal research designs as research strategies that examine psychopathology across time.
L04.4 List the ways in which ethical considerations are identified and addressed in modern psychological research.
Examining Psychopathology: The Role of Research Methods
Behavior scientists study human behavior with the scientific method, just as other sciences study volcanic activity or brain effects of cell phone use.
Psychopathology involves complexity due to interactions between biological and psychological dimensions.
Many questions lack simple answers (e.g., causes of hallucinations, best treatment for suicidality).
Important to study indirectly because we cannot directly access people’s minds; researchers develop ingenious methods to study what behaviors constitute problems.
Understanding research methods is essential for everyone; helps distinguish fact from fiction and avoid fad treatments.
Real-world questions people ask about mental health (e.g., aggression in children, sunlight and depression) motivate the need for good research literacy.
Being a good consumer of research helps determine when questions are answered by solid evidence versus sensationalism.
Important Concepts
Three general aims:
What problems cause distress and impairment?
Why do people behave unusually?
How can we help people behave in more adaptive ways?
Basic research process components:
Hypothesis: an educated guess about what will be found.
Research design: plan for testing the hypothesis; determines what is measured and manipulated.
Dependent variable: what is measured and expected to change.
Independent variable: what is manipulated or varied to influence the dependent variable.
Internal validity: whether changes in the dependent variable can be attributed to the independent variable.
External validity: generalizability of findings to other populations/settings.
All research strategies share these elements; Table 4.1 (basic components) summarizes descriptions.
Basic Components of a Research Study (Table 4.1) – Key Elements
Hypothesis: an educated guess or statement to be supported by data.
Research design: plan for testing the hypothesis.
Dependent variable: aspect measured and expected to change.
Independent variable: factor manipulated or thought to influence the change.
Internal validity: confidence that the independent variable caused the change in the dependent variable.
External validity: generalizability of results beyond the immediate study.
Hypothesis, Variables, and a Concrete Example
Hypotheses are testable; e.g., comparing CBT programs vs treatment-as-usual (TAU) for anxiety in children with ASD.
An example study (Wood et al., 2020) involved 214 children aged 7–13 who were randomized into three groups (two CBT programs and TAU).
167 randomized; 145 completed; 22 withdrew; eligibility criteria included confirmed ASD.
In the Wood et al. study:
Dependent variable: anxiety levels measured by Pediatric Anxiety Rating Scale and other measures.
Independent variables: the three treatments (two CBT modalities and TAU).
Interpretation: adapted CBT showed additional advantages (e.g., improved social-communication and adaptive functioning) beyond traditional CBT.
Concept: testability is crucial to science because it allows conclusions about what treatment modality best improves anxiety in ASD.
Internal and External Validity; Confounds
Internal validity concerns confounds: other factors that could influence the dependent variable.
Example: IQ differences across groups could confound anxiety outcomes in CBT studies.
Confounds are factors that systematically differ among groups and affect results.
Strategies to enhance internal validity:
Control groups: match experimental and control groups on relevant factors; isolate the treatment effect.
Randomization: equal chance of assignment to groups; minimizes selection bias and balances group characteristics.
Analogue models: laboratory analogues mimic real-world phenomena to increase internal validity.
Balance with external validity: increasing internal validity can reduce generalizability; often solved by multiple related studies.
Statistical vs Clinical Significance
Statistical significance: probability that observed differences are not due to chance (math-based).
Clinical significance: whether the size of the effect is meaningful for those affected.
Example (anorexia nervosa): a study found a significant difference in smoothie consumption after positive vs neutral mood induction, but the clinical meaning is questionable because levels did not reach nondisordered comparison.
Effect size: measures the magnitude of the treatment effect; more informative than p-values alone; methods for calculating effect size consider individual differences, not just group averages.
Social validity (Wolf, 1978): assessment of the importance of changes from the perspective of the person treated and significant others.
The danger of the “average” client (Kiesler, 1966): patient uniformity myth; group means can mask important individual differences.
Implication: practitioners must consider heterogeneity and not assume treatments that are statistically significant will work for every individual.
