Validity and Reliability of True-False Tests

Abstract

A model of examinee behavior based on knowledge and random guessing is used to generate hypotheses about how true-false scores work. Although others have expressed reservations about the simplicity and utility of such a model, it leads to informative ideas. The confirmation of six hypotheses forms a network of support for the contention that true-false scores contain an error component (due to guessing) that makes these scores less reliable than those based on 5-choice items. Examinee response style, a propensity to favor the selection of true or false responses when the answer is unknown, can invalidate a total true-false score. When the answer key for items unknown to an examinee is unequally split between those keyed true and false, this interaction produces scores inaccurate by as much as a sample standard deviation.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 35 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.