- 3
The questionnaire
6.
The design of the questionnaire in any survey
is as important as the proper sampling technique to be
employed. Presumably all the questiona were put to the
interviewees in Chinese, and the Chinese version is cluasT,
to say the least.
7-
Question 1
Presumably the purpose of this
The
question is merely to find out whether the respondent has
heard of the six organisations, with no attempt to establish
his mowledge of these organisations. One may have heard of
a lot of things without knowing what they really are.
question is a superficial one. According to Table 1, it is
surprising to note that the Reform Club is more well-known
than the Civic Association.
8.
Question 2 The Chinese translation is a bit
misleading, particularly when the question is read out quickly. An elected member() can be easily
confused with an appointed member', who is selected from
members of the public for appointment. In English, the con-
cepts of election' and 'selection' are lucidly clear, but
the Chinese translation in the questionnaire fails to bring
out this very important difference.
9.
Question (a) The question is not properly
phrased, and the Chinese translation does not bring out the
meaning of the English version. This is why the five answers
given do not tally with the question.
CONFIDENTIAL
4
10.
Question 4(c) - In the same way as question 2,
the different concepts of 'election' and 'selection' are not
brought out clearly.
11.
Cuestion 5- The Chinese version is misleading.
H
*', in this particular context, can mean what one would
W
expect from an elected member. The question should have
been couched in this way in Chinese, "If there are alectad
members in the Legislative Council and you are qualified
to vota to decide who should be elected, will you vote?"
Question 6 This question is too long. The Chinese
translation of the sentence beginning with 'In other words
is confusing.
12.
13.
Other ccaments There is no real indication of the
quality of enumeration and the approach adopted by the
interviewers. For example, how much trouble did they take to
explain the meanings of the terms/names to the respondents?
Without some indication of this it is difficult to evaluate
the validity of the positive answers or the extent of ignorance
which is indicated by the 'don't know' category.
The "Introduction" of the resort
14.
Paragrach 5 Reference is made to the "normal degree
of statistical error present in all sample survey data".
All that is normal in this kind of situation is that there is
error (sampling and non-sampling). Non-sampling error cannot,
under present technique, be calculated. Sampling error can.
CONFIDENTIAL
- 5 -
*
This caission is surprising and one would have expected to
fox
see it done, especially in relation to main findings: example, in paragraph 13 of "Summary of Findings", in relation to the number of persons who were in favour of having elected members in the Legislative Council.
"Summary of Findings"
15.
Paragraph 11
The statement that 'Quite a significant
proportion of adults thought that Legislative Council and
Executive Council have elected members' differs from the percen-
only 16% thought so, and 48% did not. tagea given in Table 1- know. This "don't know" figure as here presented is a case of inadequate presentation. Whereas the positive replies are analysed under the three different Councils, the "don't know's
The figures could only be meaningful ta are not so broken down. together if one assumed that respondents' knowledge (or lack of related to all 3 Councils together or none of them.
imagines that this was not the case.
16.
One
Paragraph 13 - The last clause but hardly anyone was
against the idea' is not factual reporting.
According to
Table 2, 48% of the respondents were indifferent.
17.
Paragraph 15 - The 'don't know' figure (10%) is
a case of inadequate presentation. If they do not know the reasons for having elected members, what makes them say sot.
Paragraph 16 A differentiation was made between "ignorant about the cubject" (11%) and "don't know" (156); these were both included in the almost 50% of total adulta.
18.
CONFIDENTIAL