Measurement in Psychology & Intelligence Testing

25 vocabulary flashcards summarizing key terms on measurement principles—reliability, validity, standardization, and bias—in psychological testing.

Measurement in Psychology

The process of quantitatively assessing invisible constructs such as thoughts, feelings, and behaviours to predict outcomes and diagnose disorders.

Individual Differences

Continuous variations among people on physical, sensory-motor, cognitive, personality, and social functioning dimensions.

Reliability

The consistency of a test; its ability to yield the same score for the same person under identical conditions.

How do we know if a questionnaire measure is reliable ?

 

Methods to quantify reliability

Several methods available:

 

Test retest reliability

 

Alternate parallel forms

 

Split half reliability

 

Inter-rater reliability

 

Validity

The accuracy of a test; the degree to which it actually measures the construct it is intended to measure.

Major types of validity:

 

• Construct validity
  measures X, related to past measures of X

 

• Content validity
  captures all aspects of X

 

• Criterion validity
  predicts what X should theoretically predict

 

Standardization

The process of administering a test to a large, representative sample to establish norms that give meaning to individual raw scores.

STANDARDISATION

 

• Involves giving test to a large random sample to find out:

• average performance

• variation in performance around the average

• Tells us how well someone has performed

• Important to know what population the test is standardized on

Bias-freeness

The extent to which a test is fair and does not advantage or disadvantage particular groups unrelated to the construct being measured.

• Differences between groups on other factors may explain differences in test performance, not lack of ability

Test-retest Reliability

Reliability evidence obtained by administering the same test to the same group on two occasions and correlating the scores.

Alternate (Parallel) Forms Reliability

Reliability obtained by creating two equivalent versions of a test and correlating scores between the versions.

Split-half Reliability

Reliability determined by dividing a test into two halves (e.g., odd vs. even items) and correlating the two sets of scores.

Inter-rater Reliability

Consistency between independent observers’ ratings of the same behaviour, shown by high agreement or correlation.

Construct Validity

Evidence that a test relates highly to other accepted measures of the same construct.

Content Validity

The extent to which a test samples all facets of the construct it aims to measure.

Criterion Validity

The extent to which a test predicts outcomes (criteria) that theory says it should predict.

Norm Group

The large, randomly selected sample whose scores establish the mean and standard deviation used for interpreting individual results.

Mean (µ)

The average score of the norm group, used as the reference point in standardization.

Standard Deviation (σ)

A statistic that describes the spread of scores around the mean in the norm group.

Normal (Gaussian) Distribution

A bell-shaped distribution in which about 68% of scores fall within ±1σ, 95% within ±2σ, and 99% within ±3σ of the mean.

IQ Norms

Standard values for intelligence tests, typically set at µ = 100 and σ = 15.

Stanford–Binet Intelligence Scale

A widely used standardized test for measuring human intelligence, often used as a benchmark for construct validity.

Wechsler Scales

A family of standardized intelligence tests (e.g., WAIS, WISC) frequently used to assess cognitive ability across the lifespan.

Social Desirability Bias

The tendency of respondents to answer questions in a manner that will be viewed favourably by others, potentially distorting self-report data.

Observable Behaviour Coding

Recording behaviours from video or live observation for later analysis, often requiring high inter-rater reliability.

Predictive Use of Tests

Employing psychological measures to forecast behaviours or life outcomes such as academic performance or job success.

Continuous Dimension

A variable on which individuals differ by degree (e.g., height, reaction time) rather than by discrete categories.

Measurement Bias

Systematic factors causing certain groups to score higher or lower on a test for reasons unrelated to the construct being assessed.