Concept Check 4.1 – Quick Exercises (Answers Provided in Summary)
1. Independent variable
2. Confound
3. Hypothesis
4. Dependent variable
5. Internal validity, External validity
Types of Research Methods
Researchers study behavior using several forms:
Case studies
Correlational research
Experimental research
Single-case experimental studies
Studying Individual Cases (Case Studies)
Case study method involves intensive study of one or more individuals with a disorder.
Not controlled; limited internal validity; susceptible to confounds; limited external validity.
Historical importance: Freud, Breuer, Masters & Johnson, Wolpe scored pivotal early work via case studies.
Limitations:
Coincidental factors may be mistaken for causal factors.
Media sensationalization can bias interpretation.
Case studies help generate hypotheses but do not establish causality.
Research by Correlation
Correlation assesses whether two variables relate to each other; does not imply causation.
Correlation coefficient r quantifies strength and direction of a linear relationship.
Example: Positive correlation where variables rise together; Negative correlation where one rises as the other falls.
r values range from
with stronger relationships as |r| increases.
Directionality problem: correlation does not reveal which variable causes the other or whether a third variable causes both.
Epidemiology: correlational research focused on incidence, prevalence, and distribution of disorders in populations.
Examples:
COVID-19 epidemiology identifies vulnerable groups (older adults, preexisting conditions).
Prevalence vs incidence definitions:
Prevalence: number of people with a disorder at a given time.
Incidence: number of new cases during a period.
Examples: binge drinking prevalence among U.S. college students ~40%; incidence of binge drinking changing over time.
Epidemiology is valuable for directing research directions even though it cannot conclusively establish causation.
Research by Experiment: Causality
Experiments involve manipulating an independent variable and observing effects to infer causality.
If a correlation exists between social supports and disorders, experiments can test whether increasing supports changes disorder prevalence.
Group Experimental Designs involve treatment of a group and follow-up to assess changes; a key challenge is what would happen without treatment (counterfactual).
Clinical trials: a formal, systematic evaluation of treatment effectiveness and safety following standardized protocols.
Randomized clinical trials: participants randomly assigned to groups.
Controlled clinical trials: include a control condition.
Randomized controlled trials (RCTs): both randomization and control conditions.
Control Groups, Placebo, and Double-Blind
Control groups: similar participants who do not receive the experimental treatment, used to rule out alternative explanations.
Placebo effect: improvement due to expectation of treatment; addressed with placebo control groups.
Placebo control groups in psychology: give some form of the actual treatment (e.g., homework) without the active component to control for placebo/expectation.
Double-blind control: neither participants nor researchers know which group participants are in; reduces allegiance bias.
Allegiance effect: researchers’ expectations may bias outcomes; double-blind minimizes this risk.
Limitations: even with double-blind, some treatments (e.g., meds with side effects) may reveal group assignment.
Comparative Treatment Research; Process vs Outcome
Comparative treatment research compares different treatments across comparable groups.
Process research asks: Why does it work?; looks at mechanisms of change (biological mechanisms, cognitive processes, etc.).
Outcome research asks: What changes occur?; overall effectiveness.
Example: relaxation training may not speed progress; cognitive restructuring and exposure could be more critical components in youth anxiety treatment.
Single-Case Experimental Designs
BF Skinner contributed single-case experimental designs: systematic study of individuals under varying conditions.
Rationale: depth of information about an individual can be more informative than averages from many individuals.
Key features:
Repeated measurements: multiple assessments before and after intervention to assess level, variability, and trend.
Withdrawal designs: treatment is withdrawn to test if effects revert; ethical and practical concerns exist.
Multiple baseline design: treatment started at different times across settings/behaviors/individuals to infer causality without withdrawal.
Wendy example (anxiety): repeated measures show variability and trend; helps distinguish true treatment effects from natural fluctuations.
Advantages: does not require withdrawal; mirrors real-world clinical practice.
Disadvantages: external validity concerns due to small Ns; generalizability can be improved by conducting multiple baselines across several individuals.
Example study (Durand, 1999; Durand & Hieneman, 2008; Durand et al., 2013): functional communication training reduced challenging behaviors in children with autism and improved communication.
