Research Methods & Scientific Processes

Laboratory Experiment

Highly controlled environment; IV is manipulated. +High internal validity, −Low ecological validity.

Field Experiment

IV manipulated in a natural setting. +More realistic, −Less control of extraneous variables.

Natural Experiment

IV occurs naturally, not manipulated. +Useful when unethical to manipulate IV, −No random allocation.

Quasi-Experiment

IV based on existing differences (e.g., gender). +Can study unique groups, −No causal conclusions.

Naturalistic Observation

Behaviour studied in natural environment. +High ecological validity, −Lack of control.

Controlled Observation

Some variables are controlled by researcher. +Easier to replicate, −Lower ecological validity.

Covert Observation

Participants unaware they're being observed. +Less demand characteristics, −Ethical concerns.

Overt Observation

Participants know they're being watched. +Ethically acceptable, −Demand characteristics.

Participant Observation

Researcher becomes part of the group. +Greater insight, −Researcher bias.

Non-Participant Observation

Researcher remains separate. +Objective, −Less insight into behaviour.

Questionnaires

Set of written questions. +Efficient data collection, −Social desirability bias.

Interviews

Structured: fixed questions; Unstructured: flexible. +Rich data (unstructured), −Time-consuming.

Correlations

Analyse relationship between co-variables. +Can identify links, −No cause-and-effect.

Difference Between Correlational Studies and Experiments

Experiments manipulate variables | Correlational studies do not—they measure co-variables, correlations can't show causation.

Content Analysis

Systematic coding of qualitative data into categories. +Can convert qualitative to quantitative, −Subjective interpretation.

Case Studies

In-depth investigation of one person/small group. +Rich detail, −Low generalisability.

Aim

Statement of study's purpose. Distinct from hypothesis, which is a testable prediction.

Hypothesis

Directional: predicts specific effect; Non-directional: predicts a difference without stating direction.

Population vs. Sample

Population = larger group of interest; Sample = smaller group studied to represent population.

Random Sampling

Everyone has equal chance. +Unbiased, −Time-consuming.

Systematic Sampling

Every nth participant. +Objective, −May not be representative.

Stratified Sampling

Proportional representation of subgroups. +Highly representative, −Complex.

Opportunity Sampling

Those available at time. +Convenient, −Biased sample.

Volunteer Sampling

Participants self-select. +Easy, −Volunteer bias.

Sampling Bias

When the sample is not representative. Affects generalisability.

Pilot Studies

Small-scale practice run to identify issues in procedures, measures, etc.

Experimental Designs

Repeated Measures, Independent Groups, Matched Pairs.

Repeated Measures

Same participants in each condition. +No participant variables, −Order effects.

Independent Groups

Different participants in each condition. +No order effects, −Participant variables.

Matched Pairs

Different but matched participants. +Controls participant variables, −Hard to match accurately.

Observational Design

Use of behavioural categories, event sampling, and time sampling.

Behavioural Categories

Clearly defined behaviours to record during observation.

Event Sampling

Counting each time a behaviour occurs. +Good for infrequent behaviour, −May miss behaviours.

Time Sampling

Recording behaviour at regular intervals. +Reduces overload, −May miss important events.

Open Questions

Allow detailed, qualitative responses. +Rich data, −Harder to analyse.

Closed Questions

Fixed responses (e.g., yes/no). +Easy to quantify, −May lack depth.

Independent Variable (IV)

Variable that is manipulated.

Dependent Variable (DV)

Variable that is measured.

Extraneous Variables

Uncontrolled variables that may affect the DV.

Confounding Variables

Variables that systematically affect the DV and interfere with results.

Operationalisation

Making variables measurable (e.g., "aggression = number of punches").

Random Allocation

Ensures each participant has equal chance in each condition. +Reduces bias.

Counterbalancing

Controls order effects in repeated measures design (e.g., ABBA).

Randomisation

Using chance to reduce investigator bias in design (e.g., random order of trials).

Standardisation

Keeping procedures the same for all participants. +Improves reliability.

Demand Characteristics

Participants guess aim and change behaviour.

Investigator Effects

Researcher unconsciously influences results.

Ethical Issues

Informed consent, deception, protection from harm, right to withdraw, confidentiality.

BPS Code of Ethics

Provides guidelines for ethical psychological research and how to address issues.

Dealing with Ethics

Debriefing, right to withdraw, anonymity, ethical committees.

Peer Review

Process of assessing research before publication to check quality and credibility.

Psychology and the Economy

Research can inform public policy, education, mental health services (e.g., reducing absence, improving treatments).

Reliability

Consistency of a measure. Includes test-retest and inter-observer reliability.

Test-Retest

Measure tested twice. High correlation = reliable.

Inter-Observer Reliability

Two observers' records compared. High agreement = reliable.

Improving Reliability

Standardised procedures, clearer behavioural categories, training observers.

Face Validity

Does the test appear to measure what it's supposed to?

Concurrent Validity

Does the test correlate with a well-established measure?

Ecological Validity

Extent findings generalise to real life settings.

Temporal Validity

Extent findings apply over time.

Improving Validity

Use of standardisation, control of EVs, blind procedures, realistic tasks.

Objectivity

Unbiased research. Improves scientific credibility.

Empirical Method

Gathering evidence through observation and experiment.

Replicability

Ability to repeat study and get same results.

Falsifiability

Scientific theories must be testable and disprovable.

Theory Construction

Building explanations based on evidence.

Paradigm

A shared framework of assumptions in a scientific field.

Paradigm Shift

When a dominant theory is replaced following contradictory evidence.

Scientific Report Sections

Abstract, Introduction, Method (design, sample, procedure), Results, Discussion, References.