expensive. However, if the questionnaire is complicated, such a method should not be used as there would be difficulties in quality control.
- Face-to-face interview.
(3) Collection and analysis of data -- This is a purely technical procedure.
(4) Interpretation of the results -- With statistical methods, the inclinations of opinions on a certain question shown by people of different occupations, ages, sexes and educational backgrounds are analysed and then an interpretation of the survey results is made. This procedure, on the one hand, involves technical problems and on the other, depends on the analysts' understanding of the political and social implications of that question.
With respect to the above-mentioned procedures: (1) the objective of the survey must be established by the commissioning organization; (2) as to the scope of the survey and the design of the questionnaire, there are two possible practices: (i) the commissioning organization decides first and then advice will be given by the research company; or (ii) the research company is solely responsible, e.g. the public opinion poll conducted in 1987 by the Hong Kong Government on the reform of the political structure; (3) the technical aspects are usually decided by the research company in consultation with the commissioning organization; (4) the actual interview and compilation of statistics are carried out by the research company; and (5) as to the interpretation of the results, there are also two possible practices: (i) the results may be analysed by the research company alone; or (ii) they may be analysed jointly by the research company and the commissioning organization.
3.4 Monitoring and quality control
Professional research companies usually have set procedures to control the quality of surveys. However, in view of the controversy caused by the public opinion poll conducted in 1987 by the Hong Kong Government on the reform of the political structure, when proposing that a similar sample survey should be conducted on the Basic Law (Draft), some people suggested that an independent and authoritative monitoring group comprising eminent persons from various sectors should be established to monitor the survey. Only with this practice will the results of the survey be accepted. However, there were no specific proposals on the composition, powers and functions of this monitoring group.
34
