Philippine Daily Inquirer
A lawyer and former Leyte representative yesterday questioned the results of the latest Social Weather Stations (SWS) survey, saying it was skewed to favor Sen. Grace Poe over Vice President Jejomar Binay.
Former Biliran, Leyte, Rep. Glenn Chong, a convenor of the Citizens for Clean and Credible Elections, said the distribution of the sample size of 1,200 respondents did not reflect the actual voter distribution in the four geographic sites commonly used by pollsters: Metro Manila, the Balance or Rest of Luzon, the Visayas and Mindanao.
In a non-commissioned survey conducted June 5 to 8, SWS asked 300 respondents each from the four sites to give up to three names of persons they thought should succeed President Aquino. Poe was named by 42 percent of the respondents; Binay, 34 percent; and Interior Secretary Mar Roxas, 21 percent.
“[There was] a skewing of sample size in favor of, or against, the Top 2 main contenders,” Chong said in a statement posted on his Facebook account, referring to Poe and Binay.
Saying that he was not for Binay, Chong said his suspicions had two assumptions: one, that the “obvious” campaign for Poe and the “demolition” of Binay would be “most apparent and effective” in areas closest to Metro Manila; and two, that Poe traces her roots to the Visayas while Binay traces his within the Balance or Rest of Luzon.
Using Commission on Elections data as of October 2013, Chong said the proper number of respondents should have been 138 for Metro Manila; 248 for the Visayas; 537 for the Balance or Rest of Luzon; and 277 for Mindanao.
According to Chong, Metro Manila has a voting population of 5,996,193 or 11.53 percent of the national voting population of 52,006,910, so the “correct sample size” should be only 138 respondents which is 11.53 percent of 1,200 respondents.
“By arbitrarily polling 300 respondents in Metro Manila or 162 more respondents than necessary, the survey results were affected by up to a maximum of 13.5 percentage points (12 respondents = 1 percentage point). This situation clearly favors Poe as against Binay under Assumption No. 1,” said Chong, who rounded off most of his computations.
Visayas meanwhile has a voting population of 10,735,024 or 20.64 percent of the total, so the correct sample size should be 248 respondents. The additional 52 respondents affected the survey results by up to a maximum of 4.3 percentage points, which “clearly favors Poe against Binay under Assumption No. 2,” the former congressman said.
The Balance or Rest of Luzon meanwhile has a voting population of 23,277,506 or 44.76 percent of nationwide voting population and should have had 537 respondents or 44.76 percent of 1,200 respondents.
With 237 less respondents than necessary, the survey results were affected by up to a maximum of 19.75 percentage points, “(which) clearly favors Poe against Binay under Assumption No. 2,” Chong said.