Advances in the Study of Bilingualism

than the Welsh participants, at first sight providing a possible explanation
for the superior self-rated proficiency of the Miami participants in English,
and for the more extensive use of English ML in the Miami data. However,
qualitative evidence based on observation and experience of data transcrip-
tion points to the fact that there may have been a tendency for the Welsh-
English speakers to underestimate their proficiency in English for political
reasons, and that the Spanish-English speakers’ self-assessment may in some
cases have expressed their desired level of proficiency rather than its actual
level. Our assessment of proficiency based on the transcripts leads to these
conclusions. If this is the case then as suggested above, proficiency may not
be the only factor determining the choice of ML.

132 Part 3: Bilingual Language Use

0%

5%

11%

85%

0%

5%

26%

68%

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

Some words Basic

conversaon

Fairly confident Confident

Sp-Eng

Wel-Eng

Figure 6.4 Language profi ciency self-report for English

26%

52%

22%

0%

72%

15% 12%

1%

0%

10%

20%

30%

40%

50%

60%

70%

80%

Spanish/Welsh English Both Other

Sp-Eng

Wel-Eng

Figure 6.5 Primary school language for the Spanish-English and Welsh-English corpora