94
AHMED M. ABDEL-KHALEK
Table 3. Loadings of the two factors after oblique rotation by Promax among the 16 groups†
Salient saturations
Factor I
Factor II
Sample
No.
N(A)
N(B)
C
E(A)
E(B)
R
1.
919
913
940
907
830
872
2.
919
924
945
975
864
913
3.
909
936
952
880
754
833
4.
921
947
957
923
849
874
5.
879
940
970
950
882
839
6.
932
949
935
878
816
822
7.
910
919
944
875
672
687
8.
916
924
953
933
829
864
9.
957
931
956
904
880
905
10.
906
929
940
851
926
898
11.
872
912
865
898
839
864
12.
917
920
827
785
838
946
13.
968
903
845
859
887
931
14.
906
934
942
822
726
849
15.
949
950
953
870
785
821
16.
884
899
906
895
815
941
* Decimal points omitted.
+ Each row contains the saturations of one sample. The omitted loadings on the irrelevant factor were not significant.
higher than that of British means (21.37 ± 11.94 for males, n = 100; 21.38 ± 11.82 for females, n == 100) as stated by Franks et al. (1960).
The differences between the present results on Egyptians and the British samples on the E scales of the EPI were not as large as those on the N scales. On the other hand most of the normal Egyptian groups had higher mean scores than British samples on Rhathymia (33.97 ± 12.01 for males, n = 100; 34.45 ± 11.44 for females, n = 100) as stated by Franks et al. (1960). The most prominent result in the three E scales, was the low mean scores among Egyptian neurotics. The majority of them were anxiety states; i.e. dysthymics. This finding verifies a hypothesis stated earlier by Eysenck (1947).
Two significant factors were extracted in all of the 16 groups from the matrices of intercorrelations between the tests. The variance accounted for ranged between 75-86% of the total variance. The unrotated factors were rotated into orthogonal simple structure according to the Varimax method of Kaiser (1958) to gain a psychological meaning.
The first Varimax factor among the 16 groups was a clear and high-loaded N-factor. Its saturations ranged between 0.846 and 0.949. The second Varimax factor was a high- loaded E-factor. Its loadings ranged between 0.705 and 0.935. Then a transformation to oblique simple structure was carried out by using the Promax method. Table 3 presents the result.
The resulting changes from Varimax to Promax solution were minimal and there were obvious similarities between them. The two factors had the same characteristics, i.e. N and E high-loaded factors. Following the Promax oblique rotation, interfactor corre-
Table 4. Interfactor correlation (fr) among the 16
groups
Sample No.
(fr)
Sample No.
(fr)
1.
-0.211
9
-0.187
2.
-0.324
10
0.159
3.
-0.182
11
-0.153
4.
-0.241
12
-0.342
5.
-0.338
13
0.215
6.
-0.321
14
-0.148
7
-0.289
15
-0.088
8.
-0.288
16
-0.322
271
ص ٤٦١