Combinations of two measure types correlate very highly with the composite (.96, .91, .98), with little difference between choices¶

Claim · All claims
Evidence · 1 study · associational r1 · q2 · i3 large
Currency · newest study 1921 · not yet checked against its sources · not reviewed by a person

Subclaims¶

q2 i3 Stanford Revision plus group tests correlated .96 with the composite, Stanford Revision plus teachers' estimates .91, and group tests plus teachers' estimates .98 (30 cases each); under the survey's conditions it would make little difference which two-type combination was selected. → Chassell 1921

Evidence¶

Chassell 1921¶

Chassell, C. F., & Chassell, L. M. (1921). A survey of the three first grades of the Horace Mann School by means of psychological tests and teachers' estimates, and a statistical evaluation of the methods employed. Part II. The Journal of Educational Psychology, 12(5), 243-252. https://doi.org/10.1037/h0068285

q2 · i3 · associational · r1

Correlational comparison of two-type combinations with the composite in the same survey; the printed correlations r = .96, r = .91, and r = .98 are large effects. The article attributes the high correlations to common elements: the four elements of two types are "identical with four out of the six elements in the composite", which also lowers the regression values (.28, .41, .20).

"Thus the Stanford Revision combined with the group tests gives a correlation of .96 with the composite, the Stanford Revision combined with the teachers' estimates, of .91, and the group tests combined with the teachers' estimates, of .98," (30 cases being included in each instance)."

Discussion¶

Related Claims¶

• Group tests combined with teachers' estimates correlate .77 with the Stanford Revision, questioning the necessity of individual examination — related
• Group tests correlate much lower with the Stanford Revision (.45 and .53) and only .37 with each other — related
• Any single measure type approximates the composite (.89, .87, .85) but has only limited value used alone — related
• Teachers' estimates show the highest single-measure correlation with the Stanford Revision (.72, 60 cases) — related