Research Design返回

This was a random telephone survey conducted by telephone interviewers under close supervision. To minimize sampling bias, telephone numbers were first drawn randomly from the residential telephone directories as "seed numbers", from which another set of numbers was generated using the "plus/minus one/two" method, in order to capture the unlisted numbers. Duplicated numbers were then filtered, and the remaining numbers were mixed in random order to produce the final telephone sample.

All data were collected by our interviewers using a Computer Assisted Telephone Interview (CATI) system which allowed real-time data capture and consolidation. To ensure data quality, on top of on-site supervision, voice recording, screen capturing and camera surveillance were used to monitor the interviewers' performance.

Target respondents of the study were Cantonese-speaking population of Hong Kong who fulfilled either one of the following criteria: 1) parents aged 18-year or above with at least one child aged 17-year or below; 2) young adults aged between 18 and 34-year with no children; and 3) children aged between 9 and 17-year. The target sample size for each group was set at 500. In order to capture as many pairs of parent-and-child of the same family as possible, when telephone contact was successfully established with a target household, one person fulfilling either criterion 1 (parent) or 2 (young adult) was invited to participate in the survey first. If more than one qualified subject had been available, selection was made using the "next birthday rule" which selected the person who had his/her birthday next. All parent respondents were asked at the end of the interview if they would give consent to their child of age 9-17-year to participate in this survey. If more than one child had been available, selection was made using the "next birthday rule" again. If the child was not available at that time, appointment and subsequent recalls would be made accordingly. In other words, provided that parental consent was given and that the child was qualified and completed the interview, both the qualified parent and child from the same household would be interviewed. However, due to various reasons, and mainly because some parents hesitated to let their children take part in this survey, we failed to interview both targets for every qualified household. Therefore, we had to supplement our target group of children between 9-17 years old with another separate random selection process. Parental consent was still a requirement before we started the interview and if more than one subject had been available in the household, "next birthday rule" would be applied.

The fieldwork was conducted during the period of May 7 to 15 and June 2 to 13, 20081 . A total of 1,516 Hong Kong citizens aged 9-year or above were successfully interviewed. Among them, 506 were parents, 506 were young adults and 504 were children, of whom only 143 belonged to parent-and-child pairs.

As shown from the calculations in Appendix I, the effective response rate of each group, namely parents, young adults and children, was 77.8% (Table 1), 77.9% (Table 3) and 79.3% (Table 5) respectively, and the sampling error for percentages based on each target group is less than 2.2%. In other words, the sampling error for all percentages is less than plus/minus 4.5% at 95% confidence level. Please refer to Appendix I for detailed breakdown of the contact information.

1Due to the disastrous Sichuan earthquake of May 12, 2008 which killed hundreds of thousands of people and thus affected the mood of all Chinese people in the world (including Hong Kong people), the fieldwork operation was suspended for two weeks.