HKU POP releases popularity figures of CE and HKSAR Government and a District Council pre-election surveyBack

The figure of "decided voters" in the original release was incorrectly based. It should be corrected as 37%. We apologize for the mistake.

A brand new "DC Guessing Game" hosted by the Public Opinion Programme (POP) at the University of Hong Kong is now available at "PopCon" (http://popcon.hk). Every day, people can make guesses on the final result of the DC election down to constituency level, in order to earn credits and win prizes including iPod Touch and iPod Nano. Up till 12 noon today, over 38,000 guesses have been received, and the time series charts are drawn for all constituencies. They will be updated real time until the last minute of the election. Moreover, the results of five key constituencies are also shown real time in the homepage of "PopCon", they include "Kwun Lung" in Central and Western District, "Whampoa East" in Kowloon City District, "Lung Sheung" in Wong Tai Sin District, "Lok Tsui" in Tuen Mun District and "Lai Wah" in Kwai Tsing District. The accuracy of such public wisdoms would be analyzed and compared to results of opinion surveys by POP after the election. The "DC Guessing Game" is jointly developed by POP and Internet creative house Swarm.

The latest surveys conducted by POP show that the popularity of CE Donald Tsang has gone back to the level before he delivered his Policy Address. His support rating now stands at 47.9 marks, approval rate at 25%, disapproval rate at 65%, giving a net popularity of negative 40 percentage points. As for the SAR Government, there is not much change compared to one month ago, and its net popularity still stands at negative 24 percentage points. Regarding this Sunday's District Council election, 66% of the registered voters intend to vote, a big drop of 10 percentage points compared to that of 2007. Moreover, in terms of candidate choice, those who have already decided how to vote has also dropped by 10 percentage points. This means that the turnout rate in the coming election may be lower, and the effect of loyal voters more significant. As for the main factors affecting candidate choice, 47% said political platform and alignment are most important, 40% said past performance. The situation is a reversal of 2007, probably because this year's winners will have the chance to get a seat in the Legislative Council Functional Constituency (to become so-called "Super District Councillors") next year. The maximum sampling error of all percentage figures is between +/-2 and +/-4 percentage points at 95% confidence level, while the sampling error of rating figures needs another calculation. The response rate of the surveys is between 70% and 72%.

As a general practice, all figures released by POP today have been weighted according to provisional figures obtained from the Census and Statistics Department regarding the gender-age distribution of the Hong Kong population in mid-year 2011.

1. Popularity figures of CE Donald Tsang and the HKSAR Government

Date of survey	Overall sample size	Response rate	Maximum sampling error of percentages[6]
13-20/10/2011	1,013	71.8%	+/-3%

[6] Errors are calculated at 95% confidence level using full sample size. "95% confidence level" means that if we were to repeat a certain survey 100 times, using the same questions each time but with different random samples, we would expect 95 times getting a figure within the error margins specified. Questions using only sub-samples would have bigger sampling error. Sampling errors of ratings are calculated according to the distribution of the scores collected.

Recent popularity figures of CE Donald Tsang and people's satisfaction of the overall performance of the HKSAR Government are summarized as follows:

Date of survey	29/8-6/9/11	13-20/9/11	26/9-4/10/11	12/10/11	13-20/10/11	Latest Change
Sample base	1,012	1,001	1,025	1,032	1,013	--
Overall response rate	64.6%	65.5%	66.1%	65.6%	71.8%	--
Finding[7]	Finding	Finding	Finding	Finding[10]	Finding and Error[7]	--
Rating of CE Donald Tsang	49.0	47.3[9]	48.4	50.6[9]	47.9+/-1.4	-2.7[9]
Vote of confidence in CE Donald Tsang	24%	24%	24%	25%	25+/-3%	--
Vote of no confidence in CE Donald Tsang	65%	67%	69%	66%	65+/-3%	-1%
Net approval rate	-41%	-43%	-45%	-41%	-40%	+1%
Satisfaction rate of SARG performance[8]	--	22%	--	--	24+/-3%	+2%
Dissatisfaction rate of SARG performance[8]	--	46%	--	--	48+/-4%	+2%
Net satisfaction rate	--	-24%	--	--	-24%	--
Mean value[8]	--	2.6+/-0.1 (Base=603)	--	--	2.6+/-0.1 (Base=632)	--

[7] All error figures in the table are calculated at 95% confidence level. "95% confidence level" means that if we were to repeat a certain survey 100 times, using the same questions each time but with different random samples, we would expect 95 times getting a figure within the error margins specified. Media can state "sampling error of rating not more than +/-1.4, sampling error of percentages not more than +/-4% at 95% confidence level" when quoting the above figures. The error margin of previous survey can be found at the POP Site.
[8] Collapsed from a 5-point scale. The mean value is calculated by quantifying all individual responses into 1, 2, 3, 4, 5 marks according to their degree of positive level, where 1 is the lowest and 5 the highest, and then calculate the sample mean. Starting from March 2011, this question only uses sub-samples of the tracking surveys concerned. The sample size for this series is 639.
[9] Such changes have gone beyond the sampling errors at the 95% confidence level, meaning that they are statistically significant prima facie. However, whether numerical differences are statistically significant or not is not the same as whether they are practically useful or meaningful.
[10] These questions only use sub-samples of the survey concerned. The sub-sample sizes of questions on CE's support rating and hypothetical voting were 770 and 843 respectively.

