Survey says…

After the EDSA Revolution, we only had the Social Weather Station (SWS) to let us know how voters ranked candidates and what their sentiments were on issues. Then, in the late 1990s, came Pulse Asia — a sort of breakaway from SWS — which eventually gained traction in the 2000s. Now, we have so many survey results from relatively unknown survey companies bombarding our feeds and screens, purporting to provide candidates and voters alike with insights into political polling and voter sentiment during election season.

But surveys are not as simple as they are made to appear when reflecting voter sentiments and trust/approval ratings. Most of the time, results are just posted online or reported on television or over the radio without any guidance to the public on how they should be interpreted.

To properly appreciate survey results, it matters when the survey was conducted (the survey period), how many respondents were used and how they were chosen, and what issues or events were trending or prevalent in the news cycle in the weeks before and during the survey. The number of respondents affects not only the cost of the survey but also its probability of accuracy. The issues and events of the weeks preceding — and during — the survey could factor in or influence, to a certain extent, the way respondents answer.

Then there are commissioned versus non-commissioned surveys. Commissioned surveys are paid for by a specific client (like a candidate or party) to gather data on a particular topic or question of interest. Questions are often tailored to the client’s needs, such as gauging voter preferences for a specific candidate or assessing public opinion on a policy the client supports. Non-commissioned surveys, on the other hand, are initiated and funded by the survey organizations themselves. They are designed to capture public opinion on issues of general interest and prevailing sentiments on social, economic, or political matters.

From the perspective of a candidate (or anyone thinking of running for election), you would ideally want to commission a baseline survey on your chances of winning a post — or for which post you are most likely to win (if you are eligible and considering multiple positions). You might also want to know what issues will concern voters going into an election so you can craft a platform responsive to their sentiments. For national posts, you would need another survey midway to see if you’ve gained ground or if adjustments are needed as election day approaches. Businessmen may commission surveys to decide whom to support and donate campaign funds to.

The next time you see a post of rankings or ratings, take it with a grain of salt. Its statistical value lies primarily as a tool for candidates, entities, and policymakers. To my mind, surveys have no real value to voters. In many ways, political polling has been used to condition the public’s mind during election season by reinforcing random choices made by strangers. If surveys are to be made public, they should be accompanied by a detailed explanation — or at the very least, a disclaimer — of what such numbers actually represent: the preference of 1,200 or 2,400 people, made to look like the will of an entire nation.