Research Methods

What does empirical mean?

Based on observation and measurable data rather than opinions or beliefs.

Why do we conduct research?

To build knowledge, test treatments, improve health policy, improve clinical practice, and become critical consumers of research.

What is a psychological theory?

A scientific explanation that organises ideas about psychological phenomena. It is based on evidence, can be tested, and makes predictions.

Directional hypothesis

Predicts that there will be a relationship or difference and predicts the direction of that relationship. Example: More sleep → lower stress.

Non-directional hypothesis

Predicts a relationship or difference but does not predict the direction. Example: There is a relationship between sleep and stress.

Alternative hypothesis (H₁)

Predicts there is a real effect, relationship or difference.

Null hypothesis (H₀)

Predicts there is no effect, relationship or difference. Researchers usually test whether there is enough evidence to reject the null hypothesis.

Directional vs Non-directional

Directional = predicts which way Non-directional = predicts only that a relationship exists

Alternative vs Null

Alternative = there is an effect. Null = there is no effect.

Deductive thinking

Starts with a theory, develops a hypothesis, then collects data to test it. Theory → Hypothesis → Data

Inductive thinking

Starts with observations, looks for patterns, then develops a theory. Data → Patterns → Theory

Deductive vs Inductive

Deductive = top-down (Theory → Data) Inductive = bottom-up (Data → Theory)

Reasonable vs Judgmental

Reasonable = uses logic and evidence before making conclusions. Judgmental = relies on opinions or hunches.

Sceptical vs Naïve

Sceptical = questions claims and looks for evidence. Naïve = accepts claims without evaluating evidence.

Objective vs Subjective

Objective = based on evidence and unbiased observations. Subjective = influenced by opinions, beliefs or emotions.

What is a variable?

Something that can change and be measured.

Independent Variable (IV)

The variable manipulated by the researcher.

Dependent Variable (DV)

The outcome measured to see whether it changes because of the IV.

Covariate / Extraneous Variable

A variable other than the IV that may influence the DV. It should be measured and controlled if possible.

Confounding Variable

An uncontrolled variable that may explain the results instead of the IV.

IV vs DV

IV = What the researcher changes. DV = What the researcher measures.

Covariate vs Confounding Variable

Covariate = measured and controlled. Confounding variable = not controlled and may bias results.

Correlation

A statistical relationship between two variables.

Positive correlation

As one variable increases, the other also increases.

Negative correlation

As one variable increases, the other decreases.

Correlation does not imply causation

Just because two variables are related does not mean one causes the other.

Causation

One variable directly causes changes in another.

Third-variable problem (Spurious correlation)

A third variable explains the relationship between two variables that appear to be related.

Operationalisation

Clearly defining exactly how a variable will be measured.

Nominal

Categories or names only. Example: Smoker / Non-smoker.

Ordinal

Ranked order but unequal distances. Example: Strongly agree → Strongly disagree.

Interval

Equal intervals but no true zero. Example: Temperature.

Ratio

Equal intervals with a true zero. Examples: Height, weight, income.

NOIR

N = Nominal O = Ordinal I = Interval R = Ratio

Validity

Does the test measure what it is supposed to measure?

Face validity

Does it appear to measure what it should?

Internal validity

Did the IV cause the changes in the DV?

External validity

Can the results be applied to the real world?

Construct validity

Does the test actually measure the psychological concept being studied?

Reliability

Does the test produce consistent results?

Test–retest reliability

Does the same person get similar results when tested again?

Split-half reliability

Do both halves of the test produce similar results?

Descriptive Designs

Describe behaviour but do not establish causation. Includes:

• Naturalistic observation
• Case studies
• Single-case designs

Naturalistic Observation

Observing behaviour in a real-world setting without interfering.

Case Study

A detailed study of one person or one unusual case. Useful for generating hypotheses and studying rare conditions.

Case Series

A report of several similar cases rather than one individual case.

Single-Case Design

Uses repeated measurements of one individual over time to evaluate an intervention.

AB Design

A = Baseline B = Intervention

ABA Design

Baseline → Treatment → Remove treatment. If behaviour changes with treatment and changes back when removed, it provides stronger evidence for causation.

Correlational Designs

Examine relationships between variables but cannot prove causation. Includes:

• Surveys
• Case-control studies

Survey

Uses questionnaires or interviews to collect information about attitudes, behaviours or relationships.

Case-Control Study

Compares people with a condition (cases) to people without it (controls) and looks back for possible risk factors.

Experimental Designs

Manipulate the IV, control extraneous variables, and compare groups to establish cause and effect.

Factorial Design

Studies two or more independent variables at the same time and examines interactions between them.

Randomisation

Randomly assigns participants to groups to reduce bias and make groups similar.

Randomised Controlled Trial (RCT)

Randomly assigns participants to treatment or control groups. Considered the strongest design for establishing causation.

Quasi-Experimental Design

Similar to an experiment but missing at least one key feature such as randomisation, manipulation, or a control group.

History Effect

An outside event influences the results rather than the intervention.

Maturation Effect

Participants naturally change over time, affecting the results.

Testing Effect

Taking the test itself changes later performance.

Instrument Decay

Measurement becomes less accurate over time.

Attrition (Mortality)

Participants drop out, making the remaining sample less representative.

Regression to the Mean

Extremely high or low scores naturally tend to move closer to the average on later measurements.