3.2.1 Split-half Reliability
Measures of reliability derived from split-half reliability testing
indicated that the STP approach produced reliable measures of amplitude
and latency across all components (Tables 5 – 7). In Block one, ther values for the mean baseline-to-peak amplitude of even trials
compared to odd numbered trials were considerably higher for the STP
approach than the STW approach for all components except P3, which was
close in value (r = .74 for STP and r = .79 for STW; Table
5). However, the differences
between the coefficients are much less in block two, particularly for
the later components.
Correlations for STP peak-to-peak measures are shown in Table 6, with
moderately strong to very strong correlations for all components.
Components P2 and P3 increased in strength from Block One to Block Two.
Table 7 shows that coefficients for latency measures are very high for
all components in both blocks with all coefficients exceeding .93 except
for the P1 component where r = .79 and .89, respectively for
block one and two.