ATLASResearch
methods
Quantitative + Mixed methods/ Measurement theory

Classical test theory

Separate observed scores into true score and error

Classical test theory represents an observed score as a true score plus measurement error under a defined replication process. The true score is an expected score over hypothetical replications, not an independently known psychological truth. Reliability is a population variance ratio, and item statistics and reliability estimates can change across populations and test forms. CTT often supports familiar total-score analyses.

WHEN IT FITS

Use when evaluating a fixed test or composite through item statistics, reliability and standard errors, and a score-level measurement framework matches the intended application.

Strengths

  • Provides an accessible framework for score error and reliability
  • Supports practical evaluation of fixed tests

Limitations

  • Item and reliability statistics depend on the sample and form
  • Simple applications may obscure differing precision across trait levels

Know the boundary

A CTT true score is an expectation under replications, not proof of a person's essential ability.

USED ACROSS
Banking & financeBusiness & MBAPsychologyComputer scienceEducation