4S2

APPENDIX I

visional estimates : they are shown under the heading of ' multi-
pliers ' in the first column of Table II.

1. Insert the self-correlations, i.e. the squares of the estimated
saturation coefficients (-81, -64, etc.), in the leading diagonal of the
original table (Table I).

2. Multiply each row of correlations by the corresponding satura-
tion coefficient (e.g. 1st row: -90 X -81 = -729, -90 X 72 = -648,

I. SINGLE-FACTOR ANALYSIS

fai

A. SIMPLE-SUMMATION METHOD I rag =

TABLE I

OBSERVED INTERCORRELATIONS AND CALCULATION OF SATURATION
COEFFICIENTS

Sat. Coefis.
•90
•80
•70
•Go
•50
Totals,

•90 •80 70 •60
•50
72 -63
'54 •45
72
-56 •48 •40
•63 •56
•42
•35
•54 •48 •42
•30
"45 •40
*35 •30
2-34 2-16 1-96
174 1-50

Totals
2-34
2-16
1-96
174
1-50
970

Self-correlations : \ ist estimate J 1st completed 1 totals J Sat. Coeffs.* j 1st estimate J
-80
3'H -897
•65 2-81 -803
-50 2-46 703
'35 2-09
'597
•25
175 •500
2-55
12-25 3*50
» 3-50* (checks)

Squares Self-correlations : 1 2nd estimate J and completed 1 totals . J Sat. Coeffs.t j Final estimate J
•805 -81
3*15 •90
•645
•64
2-80 •80
•494 •49 2-45
70
•356 •36
2-10 •60
•250 -25 175 •50
2*55 12-25
= 3-50*
(checks)

Squares (for checking)
•81
-64
-49
•36
•25
2*55 (checks)

* Divisor = Vi2-25 = 3-50.

t Divisor =

s= 3-50.