The easiest thing to do with survey results is to dismiss them, to say that they “contribute to trending,” and that they “manipulate the minds of the people.” For me, these are patronizing comments that attempt reflect reality better than hard metrics with disclosed methodologies and sample sizes; the literary metaphors that override mathematical truths. It seems, at least to me, that campaign organizations and supporters of candidates are more inclined to believe inspirational messages and anti-survey rhetoric, than to use surveys as metrics to define what’s going right, and what went wrong.
It’s one of the the mantras of good campaigning, whether it’s for products or for people: show your metrics. You can do whatever you want with metrics: ignore them and entertain the possibility of disaster, or study them and entertain the possibility of success. Where companies demand key performance indicators and other metrics to show how much effort they’re expending and how successful their campaigns and projects are, politics could benefit from doing the same.
During the course of the 2010 campaign, political candidates, organizations, and supporters did not come forward with metrics of their own, to refute the exit polls made by the SWS and Pulse Asia. The polling organizations have been very open in disclosing methodologies, sample sizes, and everything else within the scope of what is scientifically sound. Any scientist worth his or her salt will do that because it lends credibility and scientific basis (among other things) to a study they devote a lot of time and effort for.
No one has ever debated the science of the survey: the refutations were always made of motherhood statements that seek to appeal to emotions, or ones that seek to coddle egos, or the illuminating affirmation that comes with standing toe-to-toe with science, calling it wrong, and making a run for it.
Think of metrics as measuring pants. You don’t affirm your waistline.
The sample sizes are defined by the calculations prescribed by statistics, and any statistician or mathematician is free to call the survey organization, ask for the technical details, and calculate the sample sizes themselves. Nobody refuted SWS and Pulse Asia based on the method of the survey, the calculation of the result, the compilation of the results, and the inferences and conclusions that came with the hard data they have. It was all politics, damaging reputations, and measuring things based on how many people attended a campaign rally. Once again, the Presidential campaigns were ran without measurement: that a spiel from a supporter and the relative density of people in a gymnasium is usually the metric of campaign organizations. Surveys? They’re just “black propaganda.”
When statisticians seek to disprove the results of a statistical study, they fight statistics with statistics. The biases and suspicions they have against the other statisticians are set aside in favor of the biases and suspicions they have against the actual statistics. You don’t go to the organization and tell them they’re biased just because you’re losing: you prove the assertion with hard metrics of your own that matches – or perhaps even goes beyond – the effort of the other party. That’s how science works. It’s not infallible, but disproving science with science is much more effective than outright dismissal, denial, or whining about it on TV, newspapers, or blog sites.
If the numbers are wrong, prove it.
Nobody fought metrics with metrics; instead, the numbers were dismissed as “black propaganda.” Nobody came forward with metrics of their own to dispute the results of the exit polls; rather, straw-man arguments are created to make the science seem “politically motivated” or “ill timed.” It reflects the inherent weakness of political parties and campaign organizations: that they do not devote the resources necessary to feel the pulse of the people, diagnose the campaign, and know where they’re going strong and what needs improvement.
All campaigns, whether it’s for elections or for advertising, use metrics as indicators of success and performance. All large corporations – some even smaller and less wealthy than some political machinery – devote the expense and effort to come up with metrics of their own to diagnose their campaigns. While it’s true that surveys are not perfect, and there’s much to be desired in how transparent survey organizations can be, they are metrics that cannot – and should not – be ignored by people who are really intent on winning or making an impact. To succeed, you have to assess. To assess, you need information. One way to get information is to measure things.
I guess that the challenge in 2013 or 2016 is for campaign organizations to do exactly that: come up with their metrics, and use every available exit poll and survey (whether they do it themselves or they have others do it for them) and not perceive them instantly as attacks, but take them up as challenges. Metrics can help them configure their campaigns for success, more than they do damage reputations. That way, they have a better gauge of how they’re doing in the real world, instead of living in their own bubbles devoid of numbers to say how well – or how bad – they are doing.
Or one can leave it all up to good advice that one never takes. Who would have thought: it figures.