Now that the court’s decided, the Inquirer and Newsbreak focus on the inside story on the goings-on among the Supremes leading up to their vote. The Inquirer claims the chief justice kept his cards close to his chest; Newsbreak asserts the Palace knew how he’d vote weeks in advance. Justice Ynares-Santiago was the other justice of interest.
The Palace tells her faithful villagers to keep their torches burning and pitchforks waving.
Plan A remains on the table.
Plan B: opposition, of course, claims victory before battle begins. The Manila Times and the Business Mirror detail the “Pangasinan Express” and it’s preparations to go full steam ahead.
Plan C is a go, Manila Standard-Today dutifuly reports.
In other news: Metro Manila faces a water shortage. The United States cancels plans to set up a testing center for nurses in the Philippines -but might change its mind on two conditions. The first: better peace and order. The second: settle the nursing test leak controversy once and for all. The sweetener: the US could set up a testing system.
Overseas, the Guardian reports that as America prepares for historic mid-term elections and the implementation of reforms to prevent the 2000 electoral controversy in Florida, efforts at poll automation and modernization in as many as 10 US states may be headed for disaster. The Nation of Thailand says it now has the inside story on the Thai coup.
In the punditocracy, the Inquirer editorial dissects the Supreme Court’s decision and says it drew a line in the sand; while the Business Mirror editorial says it’s time to move on.
Fr. Joaquin Bernas SJ suggests the legions of Singaw were sloppy and deceitful in their lawyering and Raul Pangalanan maintains there was outright fraud. Rene Saguisag says that when a case is carried, then it’s carried and so there was no close call at all. He coins a phrase I love: the Speaker “wants to be Prime Rib or something.” Tee hee!
Ting Tiongco remembers what martial law was like.
Overseas, Dahlia Lithwick explains why a recent New Jersey court decision on gay marriage isn’t a case of judicial activism. Michael Wright on what the Thai coup was all about.
In the blogosphere, Jove Francisco says he had to look up the meaning of sangfroid because of a comment I made on his blog entry for yesterday. He reports that there’s an ongoing debate over whether the President brushed off the bad news of a Supreme Court defeat, or whether she took it like a blow to the solar plexus. Some Palace officials are said to be in a state of shock -but the carnival continues and the fight goes on. Local officials are part of the President’s entourage to China (preplanned, surely, when victory was considered inevitable) and the President gets a couple of days off to eat dimsum and consult her bankers in Hong Kong.
Bloggers such as In a Jaded World continue looking at the Supreme Court decision, or reacting to it, such as From the Barrel of My Pen (a Fil-Am), Daily Musings and Philippine Politics 04 (who are both delighted), Bunker Chronicles (irritated over Lambino’s continued defiance).
A Nagueño in the Blogosphere points out Wikipedia’s made it to a Supreme Court decision.
And please don’t forget to keep tabs on The Explainer on ANC blog.
Technorati Tags: constitution, One Voice, people’s initiative, philippines, politics, president, Thailand, Washington DC
djb -but there’s no fly. the sc said something based on probability based further on singaw’s own statements. the sws survey is, at best, an extrapolation from a sampling of opinions. you can do the science or math but it seems to me:
justices = extrapolation from fact admitted by singaw itself and based on documents submitted by singaw
sws = extrapolation based on the gathered views of a sample of individuals, asked questions that may or may not correspond to what the sc was asking -with a plus or minus sampling error.
that therefore, being more specific and involving far less extrapolating, the justices’ decision carries more weight, even methodologically, than the sws.
melvinsky,
My favorite definition of the word heuristic is that of physicist Richard Feynmann who is credited with saying in his Lectures, “A heuristic argument is a hand-waving argument.”
mlq3,
Central to the coming Motion for Reconsideration is the accusation that the Supreme Court became a trier of facts in this case. Indeed, we cannot deny that whatever reasoning it used, it concluded as a fact that “if ever, not more than one million signatories saw” the initiative petition itself. Carpio makes this argument to be generous and to allow for the claim that indeed SOME of the signatories must have seen it. Ideally ALL should have seen it, but what WAS the factual number?
