Pulse Asia president Ronald Holmes on Friday, Feb. 18, pointed out that pre-election surveys are not predictions of the outcome of elections but only provide snapshots of voting preference during the conduct of the survey.
Holmes was reacting to a statement made by former Senator Antonio Trillanes IV questioning the results of the latest Pulse Asia survey.
"Pre-election surveys do not really predict election outcomes. These surveys are snapshots of voting disposition at the time when they were conducted," Holmes said in a Viber message.
The result of the Pulse Asia's Pulso ng Bayan survey that was conducted from Jan. 19 to 24 and had 2,400 respondents nationwide showed that former Senator Ferdinand "Bongbong" Marcos Jr. got 60 percent of the voter's preference for president in 2022.
Trailing behind Marcos were Vice President Leni Robredo (16 percent), Senator Manny Pacquiao (8 percent), and Manila Mayor Francisco "Isko Moreno" Domagoso (8 percent).
On Feb. 15, Trillanes questioned the January 2022 survey results that were released on Feb. 13, saying that Pulse Asia had failed to "predict" the winning candidates in the 2004, 2007, 2010, and 2013 elections.
"For the 2004, 2007, 2010, and 2013 elections, however, the surveys that we conducted closest to
election day or on election day (exit polls) mirrored the actual election results. For example, the 2007 exit poll showed Sen. Trillanes in the winning circle," Holmes said.
In an ANC interview on Feb. 14, Holmes also pointed out that "if the elections were conducted at the time that we conducted the survey, yes (presidential candidate Bongbong Marcos would win)."
"But if you talk about whether this would be retained and whether this is predictive of what will happen in May 9, my answer is a definite no," he added.
Sampling method follows global standards
Trillanes also claimed that Pulse Asia's survey methodology was flawed due to the division of the country into four regions—National Capital Region, Balance Luzon, Visayas, and Mindanao--and allocation of 300 respondents per region for a homogenous sampling design.
But Holmes explained that Pulse Asia uses a multi-stage probability sampling to ensure that their surveys are "nationally representative."
"This also ensures the heterogeneity of the sample. The sampling method conforms to global scientific standards for the conduct of national surveys," he added.
The Pulse Asia chief also belied assertions that Davao City is being used to represent Mindanao saying it is "patently false."
Holmes said that Pulse Asia resumed face-to-face interviews of respondents since September 2020. He also assured that public health protocols were strictly followed.
This was contrary to the former senator's claim that it was impossible for the survey firm to have conducted face-to-face interviews in January as it was also the time of the Omicron surge.
Trillanes also claimed that the survey figures contradict what is being felt on the ground--where there is a “negative” perception of Marcos and a “positive” one for Robredo.
"The survey field work period is indicated in our media release. The survey results can be interpreted based on the events that transpired before the survey field work. These events may have influence the public’s voting disposition captured by the survey," Holmes pointed out in reaction to the former senator's concern that the result was "not real-time" and does not reflect the voter perception on the ground.