The latest survey showed that, CE Donald Tsang scored 47.9 marks, and 25% supported him as the Chief Executive while 65% opposed him. Regarding people's appraisal of the overall performance of the HKSAR Government, the latest figures revealed that 24% were satisfied, whereas 48% were dissatisfied. The mean score is 2.6, meaning in between "half-half" and "quite dissatisfied".

2. District Council pre-election survey

Date of survey	Overall sample size	Response rate	Maximum sampling error of percentages[11]
26/10-/1/11/11	605	69.9%	+/-4%

[11] Errors are calculated at 95% confidence level using full sample size. "95% confidence level" means that if we were to repeat a certain survey 100 times, using the same questions each time but with different random samples, we would expect 95 times getting a figure within the error margins specified. Questions using only sub-samples would have bigger sample error. Sampling errors of ratings are calculated according to the distribution of the scores collected.

Whenever there are large-scale elections, POP would conduct regular surveys to track opinion changes. Major findings regarding registered voters are summarized below, together with parallel figures obtained in previous District Council elections. Findings beyond registered voters can be found in the POP Site.

	2003				2007				2011	cf similar survey 2007	cf similar survey 2003
Date of survey	15-17/9	19-22/10	3-6/11	15-19/11	17-21/9	22-25/10	5-9/11	12-15/11	26/10-1/11/11
Sample base	714	725	721	1,157	761	801	859	877	500	--	--
Overall response rate	70.1%	60.0%	60.7%	65.1%	65.5%	67.4%	66.3%	79.0%	69.9%	--	--
Maximum sampling error of percentages (at 95% conf. level)[12]	+/-4%	+/-4%	+/-4%	+/-3%	+/-4%	+/-4%	+/-3%	+/-3%	+/-4%	--	--
Finding for each question/ Sampling error[12]	Finding	Finding	Finding	Finding	Finding	Finding	Finding	Finding	Finding and Error	--	--
Proportion of voters who planned to vote in DC Election	67%	70%	75%	74%	75%	73%	76%	76%	66+/-4%	-10%[13]	-8%
Have decided which candidate to vote for	--	--	33%	37%	--	--	38%	47%	37+/-4%	-10%[13]	--
Knew which political camps are running for the election at local district	--	--	67%	71%	--	--	72%	76%	71+/-4%	-5%[13]	--
Main factor: platform and political alignment	44%	36%	36%	35%	42%	45%	41%	36%	47+/-4%	+11%[13]	+12%
Main factor: past performance	40%	43%	46%	47%	43%	43%	49%	50%	40+/-4%	-10%[13]	-7%
Main factor: personal background and performance during election	8%	7%	3%	3%	13%	6%	3%	3%	4+/-2%	+1%	+1%

[12] "95% confidence level" means that if we were to repeat a certain survey 100 times, using the same questions each time but with different random samples, we would expect 95 times getting a figure within the error margins specified. Media can state "sampling error of percentages not more than +/-4% at 95% confidence level" when quoting the above figures.
[13] Such changes have gone beyond the sampling errors at the 95% confidence level, meaning that they are statistically significant prima facie. However, whether numerical differences are statistically significant or not is not the same as whether they are practically useful or meaningful.

Among the sub-sample of registered voters, 66% said they intended to vote. (However, please note that 66% intention to vote may not convert into 66% turnout). 37% of the registered voters have already decided which candidate to vote for. Meanwhile, 71% of the registered voters said they knew which political camps are running for the election at local district. Regarding the main factor for choosing the candidate, 47% registered voters opted for a candidate's platform and political alignment to be most important. 40% considered a candidate's past performance, while 4% considered a candidate's personal background and performance during election to be most important.

Whenever there are large-scale elections, HKUPOP will conduct various surveys to track opinion changes. The research team will also conduct exit polls on election day to study voter behaviour and motivation. Since the handover, the research team has conducted different surveys for District Council elections (1999, 2003 and 2007), including pre-election surveys, exit polls and so on.

As a general practice, our last round of District Council pre-election survey is normally conducted about one week before the election. This year is no exception. However, in light of the quiet atmosphere this year, only one survey has been conducted.

As usual, the questions asked in our "District Council pre-election" include: "Are you a registered voter?", "Do you plan to vote in the District Council election this November?", "Which political camp do you incline to identify yourself with?", "What is your main consideration in selecting a candidate?", "Have you decided to vote for which candidate?", "Do you know which political camps are running for the election at your district?" and "If XXX, YYY, ZZZ, etc, other political parties and some independent candidates are sending their members to run for candidacy at your living district, which political party or independent candidate do you incline to support?" For this year, because of small sample size, the results of the last two questions are reserved for further analysis only.

• November 8, 2011 (Tuesday) 1pm to 2pm: Ratings of top 5 Executive Councillors