Chapter 3: Three Claims, Four Validities — Interrogation Tools for Consumers of Research

A set of vocabulary-style flashcards covering key terms from the lecture notes on three claims, four validities, variables, operationalization, and validity concepts.

Three Claims

The main types of research claims: frequency claims, association claims, and causal claims.

Frequency claim

Describes a statistic (value, percentage, rate) for a single variable, e.g., 'Nearly 60% of teens text while driving.'

Association claim

Argues that one level of a variable is related or linked to a level of another variable.

Causal claim

Argues that one variable causes changes in another; requires covariance, temporal precedence, and ruling out alternative explanations (internal validity).

Construct validity

How well a conceptual variable is operationalized (measured or manipulated) in a study.

External validity

The extent to which results generalize to other people, contexts, times, and places.

Internal validity

The extent to which alternative explanations for a causal relationship are ruled out.

Statistical validity

The accuracy, precision, and replicability of statistical conclusions (e.g., margin of error, confidence intervals).

Operationalization

The process of turning a conceptual variable into a measurable or manipulable variable.

Conceptual variable

An abstract variable of interest; also called a construct.

Operational variable

The measurable or manipulable form of a conceptual variable (operational definition).

Measured variable

A variable that researchers observe or record rather than manipulate.

Manipulated variable

A variable that researchers assign to participants (the independent variable in experiments).

Independent variable

The variable that is deliberately varied or manipulated to observe its effect.

Dependent variable

The variable measured to assess the effect of the manipulation.

Constant

A variable that does not vary within a study (one level only).

Levels

The different settings or values a variable can take (e.g., High, Medium, Low).

Two levels

A variable with at least two levels, commonly used to compare effects (e.g., High vs Low).

Two levels (example)

For example, the marital status variable with levels such as Single and Married.

Stress levels example

An illustrative variable with levels such as High, Medium, and Low.

Operationalization example

How a concept like texting anxiety can be measured (e.g., questionnaire score, behavioral observation, etc.).

Reliability

Consistency of measurement across time or raters (e.g., test–retest, inter-rater, internal consistency).

Validity

The appropriateness and truthfulness of inferences drawn from data; includes construct, external, internal, and statistical validity.

Third-variable problem

An unmeasured variable that could cause both measured variables, threatening internal validity.

Temporal precedence

In causal claims, the cause must occur before the effect in time.

Covariance

Two variables vary together; a necessary condition for causation.

Point estimate

A single numerical estimate of a population parameter from sample data.

Confidence interval

A range around the point estimate that expresses the precision of the estimate.

Population vs. sample

Population is the entire group of interest; a sample is a subset used to infer about the population.

Generalizability

The extent to which results apply to populations beyond the study sample.

Positive association

As one variable increases, the other tends to increase.

Negative association

As one variable increases, the other tends to decrease.

Zero association

No systematic relationship between two variables.

Operational definitions

Concrete definitions of how a variable is measured or manipulated.

Correlation vs causation

Correlation (association) does not imply causation; causal claims require stronger evidence and experimental control.

Verbs for association claims

Linked, associated, correlated, predicted, tied to (used to describe associations).

Verbs for causal claims

Affects, leads to, reduces, causes, increases (used to describe causality).

Sample generalization (external validity)

How well results from a sample generalize to the population.

Reliability types

Common forms include test–retest, inter-rater, and internal consistency.

Operationalization question

Ask how researchers measured the concept and what they mean by it.

External validity sampling

How participants were selected and how well they represent the population.

Interrogation language (for claims)

Use critical questions and appropriate verbs to assess the strength of claims.