Genetics and Behavior Across Time and Cultures
Genetics view: behavior arises from gene-environment interactions; genetic endowment interacts with experiences.
Four broad approaches to gene-environment influences:
Basic genetic epidemiology
Advanced genetic epidemiology
Gene finding
Molecular genetics
Table 4.2 outlines these approaches and questions addressed:
Basic genetic epidemiology: Is the disorder inherited; estimates heritability via family/twin/adoption studies.
Advanced genetic epidemiology: If inherited, what factors influence the disorder (developmental timing, sex differences, gene-environment interactions).
Gene finding: Location of the gene(s) influencing the disorder via linkage/association studies.
Molecular genetics: Biological function of the genes and how they affect disorders.
Family, Adoption, and Twin Studies; Endophenotypes
Family studies: assess whether a trait runs in families; proband is the affected individual studied.
Familial aggregation: e.g., blood-injury-injection phobia shows ~60% aggregation in first-degree relatives for this disorder.
Adoption studies: separate environmental from genetic influences by examining siblings raised in different homes.
Twin studies: compare identical (monozygotic) vs. fraternal (dizygotic) twins to estimate heritability.
Identical twins share ~100% of genes; fraternal twins share ~50%; concordance for traits can indicate genetic influence.
Epigenetic markers and differences can occur even in identical twins.
Classic example: antisocial behavior studies in Vietnam-era twins show greater similarity in identical twins, especially in adulthood, suggesting genetic factors gain influence over time.
Combining adoption and twin methods helps separate nature and nurture.
Genetic linkage analysis and association studies are used to locate specific genes:
Genetic linkage analysis uses genetic markers to find co-segregation with a disorder within families.
Association studies compare marker frequencies in people with versus without the disorder; can identify markers near implicated genes.
Caution: replication is often needed; single linkage findings may not replicate across families; avoid overgeneralizing from initial linkage claims.
Cross-Time Behavior; Time Trends and Prevention
Longitudinal studies examine how disorders/behaviors change over time; prospective designs follow individuals forward in time.
Cross-sectional designs compare different age groups at one time; potential cohort effects exist.
Sequential designs combine cross-sectional and longitudinal elements (following several cohorts over time).
Prevention research aims to reduce risk factors and prevent disorders:
Health promotion/positive development: broad, population-wide skill-building (e.g., Seattle Social Development Program).
Universal prevention: targets whole populations at risk factors present in that population.
Selective prevention: targets groups at higher risk (e.g., children with parent deaths).
Indicated prevention: targets individuals showing early signs of problems (e.g., depressive symptoms).
Across time, prevention strategies use both correlational and experimental designs to evaluate effectiveness.
Cross-cultural research emphasizes studying psychopathology across cultures to avoid ethnocentrism and to understand culture-specific presentations and treatments.
Cross-cultural considerations include differences in symptom expression, thresholds for pathology, and treatment models (e.g., family models in Japan, religio-medical integration in some Middle Eastern contexts).
Power of a Program of Research; Replication and Ethics
A program of research uses multiple studies with different designs to build a coherent understanding of a problem.
Replication increases credibility; findings replicated across methods and samples reduce likelihood of spurious conclusions.
Ethical considerations are central to research: informed consent, IRB approval, minimizing harm, confidentiality, deception/debriefing, and participant welfare.
Informed consent: competent, voluntary, fully informed, and comprehended by participants; difficult in certain populations (children, cognitive impairments).
IRB (Institutional Review Board): protects participants’ rights in university/medical settings.
APA Ethics Code guides research conduct; emphasizes participant welfare and scientific integrity.
Participatory action research: involving consumers (people with lived experience) in design, implementation, and interpretation to improve relevance and ethical standards.
Concept Check 4.3 – Self-Assessment (Part A and Part B)
Part A (designs):
1. Longitudinal (L)
2. Cross-sectional (CS) or Longitudinal (L) depending on context
3. Longitudinal (L)
4. Cross-sectional (CS)
5. Longitudinal (L)
6. Cross-sectional (CS)
Part B (true/false):
7. True (consent required after informing participants)
8. False (in some designs, control/placebo may require consent; all participants should consent or assent)
9. True (IRB approval required in many settings)
True (confidentiality rights apply)
False (when deception is essential, debriefing is typically required)
Genetics and Behavior Across Time and Cultures – Key Concepts
Phenotypes vs genotypes:
Phenotype: observable characteristics or behavior.