The Supreme Court’s method was simply the fact that there were 10 signatures per sheet. They multiplied this by the 100,000 copies Lambino attested to printing and CONCLUDED that “not more than one million … etc” Their METHOD was purely arithmetic and logical. But the SWS GATHERED data, analysed it statistically and showed with known precision that MORE THAN one million had seen the initiative petition as of June, 2006. The SWS method was EMPIRICAL and I claim superior to the “back of the envelope calculation” that Carpio did in Lambino v. Comelec.
SWS has shown with real data that as of June, 2006, an important numerical fact the Supreme Court bases its finding of insufficiency of form upon, was WRONG!
This is significant because it goes to the heart of the accusation that the Supreme Court SHOULD NOT be a trier of facts, even if in most other cases testimonial and documentary evidence are both necessary and sufficient to establish a FACT.
In Lambino v. Comelec, the SWS methodology is clearly superior because it uses real data.
The truth of the following statement in the Decision is seriously cast in doubt by SWS’s raw data, (though not by their Media Release Bureau or PDI’s headlines.)
Lambino v. Comelec: “Not more than one million signatories saw the petition before they signed the signature sheets.”
The 2nd Quarter SWS data indicates AT LEAST 1.18 million HAD seen such petition by June 30, 2006, with 2 months of signature collection to go. (95% confidence level, plus or minus 3%, Z=2)
Such fact may not have been known to the Court, or couched in the noncomprehending, formulaic jargon of the Media reportage of surveys.
But that is what Motions for Reconsideration are for. The Conclusion of the Decision is probably secure, even if the Court accepts its facts were wrong, because of that revision thingie…
MLQ3,
It is true that the Supreme Court based its extrapolation on the testimony of the SnB, specifically Raul Lambino’s in the oral argument session. But even SnB/Lambino are perhaps not in possession of the SWS data, because EVEN the SWS does not seem to realize what its data actually says! But anyone can do the arithmetic based on Mahar Mangahas article yesterday, the essential figures of which are:
40% of 6.8% of the total number of voters (43.5 million adjusted for population growth) saw the petition as of June 2006, according to SWS.
djb, what i don’t understand is this:
1. sc said, lambino said, they printed 100,000 copies. they submitted sheets that showed a max. of 10 could sign per sheet. ergo, if a sheet were connected to a copy, then 1 million is the maximum -already assuming every single copy was read.
so that’s a deduction from very basic facts without going into whether the fact is true or not.
otoh, a survey takes a small group selected based according to a statastical method, gets their opinions, and then extrapolates how those opinions reflect a larger opinion. but it all aims to give an impression, but isn’t factual, just a snapshot of opinion.
it infers from the opinions of the few, what the many may be thinking. but how are we to know for certain how accurate the snapshot is? only say if you compared it to something physically measurable: such as, do exit surveys match how people actually voted, based on actually counting ballots, presuming the counting was clean.
otoh, 100,000 copies -fact.
the copies were distributed -accepted as fact.
the maximum that could have read them -deduction but not states as fact, simply stated as logical conclusion.
sws – impression based on deduction but in itself, the only fact is the survey took place and sws interpreted the data.
but sws data can’t be matched with sc data. it’s apples and oranges.
It was Chief Justice Panganiban himself, in a landmark decision protecting ABSCBN/SWS exit polls in 1998 as a genre of free expression that established surveys as a valid scientific tool for estimating how people voted in an election. Besides, statistical methods are accepted in product liability suits, medical malpractice, and many other more mundane situations. His ponencia in that case established his techie reputation. He says a lot there that might controvert your own view of statistical data AS data and not opinion alone. It is DATA about OPINIONS, but it IS data. Also, the precision within which such data outlines reality is well-established and well-understood even by the CJ. Of course, there were many other factors that actually went into their judgment of a fatal insufficiency in form.
But it cannot be denied, no matter what angle I look at it from: the SWS 2nd quarter survey data controverts the Supreme Court’s finding of fact.
The two approaches are perhaps apples and oranges, but the facts they seek establish, are about millions of people and in that situation, statistical survey techniques are unrivalled. Their data is empirical, the Court’s is merely documentary and testimonial.
MLQ3,
Note that my beef here is NOT with the Supreme Court at all, since none of this was brought to its attention in the case. I am aghast at the recent statistical question design and risible misrepresentations of their own data by SWS itself, and how the newspapers and other media, lacking the mental apparatus of a sophomore in Stat101, reports it all with no real understanding of what they are writing in headlines, articles and columns. Polling IS a genre of journalism–that much is obvious now.
