Establishing Probability-based and Non-probability-based Panels: The Hong Kong Experience
The newly established HKPORI recently launched a series of multi-client independent panels in Hong Kong, both probability- and non-probability-based. The former include the “Hong Kong People Representative Panel” and the “Hong Kong Youths Representative Panel”, while the latter include the “Hong Kong People Volunteer Panel” and the “Hong Kong Youths Volunteer Panel”. More panels may be established in due course, possibly in collaboration with other organizations.
The probability-based online and telephone panels are representative of the Hong Kong population, they will be cost effective and adhere to high quality recruitment standards. The methodology of these panels draws from the leading probability-based and multi-client household panels in the United States and United Kingdom. They are the AmeriSpeak panel developed by NORC at the University of Chicago and the probability-based panel at NatCen in the UK. Both entities are leading social science research organizations dedicated to research excellence at affordable prices.
The HKPORI Probability-based Panels look to provide as pre-screened and representative samples of the Hong Kong population that allow people interested in quick, accurate responses to have access to quality data collection. Further, the structure of these panels would allow for substantial longitudinal research and opinion research on particular populations.
Besides probability-based panels, HKPORI has also established non-probability-based volunteer panels open to the entire Hong Kong population, for those who consider themselves qualified. These non-probability-based panels will grow fast and provide efficient and quick means to obtain opinions and thoughts from certain targeted populations, including the general public, youths, netizens, elderlies, ethnic minorities, and so on.
This paper describes the development of such panels in Hong Kong, their applications as well as the outcomes of resource-saving in using such tools.