Doubts raised about validity of survey

First published: 13th September 2013

On Sunday, I received a call from the Hong Kong Public Opinion Research Centre Ltd for a survey about people's perception of the performance of Legco and Hong Kong's political parties.

However, I was denied the opportunity to contribute my opinion because I do not speak Cantonese. The survey caller was able to tell me this in English.

A basic principle of statistics is that a sample of a population should be random, and any systematic bias, such as ability to speak a particular language, will invalidate the survey results.

In this case, the result would appear to be that the research centre's statistics ignore ethnic minorities.

The research centre is a subsidiary of the One Country Two Systems Research Institute.

I spoke to a Mr Chan from the centre who informed me that the survey I was unable to participate in was for "internal reference" by the institute.

It has enough resources to make about 20,000 phone calls for the survey, but chooses not to allocate resources for support of other languages, such as English, Putonghua, other Chinese dialects and foreign languages.

Mr Chan emphasised that this survey was not commissioned or funded by the government, but he could not say what the "internal reference" included, such as supporting material for the many consultancy studies performed by the institute for the government's Central Policy Unit.

Can the institute stand by the credo and working principles on its website open its research methodology to public scrutiny and explain how this apparent marginalisation of ethnic minorities affects the consultation studies it has prepared for the government?

More Information