The Public Opinion Programme (POP) was established in June 1991 to collect and study public opinion on topics which could be of interest to academics, journalists, policy-makers, and the general public. POP was at first under the Social Sciences Research Centre, a unit under the Faculty of Social Sciences of The University of Hong Kong, and was transferred to the Journalism and Media Studies Centre in The University of Hong Kong in May 2000. In January 2002, and was transferred back to the Faculty of Social Sciences in The University of Hong Kong. POP provides quality survey services to a wide range of public and private organizations, provided that they allow the POP Team to design and conduct the research independently, and to bear the final responsibilities.


In February 2003, POP was commissioned, for the third time, by Media Education Info-tech Co. Ltd. (which owns "Education") to conduct a survey on the public's perceptions of the eight universities in Hong Kong. The primary objective of this survey was basically the same as that of the past two years, i.e., to study the general public's perception of the local universities, but with a slightly different instrument designed in line with the client's new emphasis. In fact, a similar study was carried out three years ago, which was designed and coordinated entirely by the client using a different methodology. Any direct comparison between the results obtained from the first and subsequent surveys is not recommended, while comparison of results from the second, third and fourth surveys must also be made with great caution.


The questionnaires used in this and the previous two years' surveys were designed independently by the POP Team after consulting the client. Fieldwork, data analysis and interpretation were also carried out independently by the POP Team.


This year's telephone survey was conducted during the period of 28 February - 11 March 2003. A total of 1,025 Hong Kong Cantonese-speaking residents aged 18 or above were successfully interviewed. The overall response rate was 65.0% and the standard error due to sampling was no more than 1.6 percentage points. That means at 95% confidence level, the sampling error of percentage figures was less than plus/minus 3.2 percentage points.