Corrected (disattenuated) correlations exceeding unity, when read with their probable errors, show Spearman's contention that corrected coefficients tend toward 1.00 is unsound¶

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

Subclaims¶

q2 i? Corrected coefficients such as Composition at 1.29 + 0.25 exceed unity; taking 1.00 as the most probable value ignores the probable error. → Wyman 1921

Evidence¶

Wyman 1921¶

Wyman, J. B., & Wendle, M. (1921). What is Reading Ability? The Journal of Educational Psychology, 518-531. https://eric.ed.gov/?q=%22What+is+Reading+Ability%22

q2 · i? · associational · r1

Methodological argument using the study's corrected coefficients, with probable errors computed by a Kelley-derived formula printed here for the first time. The authors argue a corrected value of 1 + a with small probable error disproves Spearman's hypothesis. No effect size is printed for this argument.

"Had Spearman known the probable errors of his corrected coeffi- cients, he would have seen the absurdity of claiming that coefficients well above 1.00 supported his argument that all corrected coefficients tended toward 1.00."

Discussion¶

Related Claims¶

• The low corrected correlation between the arithmetic exercise and the English criterion affords objective evidence that arithmetic and English ability are separate capacities — related
• MAP Growth state-test correlations exceed 0.8 for most Grade 3-8 concurrent indices, with mathematics correlations higher than reading — related