128
be significantly correlated for 13-year-olds (.59), there is only a
tendency toward significance for the 12-year-olds (.46), and none for
the older students (.41 and -.7). The age variable presents a con
fusing pattern in this table, with the 18 13-year-olds showing a
significant correlation and 14 12-year-olds showing none. Sex is the
important variable here. Nine of the 14 12-year-olds are female; 11
of the 18 13-year-olds are male. For females, there is no significant
correlation (.29), but for males there is (.58). When all the students'
scores are combined, there is a very significant correlation between
the written scores in English and Spanish (.43) for the group.
Correlations of oral and written measures
Correlations of oral and written language measures with rank on
total scores are displayed on Table 4-6. Analysis of these data reveals
that for females both oral English and oral Spanish are significantly
correlated with total English (.008) and total Spanish (.04). Written
English and written Spanish are also significantly correlated with
total English (.0001) and Spanish (.0001). There is nothing surprising
about these correlations: written and oral measures of English are
components which are significantly correlated with their total. Written
and oral measures in Spanish are significantly correlated with their
total. However, for males, not only does this pattern of component
correlations hold true, but there are additional significant correlations
as well. All components, except oral English, are correlated with their
totals and with the totals of the alternate language as well. Oral and
written Spanish are significantly correlated with rank on total English