Thursday, November 16, 2006

How PDI and SWS Put Words in the Mouths of the Undecided

PDI's Amando Doronila thinks the Opposition is being overconfident about 2007 on top of which they are fighting amongst each other, they can't get their act together, they have run out of "patronage resources" -- so he advises them to commit hari-kiri instead, by mounting a full frontal attack on the Palace's strong side (where of course they would be slaughtered) --
Doro: "If the aim of the opposition is to destroy Ms Arroyo's power base, the House should be the target of the opposition. The congressional races are the decisive battleground for a tectonic shift of power in May. It's the outcome of elections for the House seats that will decide whether the Arroyo presidency will be a dead duck after May and not just a lame duck."
That's what the Palace wants the Opposition to do. There is nothing that matters more to the Palace now than the 2007 midterms. Everything it does can be understood in relation to that, and the nightmare they are trying to prevent. The whole nexus of worry is the Senate race and a Hanging Jury of Senator Judges from the Opposition.

Here is a new theorem about impeachment under the 1987 Constitution. The Lower House will never again attempt an impeachment like Erap’s -- where the Senate would NOT have convicted the President in a trial anyway. That was the lesson of the Second Envelope. That was the lesson of 2005 and 2006 impeachment attempts.

Only the people may impeach and convict the President in this realpolitik. By first electing a Hanging Senate--16 out of the 24, the two-thirds of all the Members of the Senate, sitting as Senator Judges, required to convict the President in an Impeachment Trial. It’s the only way to remove the hanging sword of Damocles over the required eighty necks in the Lower House, where not all of them are brave or moral or foolish enough to vote impeachment in the face of certain Doom from the Palace in the event of an acquittal. Next year there will be eight incumbents ready to convict. The Opposition only needs to win 8-4 in 2007 and we shall have the 2/3 in the Senate needed for conviction at trial. When that happens in the Senate, the House will become a house of full of suddenly courageous hammers, and guess who will look like the Hangnail?

MLQ3 believes that the real message to the Opposition ought to be that they better go after the undecideds -- a constituency he estimates from doing his own analysis of several past public opinion surveys on several large issues to be as large as 25% of the electorate. It seems there are a lot of Bong Austeros out there!. MLQ3's suggested strategy of reaching out to, and winning over the "Middle Forces" is far sounder than Doronila's Charge of the Light Brigade assault on Fortress Batasang Pambansa.

But let me also say something about undecideds in a public opinion survey. Getting any double digit percentage of undecideds in a survey is a sure sign that the Question or Issue put to the respondents is either not yet ripe for decision (say because the election or plebiscite is still far away) or because the Question or Issue itself will NEVER be decided in any future objective test like a plebiscite or election. It is a readily observable fact that as Election Day approaches, most election races "tighten" or "firm up" as the undecideds indeed, begin to decide. But "unscientific" questions or issues, those that will never be put to an objective test, will always have a very different behavior in the undecideds. A classic example of this was also mentioned by MLQ3 today -- the SWS survey on whether people doubted the fairness of an upcoming Supreme Court decision. That SWS poll was unscientific and deviously tricky because its question was loaded and ambiguous; its answers were scientifically collected but the interpretation of the data and the resulting headlines were fallaciously arrived at.

The SWS Media Release and the PDI headline that a majority of Filipinos were unsure of or doubted the fairness of the Supreme Court WAS true. But it was equally true that a majority of Filipinos doubted the UNfairness of the Supreme Court, because they were undecided HOW the Court would rule and whether the decision would be fair OR unfair.

Here is what basically happened in that poll. There is a Random Sampling Crystal Bowl like we had on the Explainer. But there are Red, Blue AND White Balls. The Red Balls are those who are sure the Supreme will be fair, the Blue Balls are those who are sure the Supreme Court will be unfair, the white balls are those who are unsure if the Court will be fair OR unfair, they are UNDECIDED either way.

The claim in the SWS Media Release and PDI headlines are decidedly DECEPTIVE because even though the SWS pulled a majority of white balls (undecided), what their Media Release and the PDI headline said was:

MAJORITY OF BALLS PULLED WERE NOT RED!

