Chapter 4: Methodology
(1) Sampling
Data for this survey was collected by two different methods, i.e. (a) mail questionaires and (b) questionaires completed by interviews. Different methods were adopted for the selection of samplings. (i) For mail questionaire
The sampling list was taken from the list of Voters' Registry (Shamshuipo West) 1983. Based on the random table, 5143 samples were selected as subjects for mail questionaire. (Sample size = 2.44% of the total population) Each selected individual was sent by mail, a package containing a questionaire, explanatory note and a self-addressed envelope. At the closing date, on March 31, 1984, a total of 1012 valid questionaires were returned. The return. rate was thus 19.7%. (ii) For Interviews
Sampling list was taken from the Household Information of the Census and Statistic Department. From the list of 42526 domestic units, one from every 169 units were selected sequentially. A total of 250 samples was obtained systematically. (0.588% of the population) Among the dwellers of the selected units who were over 18 the subject was again randomly selected. A total of 210 questionaires were successfully completed.
(2) Weighting
Since the future of Hong Kong is such a complex issue, it can be expected that those who responded, especially among those by mailing back questionaire, would have better education background compared to the rest of the Shamshuipo West population.
In fact, Table X shows that in Shamshuipo West, 57.7% of the resident have an education level of primary school or under. Yet only 23.9% and 38.1% of the respondents of mail questionaire and interview respectively have this education level. Therefore, without weighting, in comparison with the general population, the sampling would be bias in terms of their education level. Therefore, the findings has been adjusted by a weighting. Formula as follows:
=
Weighting co-efficient (W)
The Distribution of Education level among Shamshuipo West (E)
x k
k
=
The Distribution of Education level of
the Sample (%)
adjusting constant.
In this case,
k = 1.1015 for mail questionaire.
k = 1.0153 for interviewed questionaire. According to the above formula, a table was worked out as the following table:
Mail (k=1.1015) Questionaire
(k=1.0153)
Interview
Level of Education
E
S W=kxE/S
W=kXE/S
1. Illiterate
0.203 0.029 7.712 0.203 0.087
2.369
2. Primary school
0.373 0.210 1.957
0.373
0.290
1.306
3. Junior Forms
0.182
0.180 1.113
0.182
0.203
0.910
4. Senior Forms
0.191
0.336 0.627 0.191 0.300
0.646
5. Post-secondary
0.051
0.245 0.232 0.051 0.121
0.428
34
•FC040/1727
No comments yet.
Private notes are available after approval.