Genotype: the genetic makeup (e.g., Down syndrome: phenotype includes intellectual disability; genotype is extra chromosome 21).
Human Genome Project (1990–): mapped ~25,000 human genes; identified numerous genes contributing to inherited disorders.
Endophenotypes: genetic mechanisms underlying symptoms; e.g., genes influencing working memory problems in schizophrenia.
Family Studies, Adoption Studies, Twin Studies – Details
Family studies examine trait prevalence within families; proband is the starting case.
Adoption studies separate environmental from genetic influences by comparing siblings raised apart.
Twin studies leverage contrasts between monozygotic and dizygotic twins to estimate heritability and environmental contributions.
Key example: antisocial behavior shows higher similarity in identical twins than fraternal twins; suggests genetic factors increase influence with age, while juvenile antisocial traits show stronger environmental influence.
Genetic Linkage Analysis and Association Studies – Locating Genes
Linkage analysis uses known genetic markers to detect co-segregation with a disorder within families; if a marker tracks with the disorder, the implicated gene is near that marker.
Example: bipolar disorder studied in large Amish family suggested linkage on chromosome 11; replication attempts in other families were not consistently successful.
Association studies compare marker frequencies in affected vs. non-affected individuals to identify markers near causal genes; often more robust for weak associations.
Caution: findings require replication and cannot assume a single gene is responsible for complex disorders.
Studying Behavior Over Time; Prospective vs Retrospective; Prevention
Prospective (longitudinal) studies track changes as they occur; retrospective studies rely on participants’ memory, which can be biased.
Prevention research uses time-based designs to test strategies that reduce risk or delay onset of disorders.
Protective factors across the life course (education, exercise, social engagement, managing hearing loss, depression, diabetes, obesity) can slow or prevent dementia and other disorders.
Cross-Cultural Research and Challenges
Western-centric research may miss cultural variations in symptomatology and treatment acceptance.
Cross-cultural studies must account for variations in symptom expression, thresholds for pathology, and cultural treatment norms.
Cultural factors can influence how disorders are described and treated (e.g., medication adherence, family involvement in care).
Power of Replication and Ethical Considerations
Replication strengthens confidence in findings; one study is rarely enough to establish truth.
Ethical considerations include balancing internal validity with participant welfare; informed consent and debriefing are key.
Participatory approaches involve patients/consumers in designing and interpreting research to improve relevance and ethics.
Key Terms (Glossary Highlights)
hypothesis: an educated guess to be tested.
research design: plan for testing the hypothesis.
dependent variable: what is measured and expected to change.
independent variable: what is manipulated.
internal validity: confidence that changes are due to the IV.
external validity: generalizability of results.
testability: the extent to which a hypothesis can be tested.
confound / confounding variable: a factor that can distort the causal inference.
control group: a group not receiving the experimental manipulation for comparison.
randomization: assigning participants to groups by chance to reduce bias.
analogue models: laboratory analogues of real-world phenomena.
generalizability: extent findings apply to populations beyond the sample.
statistical significance: likelihood that observed effects are not due to chance.
clinical significance: whether the effect size is meaningful in real-world terms.
effect size: magnitude of the treatment effect.
patient uniformity myth: averaging across participants can obscure individual differences.
case study: in-depth study of one or a few cases; limited experimental control.
correlation: a statistical relation between two variables.
positive correlation: as one variable increases, the other tends to increase.
correlation coefficient: measuring relationship strength/direction.
negative correlation: as one variable increases, the other tends to decrease.
directionality: uncertainty who causes whom in a correlation.
epidemiology: study of incidence, prevalence, and consequences of disorders in populations.
experiment: manipulation of an independent variable to observe effects.
placebo effect: improvement due to expectancy rather than the treatment itself.
