Constant‐Ratio Rule for Confusion Matrices in Speech Communication

1957 ◽  
Vol 29 (6) ◽  
pp. 715-720 ◽  
Author(s):  
Frank R. Clarke
Keyword(s):  
Speech Communication ◽  
Constant Ratio ◽  
Ratio Rule ◽  
Confusion Matrices

Further Test of the Constant‐Ratio Rule in Speech Communication

1957 ◽  
Vol 29 (12) ◽  
pp. 1318-1320 ◽  
Author(s):  
Frank R. Clarke ◽  
Clint D. Anderson
Keyword(s):  
Speech Communication ◽  
Constant Ratio ◽  
Ratio Rule

The constant-ratio rule and visual displays.

1961 ◽  
Author(s):  
Milton H. Hodge ◽  
Morris J. Crawford ◽  
Mary L. Piercy
Keyword(s):  
Constant Ratio ◽  
Visual Displays ◽  
Ratio Rule

scholarly journals Evaluation of the constant ratio rule for consonant confusions

1987 ◽  
Vol 81 (S1) ◽  
pp. S3-S3
Author(s):  
Theodore S. Bell ◽  
Donald D. Dirks ◽  
Gail E. Kincaid
Keyword(s):  
Constant Ratio ◽  
Ratio Rule

Four-alternative choice violates the constant-ratio rule

2010 ◽  
Vol 84 (1) ◽  
pp. 381-389 ◽  
Author(s):  
Douglas Elliffe ◽  
Michael Davison
Keyword(s):  
Constant Ratio ◽  
Ratio Rule ◽  
Alternative Choice

scholarly journals Some further tests of the constant-ratio rule

1967 ◽  
Vol 2 (10) ◽  
pp. 429-437 ◽  
Author(s):  
Milton H. Hodge
Keyword(s):  
Constant Ratio ◽  
Ratio Rule

Confusion Matrix Analysis for Form Perception

1967 ◽  
Vol 9 (5) ◽  
pp. 439-446 ◽  
Author(s):  
Raymond D. Engstrand ◽  
George Moeller
Keyword(s):  
Confusion Matrix ◽  
Matrix Analysis ◽  
Form Perception ◽  
Experimental Conditions ◽  
Geometric Figures ◽  
Field Response ◽  
Ratio Rule ◽  
Visual Form Perception ◽  
Confusion Matrices ◽  
Numeric Response

The Constant-Ratio Rule (CRR), an empirical technique for analysis of confusion matrices, was developed for use in predicting intelligibility of speech syllables. This study investigated the validity of the rule when applied to the data from experiments on visual form perception. English letters and simple geometric figures were tachistoscopically presented in the center of a viewing field. Response proportions for subsets of this master set of stimuli were predicted by CRR. Results indicated that the rule (1) accurately predicted numeric response proportions for subsets of stimuli when experimental conditions were similar and (2) predicted ordinally accurate data when experimental conditions varied within the limit which might be encountered in “operational situations.” These results, as well as arithmetic factors which can result in errors in prediction, are discussed.

Sign in / Sign up

Export Citation Format

Share Document