Haha! This was a case where SWS and PDI ended up putting words in the mouths of the Undecideds! (previous post on this topic).

Thanks MLQ3--for the opportunity to explain this in such a simple and direct way.

The exact question SWS asked in the poll was:

“Recently the Sigaw ng Bayan and ULAP raised their petition to the Supreme Court. Do you have much trust, are unsure, or have little trust that the Supreme Court would make a fair decision as to whether or not to grant the petition of Sigaw ng Bayan and ULAP regarding people’s initiative?”

The results were: 13% (much trust); 62% (unsure); 24% (little trust)

This is equivalent to a crystal bowl with 13% red balls, 62% white balls and 24% blue balls.

Now if SWS pulls mostly white balls, it is a true statement to say that a majority of the balls are "NOT RED." But that would be deviously disingenuous, just like the question itself. This is not a proper "binary" question because the creation of a category called "unsure" as part of the question itself is a self-referential flaw. There is no way to deduce why any respondent might answer "unsure". He could be unsure the SC will be fair, or unsure it will be unfair, or unsure what the question means at all, or unsure of how much trust or distrust he has for the Supreme Court! It might even mean a respondent was unsure what Sigaw or Ulap or people's initiative was. We are simply unsure what "unsure" could possibly have meant to the respondents when the survey was taken. But one thing for sure the Media Release and PDI Headline did not properly interpret the statistical data!

The PDI headline was: “62% of Filipinos doubt fairness of SC, says poll” while the SWS Media Release misrepresented its own data with, “62% unsure if SC will rule fairly on people’s initiative.”

This was a really sneaky way of putting words in the mouth of the undecided, because it is literally also true from the data that 62% were unsure if SC would rule UNFAIRLY on people’s initiative. This is an example of a loaded ambiguous question that makes a survey unscientific and propagandistic.

3 comments:

HILLBLOGGER said...

You've been nominated by mlq3 to form part of an ideal senatorial slate.

Why not go for it Dean?

OK, running in an election will need financing but am sure there will be loads of people who will contribute, no kidding.

Listen, I usually, OK perhaps not that often, with some of the things you say particularly where Bush is concerned but I do have great respect for your ideas. You will put a bit of action in that staid Senate of ours. That will be healthy.

However, you might have to give up your US citizenship.

baycas said...

in evidence-based medicine (EBM), there is such a thing as a critical appraisal.

critical appraisal is the process of systematically examining research evidence to assess its validity, results and relevance before using it to inform a decision. it is an essential part of evidence-based clinical practice that includes the process of systematically finding, appraising and acting on evidence of effectiveness. it allows us to make sense of research evidence and thus begins to close the gap between research and practice. (www.evidence-based-medicine.co.uk)

randomized controlled trials (or the so-called RCTs) are the “IN” thing in research because medical journals employing such trials can minimise bias and use the most appropriate design for studying the effectiveness of a specific intervention or treatment.

systematic reviews are also particularly useful because they usually contain an explicit statement of the objectives, materials and methods, and should be conducted according to explicit and reproducible methodology.

both, however, are NOT AUTOMATICALLY of good quality and SHOULD BE appraised critically. so, in clinical practice, we are assured of excellent patient care availed to us by our EBM-advocate doctor.

in relation to mlq3 and djb’s dissection of public opinion polls, it is really very important to APPLY CRITICAL APPRAISAL on all surveys done by the known pollsters. djb’s blogposts and the explainer last tuesday paved the way to uncover the hidden in published surveys.

if only the public opinion surveys are TESTED for their PROPAGANDA VALUES (or unworthiness) then, pretty sure, the public will no longer be misinformed.

…how to disseminate widely and make people understand the LIES from these polls is another thing…but mlq3 and djb’s endeavor is certainly a start…

costa rica said...

You did a very good work with this report; you have good and interesting information about this topic. It’s really good!!!
Do you want to see something more? Look at this:
Glass Bongs and Bong featuring Herbal Smoke, water bongs, bongs online head shop, Marijuana Alternative,glass water bongs, Hashish, Ganja, homemade bongs, Smokeshop, cannibis, legal smoking alternatives for herbal highs and aphrodisia. http://www.headshopinternational.com