Using online social networks to recruit representative samples of internet population for evaluating health promotion multimedia

To ensure the effective communication of health promotion multimedia campaigns, we usually conduct copytests for the developed media. However, conducting traditional copytests takes time and costly. Nowadays the use of the internet is dramatically increasing. The internet is potential to be an effective survey tool. However, the challenge is how to recruit unbiased population samples over the internet.

A website for questionnaire survey was developed. We used online social networks, such as micro blogs, bulletin board systems (BBS) and popular websites in Taiwan, to recruit as many participants to take the questionnaire survey as possible. The recruited participants could reach the questionnaire survey website by following the link enclosed in invitation emails or posted messages. During the survey period, the weblog of the website was activated to trace where the participants came from.

A total of participants who completed the questionnaires was 2664. Based on the distribution of Taiwan's internet population the minimized sample size with 95% confidence level, the rate of the internet population (72.56%) and with 3% margin error, should be 1068. According to the age, gender and region distribution of Taiwan's internet population (refer to the 2010 report of the Taiwan Network Information Center), we used proportionate stratified random sampling method to selected 1068 from 2664 completed questionnaires.

We further analyzed the weblog of the questionnaire survey website. The result revealed that 36.8%, 39.8%, 18.8%, and 4.6% of the participants came from micro blogs, BBS, Health99 and other websites, respectively. Interestingly, there were significantly differences in using online social networks based on their age, region and gender. Such findings can be used to choose appropriate online channels to recruit participants for a research study on the internet.