True-false tests correlate with intelligence ratings (.406) more highly than essays (.344) or written work (.255), and combining true-false tests raises the intelligence correlation while combining essays does not¶

Claim · All claims
Evidence · 1 study (2 entries) · associational r1 · q2 · i3 large
Currency · newest study 1920 · not yet checked against its sources · not reviewed by a person

Subclaims¶

q2 i3 Average correlations with intelligence are .406 for true-false, .344 for essays, and .255 for written work; the true-false-essay difference (.062) is about equal to its standard error (.067). → Arthur I. Gates 1920 q2 i? Combining several true-false tests yields a steadily increasing correlation with intelligence, whereas combining essays makes no noticeable increase. → Arthur I. Gates 1920 (2)

Evidence¶

Arthur I. Gates 1920¶

Arthur I. Gates, Teachers College, Columbia University. (1920). The True-False Test as a Measure of Achievement in College Courses. The Journal of Educational Psychology. https://eric.ed.gov/?q=true-false+test+as+a+measure+of+achievement+in+college+courses+gates

q2 · i3 · associational · r1

Correlational analysis in Tables XII-XIV of test scores against Army Alpha or Thorndike intelligence test ratings in the same classes. The printed correlations are r = .406 (true-false, N = 22), r = .344 (essays, N = 18), and r = .255 (written work, N = 9); the true-false-essay difference of .062 is about equal to its standard error of .067.

"Correlations with intelligence tests are: for the True-False + .406, <j (tr.-obt. av.) -± .032, for the essay examinations, -f+ -314, a (tr.-obt. av.) ± .04 for the written work, -j+ .255, a (tr.-obt. av.) ± .034."

Arthur I. Gates 1920 (2)¶

Arthur I. Gates, Teachers College, Columbia University. (1920). The True-False Test as a Measure of Achievement in College Courses. The Journal of Educational Psychology. https://eric.ed.gov/?q=true-false+test+as+a+measure+of+achievement+in+college+courses+gates

q2 · i? · associational · r1

Correlational analysis in Tables XVI and XVII arranged by classes: true-false averages rise from .432 (2 tests) to .545 (5 tests) while essay averages stay near .34-.35. The article reports the increase as printed in these tables; no effect size for the increase is printed.

"Combining several True-False tests produces a marked increase in the correlations with intelligence, which is not true of the essay examinations."

Discussion¶

Related Claims¶

• A single true-false test correlates with an essay examination (.33) about as highly as essay examinations correlate with each other — related
• Case studies of sharp test disagreements suggest the better the criterion of achievement, the greater the predictive value of true-false compared with essay examinations — related
• A single true-false test correlates with the criterion of achievement (.65) more highly than an essay examination (.56) or written work (.39) — related
• No method of presenting true-false questions (oral, oral-and-visual, or visual) shows superiority in criterion correlation — related
• Army intelligence test rankings corresponded with officers' rankings of soldiers (average correlation 0.536 across twelve companies), suggesting the tests measured something beyond literacy — related
• Test results correlate more highly with estimated intelligence than with grades, including Starch's .75 versus .30 at Wisconsin — related