香港大學民意研究計劃今日如期在《港大民意網站》(http://hkupop.hku.hk)發放十大立法會議員的最新評分結果。《民意網站》一般逢星期二下午二時於網上公佈定期調查結果，公眾假期除外，並同時預告來週的發放項目。《民意網站》會定期對該等安排作出檢討及調整。

按照計劃，《民意網站》下次發放數據的日期及時間為八月二十七日星期二下午二時，屆時將會發放市民對特首董建華的最新評分、市民對特區政府的滿意程度，及其對特區及北京中央政府信任程度的最新調查結果。

根據《民意網站》今日發佈的調查結果，市民對立法會議員的最新（八月二至七日）支持度排名，首三位是劉千石、范徐麗泰及周梁淑怡，分別得59.8、59.6及58.7分；李卓人及劉慧卿則名列第四位及第五位，分別得58.5及54.3分。至於第六位至第十位的排名，依次為田北俊、李柱銘、曾鈺成、黃宏發及司徒華。市民對排名首五位議員的平均分為58.2分，而首十位則為55.6分，與上次在五月初進行的同類調查的結果比較，兩者分別下跌及2.1及1.5分。

有關「十大立法會議員」評分調查的方法，在網頁內「調查方法」的欄目中已有詳述。是次調查被評分的立法會議員，皆於首階段在七月中（七月十至十五日）進行的提名調查中在未經提示下獲得最多被訪者提名的議員。在該項提名調查中，被訪者可說出多至十名最熟悉的議員，結果首五位最多被訪者提及的議員分別是李柱銘、司徒華、劉慧卿、曾鈺成及劉千石，餘下名單請參閱有關數表。獲得提名次數最多的十二名議員將進入第二階段評分調查。在八月初進行的第二階段評分調查中，被訪者就所有入選的十二名議員以零至一百分進行個別評分，零分代表絕對不支持，一百分代表絕對支持，五十分為一半半。統計結果後，認知度最低的二名議員再被剔除，餘下十大議員。《民意網站》已把二零零零年開始的各次提名調查結果加入網站，方便各界人士參考。

《民意網站》今日所載的最新調查，全部屬於訪員執行的隨機抽樣電話訪問，目標對象皆為十八歲或以上之操粵語的香港市民，各項調查的受訪人數都超過一千人。在九成半置信水平下，十大議員中各個評分的抽樣誤差為少於正負1.4分，而提名調查中各個百分比的抽樣誤差則少於正負三個百分比。所謂「九成半置信水平」，是指倘若以不同隨機樣本重複進行有關調查100次，則95次的結果會在正負誤差之內。各界人士如對《民意網站》所載調查的方法有任何疑問，研究組的成員會樂於解答，但不會在現階段就調查結果作出評述。此等安排將會在人手充裕後再作檢討。《民意網站》由民意研究計劃主任鍾庭耀博士負責，與香港大學立場無關。

The Public Opinion Programme (POP) at the University of Hong Kong today releases on schedule via the "HKU POP SITE" (http://hkupop.hku.hk) the latest popularity ratings of the top ten Legislative Councillors. POP's normal practice is to release the results of our regular surveys every Tuesday at 2 pm via our POP Site, except during public holidays, each time with a forecast of the items to be released in the forthcoming week. We will review and adjust this operation regularly.

According to this schedule, the date and time of our next release will be August 27, 2002, Tuesday, at 2 pm, the latest results on the rating of CE Tung Chee-hwa, people's satisfaction with the HKSAR Government as well as their trust in the HKSAR and Beijing Central Governments will be released.

As regards the latest (August 2-7) ratings of the top ten Legislative Councillors, the top three were Lau Chin-shek, Rita Fan and Selina Chow, attaining 59.8, 59.6 and 58.7 marks respectively. Lee Cheuk-yan and Emily Lau ranked fourth and fifth, with support ratings of 58.5 and 54.3 marks respectively. The sixth to the tenth ranks fell to James Tien, Martin Lee, Jasper Tsang, Andrew Wong and Szeto Wah correspondingly. The mean score obtained by the top five councillors was 58.2 marks, while that for the top ten was 55.6 marks. When compared with the results captured in early May, both averages have dropped, by 2.1 and 1.5 marks respectively.

The research design of our "Top 10 Legislative Councillors" has been explained in detail under "Survey Method" in our corresponding web page. The top councillors listed in our latest survey were all those who obtained highest unprompted mentions in our first stage naming survey conducted in mid-July (July 10-15). In that survey, respondents could name, unaided, up to 10 legislators whom they knew best. Martin Lee, Szeto Wah, Emily Lau, Jasper Tsang and Lau Chin-shek were mentioned most frequently. Please refer to the relevant table for the rest of the list. Those twelve who were named most frequently were then entered into the second stage rating survey. During the second stage rating survey, which was conducted in early August, respondents were asked to rate each legislator in turn using a 0-100 scale. 0 indicates absolutely no support, 100 indicates absolute support, and 50 means half-half. After calculation, the bottom two legislators in terms of recognition rate were dropped, leaving behind the top 10. For easy reference, the POP Site has already displayed the results of all naming surveys conducted since year 2000.

All new surveys reported in the POP Site today are random telephone surveys conducted by interviewers, targeting at Cantonese speakers in Hong Kong of age 18 or above. The sample size of all surveys is over 1,000 respondents. At 95% confidence level, the sampling error of the top 10 legislators is less than plus/minus 1.4 marks, while that for percentages in the naming survey is less than plus/minus 3 percentage points. That means 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. Shall anyone have any question regarding the research design of the surveys published in the POP Site, members of the POP Team will be happy to answer them, but we will not comment on the findings at this stage. Such an arrangement would be reviewed when more resources are available. Please note that Dr CHUNG Ting-yiu Robert, Director of Public Opinion Programme, is solely responsible for the work published in the POP Site, which does not represent the stand of the University of Hong Kong.