Skip to main content
. Author manuscript; available in PMC: 2014 Jun 1.
Published in final edited form as: Psychol Assess. 2012 Dec 3;25(2):643–650. doi: 10.1037/a0030912

Table 4.

Expected g Coefficients for Various Decision (D) Study Designs in Fully Crossed Homogeneous Samples (PRI)

No. of
Raters ×
No. of
Items
RCS Cognitive g
RCS Affective g
RCS Motor g
n'r n'i Grades K-l Grades 2–3 Grades 4–5 Grades K-l Grades 2–3 Grades 4–5 Grades K-l Grades 2–3 Grades 4–5
1 3 .41 .44 .54 .55 .64 .58 .51 .65 .64
6 .48 .53 .69 .60 .70 .66 .55 .68 .71
12 .53 .58 .80 .63 .73 .70 .57 .70 .74
20 .55 .61 .85 .65 .75 .72 .58 .71 .76
3 3 .66 .68 .74 .78 .83 .79 .75 .84 .82
6 .72 .75 .85 .81 .87 .84 .78 .87 .87
12 .76 .80 .91 .84 .89 .87 .80 .88 .89
20 .78 .82 .94 .84 .90 .88 .80 .88 .90
5 3 .75 .76 .80 .84 .88 .86 .83 .90 .87
6 .81 .82 .89 .88 .91 .90 .85 .91 .91
12 .84 .86 .94 .89 .93 .92 .87 .92 .93
20 .85 .88 .96 .90 .93 .93 .87 .93 .94
8 3 .81 .81 .84 .89 .91 .90 .88 .93 .91
6 .86 .87 .91 .91 .94 .93 .90 .94 .94
12 .89 .90 .95 .93 .95 .94 .91 .95 .95
20 .90 .92 .97 .93 .96 .95 .91 .95 .96
10 3 .83 .83 .86 .91 .93 .91 .90 .95 .92
6 .88 .89 .92 .92 .95 .94 .92 .95 .94
12 .91 .92 .96 .94 .96 .95 .93 .96 .96
20 .92 .93 .97 .95 .96 .96 .93 .96 .96

Note. g = predicted generalizability coefficient for a D study with the specified number of items (I) and raters (R), when items and raters are crossed with persons (P; i.e., same group of raters for each ratee); RCS = Response to Challenge Scale; n'i = number of items; n'r = number of raters. Results are based on the same Ns reported in Tables 13 for homogeneous groups.