placebo control group: used to parse placebo effects from genuine treatment effects.
double-blind control: neither participants nor researchers know group assignments to prevent bias.
comparative treatment research: comparing different treatments.
process research: examining how a treatment works.
outcome research: examining what changes occur following treatment.
single-case experimental design: rigorous single-subject research with controls.
repeated measurement: measuring behavior multiple times to assess level, variability, and trend.
variability: fluctuations in behavior across time.
trend: direction of change over time.
withdrawal design: remove treatment to test causality.
baseline: initial measurement before intervention.
multiple baseline: stagger treatment across settings/behaviors/participants to infer causality without withdrawal.
phenotype: observable traits or behaviors.
genotype: genetic makeup of an individual.
human genome project: mapping of human genes (about 25,000).
endophenotype: genetic mechanism contributing to core symptoms.
family studies: examine traits within families.
proband: the affected individual in a family study.
adoption studies: compare relatives raised apart to separate genetic/environmental influences.
twin studies: compare monozygotic and dizygotic twins to estimate genetic contributions.
genetic linkage analysis: identify markers linked to a disorder within families.
genetic markers: DNA sequences used to track inheritance.
association studies: compare markers in affected vs. unaffected groups to locate genes.
cross-sectional design: compare different age groups at one point in time.
cohorts: groups of people born around the same time.
cohort effect: confounding of age and experiential differences.
longitudinal design: following the same individuals over time.
cross-generational effect: differences across generations when generalizing findings.
sequential design: combines cross-sectional and longitudinal strategies.
informed consent: participants’ voluntary agreement with full information about the study.
IRB: institutional review board ensuring participant protections.
participatory action research: involving consumers in research design and interpretation.
Concept Check 4.2 – Answers Snapshot
1. e (experiment)
2. c (randomized clinical trials)
3. b (correlation)
4. a (case study)
5. f (single-case experimental design)
Concept Check 4.3 – Answers Snapshot
Part A:
1. L
2. CS
3. L
4. CS
5. L
6. CS
Part B:
7. T
8. F
9. T
T
F
Summary
Research in psychopathology uses hypotheses, designs, and variables to understand disorders, causes, and treatments.
Case studies offer depth but limited experimental control; correlations show relationships but not causation; experiments test causality.
Epidemiology tracks incidence/prevalence in populations; cross-sectional and longitudinal designs examine development across time, with sequential designs offering a hybrid approach.
Genetics research progresses from establishing heritability to locating specific genes (linkage/association) and exploring endophenotypes.
Prevention research emphasizes health promotion, universal/selective/indicated approaches.
Cultural factors influence symptom presentation, thresholds for pathology, and treatment approaches; cross-cultural research aims to broaden understanding beyond Western norms.
Ethics: informed consent, IRB oversight, and consumer involvement (participatory action research) strengthen protection and relevance.
Key Terms (Consolidated)
hypothesis, 103; research design, 103; dependent variable, 103; independent variable, 103; internal validity, 103; external validity, 103; testability, 104; confound, 104; confounding variable, 104; control group, 105; randomization, 105; analogue models, 105; generalizability, 105; statistical significance, 105; clinical significance, 105; effect size, 106; patient uniformity myth, 106; case study method, 106; correlation, 107; positive correlation, 107; correlation coefficient, 107; negative correlation, 108; directionality, 108; epidemiology, 108; experiment, 109; placebo effect, 110; placebo control groups, 110; double-blind control, 110; comparative treatment research, 110; single-case experimental designs, 111; repeated measurement, 111; variability, 112; trend, 112; level, 112; withdrawal design, 113; baseline, 113; multiple baseline, 113; phenotypes, 115; genotypes, 115; human genome project, 115; endophenotypes, 115; family studies, 116; proband, 116; adoption studies, 116; twin studies, 116; genetic linkage analysis, 116; genetic markers, 117; association studies, 117; cross-sectional design, 118; cohorts, 118; cohort effect, 118; retrospective information, 118; longitudinal designs, 118; cross-generational effect, 119; sequential design, 119; informed consent, 121; Answers to Concept Checks, 4.1; 4.2; 4.3; 123