Research Methods

What is research?

Use of the scientific method (formulating and testing hypotheses) to understand phenomena and discover universal laws and principles

Research in clinical psychology does not necessarily involve assessment, diagnosis, or treatment-but it can and often does

Respect for persons: Informed consent

The process through which an individual voluntarily agrees to something-whether that’s participation in a research study, medical care, or psychotherapy

In research, must provide clear and understandable information about the purpose, risks, benefits, and alternatives to participating

The participant must have the capacity to understand the information they are provided, ask questions, and freely provide consent without fear of negative consequences

  • Some folks can’t independently consent - minors, those with reduced decision-making ability (intellectual disability, or non-English speaking)

  • Other groups, such as prisoners, require special protections to ensure full informed consent without coercion

Research questions

What causes symptoms of mental illness?

Do treatments for mental illness work? How do they work?

How many people have a mental illness?

  • Prevalence- number of cases in a population

  • Incidence- number of new cases that emerge within a specific period of time

Research designs

Cross-sectional: all participants complete research procedures once

Longitudinal: all participants complete research procedures at least twice, but often multiple times

Case study: comprehensive description of one individual or a small group of individuals

Research designs: Pros and cons

Cross-sectional: quicker, easier, and cheaper but cannot provide evidence of causation in many cases

Longitudinal: can provide stronger evidence in support of casual processes but is often expensive, difficult to conduct, and suffers from high attention

Case study: useful for studying rare phenomenon and generating hypotheses but findings cannot be generalized to others

Correlations

Correlation coefficient’s range from -1.0 to 1.0

The value of the coefficient describes the strength of the relationship (e.g., 0.40 vs 0.75)

+/- describes the direction of the relationship (e.g., -0.50 vs 0.50)

p values

The p-value associated with a correlation coefficient, or any other statistic, indicates whether the result is statistically significant

p < 0.05 means there’s less than a 5% probability that results are due to chance, thus statistically significant

it’s important to remember that “statistically” significant doesn’t always mean “clinically” meaningful

Establishing causation

Correlational studies suffer from uncontrolled confounds

Establishing causation requires an experimental design consisting of:

  1. An independent (IV) and dependent variable (DV)

  2. An experimental group - participants exposed to IV

  3. A control group - participants not exposed to IV

Importantly, group assignment should be random

Adoption Studies

Studies that compare an adopted person’s traits or symptoms to both their biological and adoptive parents

Traits like the biological parents? Must be the result of genetics

Traits like the adoptive parents? Must be the result of the environment

Limitations:

  • Adoption isn’t random

  • Adoption doesn’t always occur at birth

Twin studies

Monozygotic twins (MZ) - identical

Dizygotic twins (DZ) - fraternal

MZ twins raised together and have different behavior? Must be the environment

MZ twins raised separately and have the same behavior? Must be genetics