THE School of Statistics of the University of the Philippines in Diliman recently came out with a statement cautioning us about the results of preelection surveys.
While it is easy to interpret the statement as targeting only the conduct of 'kalye' surveys, where online influencers go around asking people about their preferences, it should be stressed that the statement has a wider application. It also applies to surveys the methodologies of which are suspect. This includes online surveys, and even those that are purportedly conducted by legitimate survey firms but are drawn from biased samples.
While statisticians and polling firms continue to caution us not to easily fall prey to bogus 'kalye' surveys, or social media polling, their concern is limited to the fact that these surveys fail to uphold scientific standards. They call our attention to surveys that are conducted with the intention of enhancing and amplifying a particular political figure.