Psychology Research Methods: Scientific Method, Design, Measurement, and Ethics

Falsifiable

A theory or hypothesis is scientific only if it can be tested and shown to be false

Five characteristics of science

Deterministic, empirical, falsifiable, provisional, public

Deterministic

Phenomena are causally determined by preceding events or natural laws that can be discovered and explained

Empirical

Based on measurable, objective, reproducible evidence, not pure reason, emotion, or subjective experience

Provisional

Scientific theories remain open to revision when new evidence becomes available

Public (science)

Science is a public good that benefits society and should be available to all

Pseudoscience

Appears scientific but relies on consensus-based claims, anecdotal evidence, and unfalsifiable claims, and lacks peer review and self-correction

Anecdotal evidence

Personal experience or testimonies used as evidence (example: 'The Earth is flat because my senses tell me so')

Construct

A hypothetical variable that is not directly observable (example: happiness, memory quality, intelligence)

Operational definition

A description of the procedures, tests, or actions by which a construct is observed or measured

Construct validity

The degree to which the operational definition is capable of measuring the construct of interest (a priority for all three claims)

Face validity

The apparent soundness of a measure: the items look appropriate for measuring the construct, whether or not they actually are

Participant variable

A pre-existing individual difference (age, gender, race) that is measured, not manipulated, and not under the experimenter's control

Predictor variable

An explanatory variable that predicts the outcome variable

Independent variable

A predictor variable that is manipulated and under the experimenter's control

Outcome variable

The variable that is measured (behaviors, opinions, physiological responses)

Dependent variable

An outcome variable that depends on the levels of a manipulated independent variable (experiments only)

Descriptive research design

Asks about the frequency of levels of a single variable; supports frequency/value claims

Correlational research design

Asks about the relationship between two or more measured variables; supports association claims

Experimental research design

Manipulates one variable and measures the effect on another; supports causal claims

Frequency/value claim

A claim from a descriptive study about one variable (example: 18.5% of US adults report mental health challenges)

Association claim

A claim from a correlational study that two measured variables are related

Causal claim

A claim from an experiment that one variable causes a change in another ('no causation without manipulation')

Observation (method)

Watching behavior: actions, ability, performance, archives. Also includes tests scored against a correct answer, like memory tests

Self-report (method)

Asking questions about ideas: attitudes, beliefs, feelings, memories

Physiological monitoring

Measuring biological responses or physical states, usually involuntary (example: cortisol, heart rate)

Method match

Actions = observation; ideas = self-report; physical state = monitoring

Categorical variable

Levels are groups or words that vary in quality or kind (example: on-campus vs. off-campus housing)

Continuous variable

Values are meaningful numbers along a scale (at least 6 values) that vary in quantity or amount (example: time to run a mile)

Reliability

The consistency of a measure: if you repeat the measure, do you get the same result?

Validity

The accuracy of a measure: does it actually hit the construct you are trying to measure?

Consistent scores but fails to predict the real behavior

Good reliability, poor validity

Comparing a new test to an established, validated test

Tests validity

Giving the same test again a month later

Tests reliability (test-retest)

Four big validities

Construct, internal, external, statistical

External validity

Whether results generalize to other people, settings, and times

Internal validity

Whether a causal conclusion is justified (no confounds)

Statistical validity

Whether the statistical conclusions are accurate and reasonable

Tuskegee Syphilis Study

1932 to 1972 US Public Health Service study in which men with syphilis went untreated even after penicillin became the standard treatment in the mid-1940s

Belmont Report

1978 report that set ethical principles for human research; a result of the Tuskegee study

Respect for persons

Acknowledge autonomy and protect those with diminished autonomy; people can make their own decisions

Beneficence

Do no harm; maximize possible benefits and minimize possible harms (good risk-benefit ratio)

Justice

Fair procedures and outcomes in selecting research participants; everyone should participate and benefit

Privacy and confidentiality

Not disclosing personal or identifying information about a participant beyond what the study requires

Right to withdraw

Participants can opt out at any point with no negative consequences

Informed consent

Giving people enough information before the study so they can make a voluntary choice to participate

Deception and debriefing

Researchers may deceive participants only when knowing the purpose would affect results; they must give the correct information at the end

Institutional Review Board (IRB)

An independent group that reviews research to make sure ethical standards are met

Replication

Repeating a study to test whether a finding is reliable

Direct replication

Recreates the original study's methods, materials, and procedures as closely as possible

Conceptual replication

Tests the same underlying idea using different methods, stimuli, or samples to examine generalizability

Why replication matters

Confirms findings are real vs. chance, error, or fraud; strengthens theories; supports external validity

Hyndman and Bisin (2026) result

Replication of Ariely and Wertenbroch (2002) found no significant effect of deadline type; the original article was later retracted