8.
indicates all too clearly
their comprehension.
that this question is simply beyond
Fourthly, as to
the order in which these options
appe ar in the Section F3,
experts have advised me that it is
unfair to present these options always in the same order SO
that every respondent would have heard the option: "no change"
before the others. I am told that it is normal survey practice
to rotate the order in which the options are presented to avoid
what is called "order bias". Otherwise the respondents would
inclined to accept the first option instead of the
be mo re
others.
This is particularly so when the entire interview took
a long time as in this case. As a result, the prejudice to
direct elections for 1988 is very great because the only option
which allowed for a "Yes" answer to the introduction of direct
elections in 1988 came last in the question.
Fifthly, experts have also told me of another flaw
interviewers were instructed to deal with
in the way the
another part of the Questionnaire, namely, Section F5. The
experts have called it "the wrong skip". According to Section
F. only those respondents who had chosen option (4) in Section
F3 would be asked further questions in Section F5 which
contained 6 options as to what changes should be made to the
Legislative Council in 1988, including the introduction of
was wrong, because even
direct elections.
This
those