Chief Justice Renato Corona: Statistically Guilty?
“There are three kinds of lies: lies, damned lies and statistics.”
– Mark Twain’s Own Autobiography: The Chapters from the North American Review
Pulse Asia released their findings on the Impeachment of Supreme Court Chief Justice Renato C. Corona from a March 2012 Ulat ng Bayan national survey and requested media assistance in disseminating the results to the public (http://pulseasia.com.ph/pulseasia/story.asp?ID=747). Of course, Philippine media obliged but interestingly, two major news/media organizations came up with two different decisions as to which part of the conclusion to focus on in their headline banners:
GMA Network Headline: “Pulse Asia: 47% of Pinoys Believe Corona Guilty” http://www.gmanetwork.com/news/story/252057/news/nation/pulse-asia-47-of-pinoys-believe-corona-guilty
InquirerNews: “43% of Filipinos Uncertain About Corona’s Guilt: Pulse Asia” http://newsinfo.inquirer.net/164531/43-of-filipinos-uncertain-about-coronas-guilt%E2%80%94pulse-asia
Pulse Asia issued only one summary statement regarding the guilt or innocence of the Chief Justice but GMA Network quoted only the first portion of the statement (“47% of Filipinos think Supreme Court Chief Justice Corona is guilty of the charges filed against him,…“) while Inquirer chose the second part only of the same sentence (“…43% are ambivalent as regards his innocence or guilt.”
Clearly, GMA and Inquirer made their own editorial judgement about what conclusion to headline and left it to the intelligence of their readers to digest the fine print.
47% or 15% Guilty, Statistically Speaking.
Pulse Asia stated that “Almost half of the Filipino population (47%) is of the view that Supreme Court Chief Justice Corona is guilty, with 33% saying he is probably guilty and 15% being certain about his guilt.” An examination of Asia Pulse Table 5 actually shows that the “47%” guilty is the sum of the 33% who said that the Chief Justice is “probably guilty” and the 15% who said that he is “definitely guilty.” In other words, to the survey question “Sa mga pinaparatang na pagkakasala kay Chief Justice Corona, masasabi ba ninyong siya ay.…” the respondents were allowed to choose from six (6) different responses (definitely innocent, probably innocent, undecided, probably guilty, definitely guilty, I don’t have enough basis to make a decision). Statistically, and this is a common practice in research, the % of responses to two or more categories are sometimes combined. Thus, the statistical guilt at 47% (33% + 15%). However, a researcher can also decide to apply a more stringent rule, in this case, to use only the “definitely guilty” (YES) or definitely innocent” (NO) % responses as the criterion for guilt or innocence.
In this case, the conclusion would have been “15% of Filipinos found the Chief Justice Guilty” or “1% of Filipinos found the Chief Justice Not Guilty.” However, the % of guilt or innocence does not seem to be the most interesting aspect of the survey. Given that the survey was conducted after the prosecution team has rested on their case, the more remarkable result seem to be that “85% of Filipinos were still undecided, probably uncertain, don’t have enough basis” about the guilt or innocence of the Chief Justice.
This is the humble submission of this representation: a stringent interpretation of the response categories % results (YES, NO, UNDECIDED) would indicate that a significant majority of Filipinos remain NOT FULLY convinced about the guilt or innocence of the respondent Chief Justice Corona. However, this representation will not go to the extent of concluding that the prosecution team failed in their task of convincing the Filipino nation of the guilt (political, judicial, statistical or otherwise) of the Chief Justice.
Let us wait for the next survey results since Pulse Asia announced that they will not stop conducting conducting surveys on the ongoing impeachment trial of Chief Justice Renato Corona. Would be interesting to find out any changes in the opinions of the Filipinos after the defense panel have closed their presentation.
Lies, Damned Lies, and Statistics: http://www.york.ac.uk/depts/maths/histstat/lies.htm