MANILA (UPDATED) - Pulse Asia's chief on Monday explained that the polling firm has been using the same survey sampling method for the past two decades.

How Pulse Asia conducts its surveys—we asked its president to walk us thru the process

• How Pulse Asia conducts its surveys—we asked its president to walk us thru the process

According to Pulse Asia President Ronnie Holmes, they use the probability sampling method, and they do not set quotas on respondents based on their age or educational attainment.

"'Yung sampling namin hindi naman nagbago nitong nakaraang dalawang dekada. Ito ay probability sampling, so ibig sabihin, lahat tayo ay may pare-parehong patas na tsansa na maaring mapili," he told Teleradyo.

(We have been using the same sampling method for the past two decades. This is probability sampling, which means everyone has a fair chance to be chosen as a respondent.)

"Hindi namin sine-set 'yung quota para sa mga edad at kung anong natapos mong pag-aaral," Holmes added.

(We do not set a quota based on age or educational respondent.)

Holmes was responding to opinion pieces written by statistician Romulo A. Virola, who said that some groups are overrepresented in the Pulse Asia's February 2022 surveys, while others are underrepresented.

• Statistically Speaking: Some questions on data and surveys for the Halalan 2022
• [OPINION] Statistically Speaking: Gen Zs, Millennials underrepresented; 'Boomers' overrepresented in Pulse Asia Feb survey
• [OPINION] Statistically Speaking: Pulse Asia Feb survey overrepresented voters who did not reach college

Virola is the former secretary-general of the then-National Statistical Coordination Board (NSCB) of the Philippines, now part of the Philippine Statistics Authority.

According to Holmes, with their sampling method, it is really possible to have more respondents from a certain age group or bracket, since they only need the respondents to be over the age of 18, a registered voter, and likely to vote in the May 9 elections.

"So sa ganitong sampling, ang mangyayari nga niyan, ang mga makukuha mong respondent ay maaring manggaling sa iba't ibang grupo at hindi necessarily proportionate doon sa sinasabing laki ng grupo na 'yun," he explained.

(In this kind of sampling, you can get respondents from different groups and the number may not be necessarily proportionate to the size of the group.)

"Pero ito 'yung limitasyon noong aming multi-stage probability sampling. Maari kang gumamit ng ibang sampling method pero kapag gumamit ka ng ibang sampling method, dagdag ang cost doon, mas maraming trabaho, at mayroon ding maaring pagkakamali 'yung ibang sampling method," Holmes added.

(This is the limitation of our multi-stage probability sampling. You can use other sampling methods but if we do, that will add more costs, more work, and there might be errors in other sampling methods.)

Holmes maintained that respondents are chosen randomly, which means the results of the survey may reflect the overall opinion of voters in the country.

"Ang sinasabi ko lang dito, dahil sa pamamaraan na aming ginagamit, hindi mo pinipili ang respondent batay sa kaniyang edad, batay sa kaniyang natapos na pag-aaral. Sila ay pinipili na sila ay adult Filipinos, registered voters at likely to vote," he said.

(What I'm saying is that because of this method, you do not choose the respondents based on their age or their educational attainment. They are chosen because they are adult Filipinos, are registered voters, and are likely to vote.)

"Random 'yan. So ibig sabihin, kung anong nakuha natin dito sa survey, maari siyang kumakatawan doon sa kabuuan ng ating botante sa ating bansa," Holmes added.

(This is random. So it means the results of the survey may reflect the overall voter preference in the country.)

Watch more News on iWantTFC

In a separate statement released Tuesday, Pulse Asia responded to the "flaws" mentioned by Virola.

On the under-representation of socio-economic classes A, B, and C

Pulse Asia said Virola employed the Unified Socio-Economic Classification (1SEC) that was at was proposed by a study group composed of academics and market research practitioners, which originally had 9 SEC clusters.

"This is different from the regular SEC used by most market research firms and Pulse Asia that divides the population into five groups, A, B, C, D, and E. Given this difference, it would be difficult to simply deduce that certain clusters from 1SEC could be classified as A, B, C, D, or E. It is also important to note that the 2017 1SEC cited by Dr. Virola continues to be subject to further validation," it said.

On the under- representation of the youth (various age groups from 18 to 41) and those who have higher educational attainment.

According to Pulse Asia, its method probabilistically selects respondents from various socio-demographic groups.

"In our reports, however, we note the margin of error for each socio-demographic group (SDG). This margin of error reflects the variance for the SDG given its share of the total sample of the survey and corrects, to a significant extent, what Dr. Virola finds as an under/over sampling of specific SDGs," it added.

Pulse Asia, likewise, said they agree with Virola calling his assumptions as "arbitrary", adding that: "We agree with his labeling of his assumptions as “arbitrary” as survey data, not only in March 2022, but in earlier and later (December 2021, January 2022, February 2022, and April 2022) pre-election surveys, show that Marcos has a marginally or significantly higher support in the groups that Dr. Virola assumed would be more supportive of Vice President Robredo."

"Thus, if we follow Dr. Virola’s logic, a larger sample from the younger and more educated groups may actually inflate Mr. Marcos’ support," it added.

Pulse Asia said it respects the criticisms of its survey results and methods, and reiterated that the survey results are time bound, and are not meant to predict actual election results.

It, however, clarified that Pulse Asia has not been bought nor infiltrated by partisan groups.

"Our organization will never submit to any form of material inducement or even intimidation that will make us deviate from or distort accepted social science principles and practice. We have also taken all the necessary safeguards to secure the integrity of our field work against any and all forms of infiltration," it said.

"Finally, the Supreme Court in 2001 struck down as unconstitutional a provision in Republic Act 9006 that prohibited the publication of survey results close to election day as it constitutes a prior restraint on the freedom of expression, among others, Pulse Asia added.

• Pulse Asia: On disinformation regarding our pre-election surveys

Former Sen. Ferdinand "Bongbong" Marcos Jr. is still the frontrunner in the May 9 presidential race, according to the latest and final survey conducted by research firm Pulse Asia in April, results of which were released Monday.

According to the polls held April 16 to 21 among 2,400 respondents nationwide, Marcos retained his March survey rating of 56 percent.

Vice President Leni Robredo, meanwhile, got 23 percent, one point down from her rating in March.

• Marcos tops Pulse Asia's final survey among pres'l bets; Pacquiao overtakes Isko

Pulse Asia Research Director Ana Tabunda said last March that Marcos was the preferred candidate by majority of younger voters as well as older voters, based on the February 2022 survey.

"Mataas din naman doon sa mas nakakatanda, pero mas mataas din sa mga mas nakababata," Tabunda said of Marcos Jr.'s supporters in terms of age groups

Pulse Asia has correctly predicted the results of the country's national elections since it began polling in 1999.