With just mere days away from the elections, the tense battle among the presidential candidates got even more heated when the survey results published by the polling body, Pulse Asia Inc. showed static movement with the voter preferences of the top two contenders “Bongbong” Marcos and Vice President Leni Robredo.
The survey results released by Pulse Asia showed that from April 16 to 21, BBM’s voter preference score was retained at 56, while VP Leni’s was reduced by 1 point to 23 (within margin of error). The lack of significant gains and losses along the scores sparked outrage among the public. Seeing how competitive the top two candidates are in capturing the attention and trust of the Filipino public – with their weekly rallies amassing hundreds of thousands of supporters, endorsements by the vast spectrum of well-known figures from the public service to showbiz sectors, and massive grassroots initiatives both in the urban and the rural communities – to a significant number of people, a 33-point gap seems to be a manifestation of a miscalculation.
"Biases" of Pulse Asia
Political and statistical analysts alike are casting doubts on the methodologies used by Pulse Asia on both its most recent and past pre-election surveys. In a Rappler article on the same topic, UP Economist JC Punongbayan affirmed the criticisms of Dr. Romulo Virola, the former secretary-general of the National Statistical Coordination Board, who indicated questionable processes of the statistical agency. He argued that Pulse Asia has been gathering results from a sample that is inequitable with the representation of various socio-economic classes, evidently manifested in the absence of results from classes A and B. Punongbayan raised this as a concern, given that according to PSA’s 2017 socio-economic classification system 1SEC, classes A and B constitute to about 9 percent of the total population. Romulo further included in his #FactsFirst discussion with Christian Esguerra, that there is also a possible underrepresentation coming from inequitable distribution of respondents within the sample that are of different age and educational attainment. Simply put, there is a possibility of “biased” preference results, given that they are from a sample that is distributed inconsistently according to population data of various socio-economic classes and demographic groups.
“Obsolete” Methodologies?
While Punongbayan and Virola are contemplating their doubts over the Pulse Asia survey, another side of the discussion is taking the skepticism up a notch. In a recent episode of Rappler's On the Campaign Trail with John Nery, Renowned Big Data Analysts Roger Do and Wilson Chua discussed their disagreements with the Pulse Asia results based on the different conclusions generated by what they consider a more up-to-date method: Social data. By analyzing social engagement patterns on campaign-related social media posts, Do and Chua have posited 2 findings that are contentious with Pulse Asia data: that BBM and VP Leni’s margin from the start of the campaign period is not as wide as previously indicated, and that VP Leni has taken the lead over since late February. As of late April, Leni Robredo has a lead margin of about 4% against BBM, based on the social media engagements in Facebook. Chua took note that the case is not isolated to the specific social media platform – the pattern is present on almost every other social media applications, including Google, Youtube, Twitter, among others. These implications indicate that generally, BBM and VPL are head-to-head with the fight towards winning in the election race..
Gathering these, a question may be raised: how reliable is Social data?
While it has still yet to make a name for itself in the Philippines, Social data already is a widely-reliable tool to indicate social patterns in the advent of the Internet Era. As an estimated 73 million Filipinos are actively using the Internet, Roger Do and Wilson Chua further explain that Social data analytics companies take advantage of the internet penetration rate opportunity to generate analyses on social trends and patterns within hours. In the context of elections, this gives the method the capability to quickly gather voter preferences in a very large sample group without the need to take major account of the common physical surveying concerns, such as sponsorship and funding, physical logistics, Hawthorne Effect, and many more. Furthermore, Social data also provides the opportunity to take a look beyond the rigid preference findings commonly presented by physical surveys. While the latter’s data shows how many are voting for which and how many are still undecided, the former can give details to the nuances beyond such preferences. By looking at the kind of reaction given by the specific individuals in the sample, Social data analysts are able to look at the different levels of interest shown to a specific candidate, and see which are the other candidates a sample is showing interest to. This provides us a clearer picture of the estimated voter preference showing not just the decided and the undecided, but also those who are still weighing the options – the soft voters.
As per Roger Do’s and Wilson Chua’s findings, a total of almost 40% among engagements last April 11 have shown “swing” reactions toward different candidate advertisements, indicating a significant percentage of individuals manifesting insufficient certainty over a specific candidate. This follows a pattern with the previous 2016 SWS exit surveys, which indicated that 33 percent of those who voted had only made their decision just days before or on the Election day itself.
Reassessing our Perspective towards Surveys
It’s okay. The numbers are over. We can relax a bit now.
Just as what Punongbayan, Do, and Chua said in the interview, surveys conducted by Pulse Asia and other proven analytics agencies have undergone professional standards of processing before being released to the public. The results are still well-grounded on real data and proven methodologies. It would be unfair to hastily brand them as unreliable. After all, Pulse Asia’s President Ronald Holmes provided explanations regarding all the questionable methods they used for their surveys. However, just how and why we utilized both in showing the estimated number of current soft and undecided voters says our whole point: we might need to take the survey results with a little less grain of salt. The job of these surveys is not to generate predictions, but to capture a picture of reality at a certain time and make the best inferences out of every data that can be gathered – and just like everything else, it is not immune to flaw, error, and inconsistency. But this is never a reason to easily sow distrust in these methods of study. The discipline of conducting surveys is founded on continuous criticism, change, and development – and baseless attacks against the credibility of these polling bodies is an attack not just to these institutions, but to the principles of democracy this country is founded on.
This is why it is a common practice to look for and compare more results from different reliable sources, to always keep an open and critical mind, and to engage in constructive discourse, in order to get a better view of the wide social phenomena. A SINGLE SURVEY DOES NOT DICTATE THIS WHOLE ELECTION STORY.
We then go back from the start. What should we do?
As there is a day left from May 9, all that is left to do is one that we frequently hear on social media: Focus lang tayo. As previously said, there are still a lot of people who have yet to decide who to vote for this coming Election. Vote using the agency that is innate in you. Surveys should not be the sword that would rule over who you vote for, but rather the tool that would help you write the next story of our country. Using your free will, and the democratic rights endowed to you by the Constitution, take the time and opportunity to know what is the utmost need of the Filipino people, and decide which among the ten is likely to fulfill such.
Lastly, let us keep calm, maintain our composure, and do not let these survey results change our dedication to promoting the values and plans that we deem as needed by our people. Rather than to spread distrust, let us instead spend our precious time left engaging with our kapwa-Pilipino, even in the most subtle ways (with respect to COMELEC’s active campaign ban). Let us help them in recognizing that a government of the people is led by one who has utmost care, compassion, and service for the people. After all, even Pulse Asia says so: nothing is yet final.
Let us vote wisely.
Comments