Research methods in psychopathology - summary of chapter 4 of Abnormal Psychology by Kring, Davison, Neale & Johnson (12th edition)

Clinical psychology
Chapter 4
Research methods in psychopathology

Introduction

Theory: a set of propositions meant to explain a class observations.
Hypotheses: expectations about what should occur if a theory is true.

Approaches to research on psychopathology

The case study

Case study: recording detailed information about one person at a time.

• Lack the control and objectivity of other research methods

The case study can be used:

• To provide a rich description of a clinical phenomenon
• To disprove an allegedly universal hypotheses
• To generate hypotheses that can be tested through controlled research

The correlational method

Variables are measured as they exist in nature.
Psycho-pathologist will rely on correlational methods when there are ethical reasons not to manipulate a variable.
Comparison of people with and without diagnoses can be correlational as well.

Measuring correlation

• The first step in determining a correlation is to obtain pairs of observations of the two variables in question.
• Once such pairs of measurement is obtained, the strength of the relationship between the paired observation can be computed to determine the correlation coefficient (r).

Statistical and clinical significance

A statistical correlation is unlikely to have occurred by chance.
A non-significant correlation may have occurred by chance, so it does not provide evidence for an important relationship.

A statistical finding is usually considered significant if the probability that it is a chance finding is 5 less in 100. p<0.05.
In general, as the absolute size of the correlation coefficient increases, the result is more likely to be statistically significant.
The significance is also influenced by the number of participants in the study.

Clinical significance: whether a relationship between variables is large enough to matter.

Problems of causality

Correlational method does not allow determination of cause-effect relationship.

• Directionality problem
• Third variable problem.
  One way of overcoming the directionality problem is based on the idea that causes must precede effect.
  • Longitudinal design
    The researchers tests whether causes are present before a disorder has developed.
  • Cross-sectional design
    Measures the causes and effects at the same point in time
  • High-risk method
    Only people with above-average risk of developing a disorder would be studied.

Epidemiology: the study of the distribution of disorders in a population.
Focuses on three features of a disorder

• Prevalence:
  The proportion of people with the disorder either currently or during their lifetime
• Incidence:
  The proportion of people who develop new cases of the disorder in some period
• Risk factors:
  Variables that are related to the likelihood of developing a disorder.

Epidemiological studies are designed to be representative of the population being studied.

Three basic methods to uncover whether a genetic predisposition for psychopathology is inherited

• Comparison of members of a family
• Comparison of members of pairs of twins
• Investigation of adoptees

The starting point of family investigations is the collection of a sample of persons with the diagnoses in question.
These people are index cases or probands.
Then the relatives are studies to determine the frequency with which the same diagnosis might be applied to them.

Cross-fostering: children are adopted and reared completely apart from their biological parents.

Association study: researchers examine the relationship between a specific allele and a trait or behavior in the population.

Genome-wide association studies (GWAS): examines an entire genome of a large group to identify variation between people.

The experiment

• Random assignment
• Independent variable
• Dependent variable

Basic features of experimental design

• The investigator manipulates an independent variable
• Participants are assigned to the two (or more) conditions
• The researcher measures the dependent variable
• Different between conditions on the dependent variable are called the experimental effect

Internal validity

Internal validity: the extent to which the experimental effect can be attributed to the independent variable.
Researchers must have a control group and random assignment.

External validity

External validity: the extent to which results can be generalized beyond the study.

Treatment outcome research: designed to address if treatment works.

• A clear definition of the sample being studied
• A clear description of the treatment being offered, as in a treatment manual
• Inclusion of a control or comparison treatment condition
• Random assignment of clients to treatment or comparison conditions
• Reliable and valid outcome measures
• A large enough sample

Efficacy of a treatment: whether a treatment works under the purest of conditions.
Effectiveness: how well the treatment works in the real world.

Dissemination: the process of facilitating adaption of efficacious treatments in the community, most typically by offering clinicians guidelines about the best available treatments along with training on how to conduct those treatments.

Analogue experiments

Investigators attempt to create or observe a related phenomenon, an analogue, in the laboratory to allow more intensive study.

• In one type of analogue study, temporary symptoms are produced through experimental manipulations.
• In another type of analogue study, participants are selected because they are considered similar to people with certain diagnoses.
• Using animals

Analogue studies provide the precision of an experiment, whereas correlational studies provide the ability to study very important influences that cannot be manipulated.

Single-case experiments

Single-case experimental design: the experimenter studies how one person responds to manipulations of the independent variable.
Single-case experimental designs can have high internal validity.

From a single-case design, reversal design (or ABAB design) the participant’s behavior must be carefully measured in a specific sequence:

• An initial time period, the baseline
• A period when treatment is introduced
• A reinstatement of the conditions of the baseline period
• A reintroduction of the treatment

If the behavior in the experimental period is different from that in the baseline period, reverses when the treatment is removed, and re-reverses when the treatment is again introduced, there is little doubt that the manipulation has produced the change.
The reversal technique is most applicable when researchers believe that the effects of their manipulation are temporary.

But, external validity

Integrating the findings of multiple studies

There is not perfect research study.
A body of research studies is often needed to test a theory.

Meta-analyses