Opinion poll has always been considered controversial. It is criticized either because it does not support the stance of one group or it does not really convey the sentiments or opinions of all in a given place during a particular time. Therefore, some argued that no one should pay serious attention with poll results which are only reflective of opinions and not facts.
But why do groups who are favored or advantaged by poll results uphold it? They even cited these results to support an argument or publicly announced that the results really reflect what the people feel about an issue that they are promoting or fighting against with. Does it mean that the value of a poll is always dependent upon what it conveys and who is benefited?
The ambivalent attitudes toward opinion poll could be true to some government officials or political candidates relative to as to how they perceived the poll results. They generally considered poll results as political tools to manipulate the thinking and decision-making abilities of the public, which I will not deny, and this is actually similar with the mass media or any means of disseminating information.
Human behavior experts would argue that it is how the information is being disseminated that influences opinion and eventually behavior such as when the public is only exposed to one side of the story–a technique known as card stacking. Similarly, it is how the poll is conducted and the results are used that make it a political tool, but it is not inherently political.
The criticisms against opinion poll usually center on the following issues: the agenda of the sponsor of the poll, the method employed, and the alleged fabrication or falsification of the results. Critics argue that a sponsor may influence the pollsters to produce results the former desired, that a method employed may be designed to produce the desired results, and worst, that the results may be manufactured to satisfy the sponsor of the poll. These allegations are possible when the ethical standards of opinion poll are not strictly observed, and when those behind it are basically driven by money and not by its importance for making informed decisions or votes.
I would categorically say that opinion poll is methodologically acceptable and can produce quality results. This is possible if the proven design or the procedure is strictly followed and without compromise by the poll supervisors and field interviewers–people that determine the quality of the poll results. Therefore, the problem if ever they will arise is always rooted on the execution of the design of the poll which is influenced by the commitment to its sanctity of those conducting it.
In my experience in an opinion poll on local governance in Negros Oriental under the tutelage of the Social Weather Station (SWS) researchers in 1997, strict observance of poll procedures means that random sampling of the sites and respondents was a must given a statistically representative sample size at various levels. Sampling started from the list of municipalities then to the barangays and down to the households using on site-sampling technique. It was at the household level that the probable respondents were randomly sampled and personally interviewed–not sampling for convenience.
In the 1997 poll conducted by the Interdisciplinary Research Group of Silliman University (see Betty C. Abregana. Negros Oriental Poll: Public Opinion on Provincial Issues and Gubernatorial Candidates, Convergence Volume 4, 1998, pp. 17-29), we asked about the preferred gubernatorial candidates of only 150 randomly sampled respondents. The specific question we asked was: “Among the names in the list, whom would you likely vote as governor of Negros Oriental if elections were held today?” The candidate identified in the poll who the respondents will vote really came out as the governor in the 1998 election. This boosted my belief on the predictive ability of a well-designed and strictly supervised poll.
The respondents were also asked how they knew about particular candidates. The first three ways reported were the following: seen or met them in person, through the radio, and mentioned by other people. In terms of desired qualities of a candidate, they will vote one who must be approachable (dali duolon), helpful to the poor (matinabangon sa pobre) and has achievements (may agi o nahimo). Being an incumbent government official (opisyal sa gobyerno karon) and supported by the church ranked only fourth and fifth, respectively, among the seven qualities listed. Clearly, the ways of knowing the candidates and their desired qualities are reinforcing each other.
But take note that poll results are time bound and influenced by the prevailing condition or event at the time the survey was conducted. So certain aspects of local voting behavior mentioned earlier were only true in the past and may be not applicable now, but lessons can be gleaned from these by political strategists. For example, personal encounter with the voters and through their social networks, particularly relatives, may be still effective campaign strategies at present particularly in rural communities, and that the candidate who has projected some or all of those aforementioned desired qualities may have greater probability to win as governor.
These conditions were true in 1998, let us find out if these will still hold true in the 2013 election.
