The exit poll was sponsored by network giant ABS-CBN Broadcasting Corp. in partnership with The Philippine STAR. It "aimed at broadcasting a scientific preview of the probable election outcome by the following day," SWS head Mahar Mangahas said.
He cited the SWS’ track record in previous elections.
Mangahas has a doctorate and is a first cousin of presidential candidate Fernando Poe Jr.
In the exit poll during the 2001 senatorial elections, the SWS correctly predicted the winners.
In 1998, the SWS estimated that Joseph Estrada would win 39.2 percent of the vote and Mrs. Arroyo, who ran for vice president, would get 50 percent.
The official count of the Commission on Elections (Comelec) showed Estrada getting 39.9 percent of the votes for president. Mrs. Arroyo got 49.6 percent.
The SWS also correctly projected the first 11 winning senatorial candidates.
"While exit polls are important for quickly assessing the election outcome, they are absolutely essential for understanding the nature of the vote," Mangahas said.
"If there were no exit polls, the only voting-characteristic that could be learned, from official counts, is merely the location of the votes."
For instance, the exit poll of the 1998 presidential election found that:
• Joseph Estrada won on the basis of the Class D and E votes, since the top vote-getter among the ABC classes was Raul Roco.
• The only significant religious votes were those of Iglesia ni Cristo members, who went 81 percent for Estrada, and Muslims, who gave him 63 percent.
• The Christian religious groups who truly supported Estrada were Jesus Is Lord who gave him 52 percent, whereas El Shaddai only gave him 39 percent, the same as the national average.
• Twenty-one percent of the voters made their presidential choice only on election day itself.
• Forty-nine percent of the voters made use of a sample ballot.
SWS said it observed requirements set by the Fair Election Act of 2001 in conducting the poll — such as conducting interviews at least 50 meters away from polling places, that interviewers should wear distinctive clothing, and that the poll should cite the number of respondents and the places where the poll was taken, and it should also state that the results are unofficial and do not represent a trend.
Mangahas also stressed that "the conduct and publication or broadcast of election surveys is covered by the constitutionally guaranteed freedoms of speech, of the press, and of expression, based on landmark rulings of the Supreme Court."
A total of 4,627 voters nationwide were interviewed from 9 a.m. to 6 p.m. last Monday, May 10 for the latest survey.
"Although traditionally called an exit poll, the SWS survey was actually conducted in the privacy of respondents’ homes, rather than literally outside the polling centers as the voters come out," Mangahas explained.
Respondents were interviewed by SWS’s long-time fieldwork contractor TNS-Trends (formerly NFO-Trends). SWS fielded inspectors to ensure that the poll was not manipulated.
The exit poll’s scale was large enough for analysis of votes in each of the nation’s 16 regions, counting Regions I to XII plus the National Capital Region, the Cordillera Administrative Region, the Autonomous Region in Muslim Mindanao, and Caraga, Mangahas said.