Two other tests were undertaken to ascertain that it was judicious to proceed with the exploratory
factor analysis. These tests were the %DUWOHWW¶V WHVW RI VSKHULFLW\ DQG WKH .DLVHU-Meyer-Olkin test.
%DUWOHWW¶VWHVWRIVSKHULFLW\(1950) tests the null hypothesis that the correlation matrix is an identity
matrix (i.e. that there is no relationship among the items). The null hypothesis states that there are all
¶VRQWKHGLDJRQDORIWKHPDWUL[DQG¶VRQWKHRII-GLDJRQDO/DUJHUYDOXHVRI%DUWOHWW¶VWHVWLQGLFDWHD
greater likelihood that the correlation matrix is not an identity matrix and that the null hypothesis can
be rejected. The Kaiser-Meyer-Olkin (KMO) test (Blackmore & Blackwell, 2006) is a measure of
sampling adequacy that compares the magnitudes of the calculated correlation coefficients to the
magnitudes of the partial correlation coefficients. If the items share common factors, then it would be
reasonable to expect that the partial correlation coefficients between the pairs of items would be small
when the linear effects of other items have been removed. The KMO measure can range between 0
and 1, with smaller values indicating that the correlation coefficient is small relative to the partial
correlation coefficient, and therefore a factor analysis may be unwise. When evaluating the size of the
overall KMO, Kaiser (1974), suggested to use the following criteria for these values: above .90 is
³PDUYHOORXV ́, iQWKHVLV³PHULWRULRXV ́, iQWKHVLVMXVW³PLGGOLQJ ́, lHVVWKDQLV³PHGLRFUH ́,
³PLVHUDEOH ́RU³XQDFFHSWDEOH ́(p.35).
This study showed the following statistics: KMO=.854 DQG%DUWOHWWVWDWLVWLFĮ .000. Hence, it was
concluded that the correlation matrix was factorable.
6.3.2 NUMBER OF FACTORS
Based on the Kaiser criteria (eigenvalues > 1), the data showed four distinct factors, explaining 62.4
% of the total variance. The Scree plot likewise hinted at a four-factor structure. Table 6.3 shows the
factor pattern matrix.