HKUPOP today releases the interim report of a study related to the 2016 Legislative Council Election Rolling SurveyBack

Press Release on August 22, 2016

The Public Opinion Programme (POP) of The University of Hong Kong in its 2016 Legislative Council Election Rolling Survey follows the current practice of most electronic media to read out only the first candidate’s name in each list but not the first two names have caused some debates. In view of this, POP has spent extra resources to conduct a scientific experiment for five consecutive days from August 11 to 15. Apart from interviewing 1,027 registered voters using the standard method, our interviewers also randomly selected 469 respondents to conduct a test by “reading out two names”, some of them read out “one name then two names” while some read “two names then one name”. POP today releases the interim report of this study to the public, and also sends a copy of it to those who have directly expressed their opinions to POP. POP will consider all feedbacks first, before deciding whether or not to adjust the survey methodology. The public can now click on to read the report. The concluding remarks of the interim report are as follows:

5.1 The results of the test conducted from August 11 to 15, 2016 show that, among the 84 candidate lists in all 5 geographic constituencies, the p-values of the t-tests between “read one name” and “read two names” lie between 0.072 and 0.996, meaning that at 95% confidence level there is no significant difference between the two sets of figures. In other words, the limited samples do not provide sufficient statistical evidence to show that “read one name” and “read two names” methods would significantly affect the survey results. Of course, if the sample size of the test could be many times bigger, or if the test could be run much longer, the results could be different.

5.2 Since there is no statistical difference between the two sets of results, we can choose to continue to “read one name” or change to “read two names” method without worrying about adjustments to the data previously collected. If we will change to “read two names”, POP will have to reallocate some resources and downsize other election studies. We will consider all feedbacks to the interim report and decide as soon as we can.

5.3 POP has approached the problem of reading names in the most scientific, open and honest way. We will balance the validity of scientific evidence and public expectation when making decisions. We understand that opinion surveys are assets of civic society, so if the public does not agree to its design, its result would be useless, and the operation of opinion surveys would also be affected. In any case, POP remains absolutely neutral to all candidates and groups, and treats them equally. POP has already signed agreements with all sponsors prior to the survey to safeguard POP’s research autonomy.

POP welcomes suggestions from all parties on whether and how to adjust the methodology for the remaining survey, after reading the result of our experiment results. We also welcome anybody to conduct similar experiments to scientifically verify the results.

Media enquiry:     Frank Lee     Tel: 3917 7724     Email: [email protected]