Orientation discrimination in WS 12
correct response time across the groups and conditions. This is a conservative method
of score adjustment. Mean RT by group and condition is illustrated in Figure 4. The
analysis showed a main effect of obliqueness, F(1, 42)=29.08, p<.001, partial η2 =
.41, which was caused by slower responses to the oblique trials. The effect of group
was not significant, F(1, 42)=1.48, p=.23, partial η2 = .04. However, there was a
significant group by obliqueness interaction, F(1, 42)=4.20, p=.05, partial η2 = .09.
Analysis by groups indicated a significant effect of obliqueness in both groups; this
was larger in the TD group, t(21)=4.92, p<.001, than in the WS group, t(21)=2.55,
p=.02. Response times of the 2 groups to the nonoblique trials were very similar,
t(21)=.06, p=.96, in comparison to a group difference which approached significance
in responses to the oblique trials, t(21)=-1.78, p=.09.
Figure 4 about here
Z-score analysis
Having established that the performance of the WS group on the Squares
discrimination task differed from that observed in typical development, we were
interested in how this difference compared to the level of performance measured in
WS on the Squares construction task employed by Farran et al. (2001). A direct
comparison of the absolute levels of performance of individuals with WS between
these two tasks is not appropriate as one would not be able to ascertain the extent to
which any differences in the absolute level of ability reflected impaired or unimpaired
performance. Therefore, the two original data sets were transformed into z-scores on
the basis of the performance of controls as this accounts for any difference in the
relative difficulty of the two tasks which occur in typical development. In the present
study, the correct response scores of the WS group were standardised separately for
each task, based on the distribution of performance of each respective control group.