130
(.003 and .02); written English is significantly correlated with rank
on total Spanish (.008) as well. For the total group, rank on total
English is significantly correlated with rank on total Spanish.
Comparison of Male and Female Students
Table 4-7 shows the means, standard deviations, and the range of
the students' scores by sex. There are no differences in Spanish;
there are differences between male and female scores in English.
Statistical comparisons of means of males and females do not reveal
any significant differences. Yet analysis of correlations of total
language scores do indicate significant differences between males and
females. Analysis of ethnographic data confirms these findings and
will be discussed later.
Entire Cuban Group
Although only 41 of the 63 Cuban students participated in the
individual section of the research, additional data were collected
on the entire group. Information on these students is provided in
order to present a more complete picture of the relationship of first
language ability to second language acquisition for the Cuban popula
tion at this junior high school. Data were collected in a variety of
ways. While observing the participating Cubans, it was impossible not
to observe the others; in informal conversations, the nonparticipants
joined the participants. All students in the program were administered
the standardized group test, LAB, as previously described. An analysis
of students' achievement by rank order, as well as change over time,