THE PERCENTAGES OF OF AQUINO AND ESTRADA BARELY CHANGED. HOWEVER BINAY LEADS ROXAS, 40.7% VS. 38.3%. [SEE http://www.sws.org.ph.] PRELIMINARY RESULTS SHOWED THEM EARLIER TIED AT 40%.
WITH STILL 10% OF CLUSTERED PRECINCTS AWAITED AT COMELEC, THE ERROR MARGINS OF THE EXIT POLL RESULTS, WHEN TAKEN INTO ACCOUNT, INDICATE A CLOSE CONTEST DOWN TO THE WIRE. HOWEVER A BINAY WIN MAY HAVE A BIGGER SPREAD THAN A ROXAS WIN.
Social Weather Stations, in cooperation with the School of Government
of Ateneo de Manila University, conducted a national exit poll sponsored
by the PLDT-SMART Foundation, Inc., for exclusive broadcast by the
Associated Broadcasting Company (TV5), called the “TV5-SWS EXIT
POLL.” The partial [underscoring mine] results for the national positions
of President and Vice-President, from the partial sample as of May 11,
1:00 a.m. and aired on TV5 as of 7a.m. today, are shown below. It will
be a long week/month for Jojo Binay and Mar Roxas. The exit poll puts both
at 40 percent.
The following information is available at http://www.sws.org.ph
11 May 2010
The TV5-SWS 2010 Exit Poll
I. THE VOTES
TV5-SWS 2010 EXIT POLL
PHILIPPINES (Partial sample as of May 11, 1:00 a.m., aired on
TV5 as of 7a.m.)
II. THE VOTERS
TV5-SWS 2010 EXIT POLL
Social Weather Stations, in cooperation with the School of Government of Ateneo de Manila University, conducted a national exit poll sponsored by the PLDT-SMART Foundation, Inc., for exclusive broadcast by the Associated Broadcasting Company (TV5), called the “TV5-SWS EXIT POLL.”
The TV5-SWS Exit Poll adheres to the requirements set by the Fair Election Act’s Section 5.5 on exit polls. This requires interviewers to (a) stay at least 50 meters away from polling centers, (b) wear distinctive TV5-SWS Exit Poll clothing and carry TV5-SWS Exit Poll identification cards; and (c) inform respondents that they may refuse to answer. The Act says that the results “may be announced after the closing of the polls on election day, and must clearly announce the total number of respondents, and the places where they were taken” and that the “announcement shall state that the same is unofficial and does not represent a trend.”
SWS previously conducted exit polls in the national elections of 1995, 1998, 2001 and 2004. Those previous polls were all conducted on election-day, in the homes of a random sample of respondents who had already voted earlier in the day. Conducting an exit poll in homes rather than at polling locations is valid under the Guidelines for Exit Polls and Election Forecasts of the World Association for Public Opinion Research (WAPOR), written by the WAPOR Exit Poll Committee of which I was a member, and approved by the WAPOR Council in 2006.
This year, however, the TV5-SWS Exit Poll is conducted not in homes but at the polling centers, 50 meters away. This enables the interviews to be done, as in the developed countries, throughout the day. The interview results are transmitted to the SWS office in Quezon City, and processed to produce the exit poll results for broadcast on TV5 as soon as completed, and lawfully allowed.
SWS conducted the TV5-SWS Exit Poll with its own regular staff and a special complement of direct hires for added interviewing and data processing capacity. Starting in mid-2004, SWS has accomplished all field and data processing operations internally, without outsourcing.
The questionnaire for the TV5-SWS Exit Poll was designed in cooperation with the Ateneo School of Government (ASoG). In addition to asking about the votes for President, Vice-President and Senator, it has several question items, in addition to standard demographics (gender, age, education, socioeconomic class, religion, charismatic group, and home language) to enable substantive analysis of the 2010 vote. The TV5-SWS Exit Poll analysis will be done by SWS in cooperation with ASoG.
SWS is entirely responsible for the sampling of the TV5-SWS Exit Poll. The poll was conducted in 802 Voting Centers (VCs) nationwide, covering all provinces from Batanes to Tawi-Tawi, and chartered cities of the Philippines. The regional distribution of the TV5-SWS Exit Poll sample of VCs, after proportional allocation according to the official 2010 number of Registered Voters (RVs) in the said provinces and cities, is as follows:
NCR – 96
Within each province/city, the sample-VCs are chosen with probability proportional to their number of RVs. In each VC, SWS fielded normally two interviewers (sometimes one, if the VC is small); the total field group, including support staff, is over 1,800 persons. SWS has 110 encoding stations at its main office to accommodate the incoming interview reports.
The interviewers of the TV5-SWS Exit Poll have been trained to behave in a strictly non-partisan manner at all times. They are to approach voters to be interviewed at a steady pace throughout voting hours. They may not accept voters who volunteer to be interviewed. For the sake of privacy, interviewees’ names are not recorded; interviewees are given a thank-you card with the name of their interviewer. The TV5-SWS Exit Poll aims for an average of 40 interviews per VC, or a national sample of about 32,000 interviews. The national sampling error, though still to be ascertained, is expected to be below 1%.
The results of each VC are weighted by its present number of RVs, multiplied by the actual voting turnout percentage in the last presidential election year (2004), to progressively obtain provincial/city results, regional results, and national results. However, the voting results for national positions may be reported at the national level only, at which the sample is maximized; it is not necessary for an exit poll to reach conclusions about sub-jurisdictions.