But the citizens should be armed with the knowledge, simple non rocket science knowledge, about how to read and interpret the raw data of surveys. Luckily, SWS makes their questions and respondent data available to anyone. That bespeaks highly of their integrity. But I think in the Media Release Bureau are people who are not social meteorologists, but political editors and spinmeisters. Mahar’s piece was “guilty conscience” piece for raising doubts about the Supreme Court’s integrity..a trying too hard to look adoring piece…”SWS is consistent with the SC ruling” … glad he did that or I would never have noticed this little kerfuffle in the data.
djb, let’s do a show on understanding surveys!
jher, thank you so very much, i was afraid the shows on the parliamentary system hadn’t helped anyone clarify their thinking.
There is no question that theoretically the SC is not supposed to be a trier of facts. I personally am in a court case where inferences from a recent SC case are being used as case law, not a decision.
“It has been pointed out already that no knowledge of probabilities, less in degree than certainty, helps us to know what conclusions are true, and that there is no direct relation between the truth of a proposition and its probability. Probability begins and ends with probability.” John Maynard Keynes
Empirical date from a survey does not confirm facts it simply confirms the number of people’s understandng of the question.
The signature sheet had a question at the top. Only 100,000 copies of the petition were made and attached to the question and signature sheet. I personally have participated in a signature campaign to have the original VAT law recalled and we set up a signing station in Ayala, Makati itself. The actual process of explaining to people in forums and in the streets for people to understand that very basic right to recall a law of citizens rights on taxation. It is a grueling and time consuming process.
Now anyone who has involved himself in an election campaign or in
political action will know the time and effort required to explain issues and to get people to sign on even in areas where people should know better gives doubt as to the validity to any claim that even 1M people actually read and understood what they were signing. Even some urban poor areas here in Metro Manila were lined up by the office of the mayor to simply get their handouts and sign on the dotted line.
An exit poll is effective all things being equal but not in close fights. SWS exit poll on FPJ was wrong in some areas. In the U.S. there is a ban on broadcasting exit polls because of the time zone difference in national elections.
A broad majority of people in the U.S. believe that O.J. Simpson
killed his wife. A jury of his peers found him not guilty. Next time we could simply take surveys instead to decide on legislation. That means on every law on the local and national level the people decide which is the purest form of democracy in theory. Imagine the money we could save by simply using SWS.
Unfotunately what mathematicians and economists forget in so called extrapolations is the fact that they are based probabilities and all variables cannot be considered. Individual biases come into play.
Hence like former Treasury Secretary Rubin said that all elections are like the stock market. They operate on fear and greed. The Fraudsters of PI promised a utopia to the poor and unfortunate. They played on their poverty and created the evil in the the format of a bicameral and legislative government.
Justice Puno pointed out the example of Brown vs Department of Education. That was about equal rights before the the law. One of the lawyers who filed that case was Thurgood Marshall. He later became the first black man appointed to the U.S. Supreme Court. This case effectively canceld out the separate but equal decision of the SC.
One thing that mathematicians should understand that the SC is a Constituional Court and our laws are based on Common Law. That means it is based on empirical history of events. History does not stand still.
Justice Puno said that there is a need to look at the economic provisons in light of global realities. What has changed in world from 1987 till today.
Can you get a survey from all classes of people as to what Puno is talking about?
How would the survey question be written out. Do you believe globalization will uplift all the poor?
http://www.blupete.com/Literature/Essays/BluePete/LawCom.htm
“The common law is as a result of a natural sequence which hardened first into custom and then into law. It did not come about as an act of will, as an act of some group aware only of the instant moment, unaware of the nature and history of man. It come about as a result of a seamless and continual development, through processes we can hardly begin to understand; it evolved along with man.”
MLQ3,
I agree! Public opinion polling and statistical surveys need EXPLAINING. We shall do it! And I hope your readers here will help by doing the following. Please suggest questions about this topic that ought to be answered in a way that any intelligent layman can understand. What is it that YOU are puzzled about or really want to understand about the darn surveys, which have adopted this “Voice of Authority or Omniscience” that sometimes we agree with, but sometimes we don’t. Long before the year 2020, the population will top 100 million Pinoys. Divining their will, or even their separate opinions about leaders and issues can’t wait on the elections of every few years. That is where the power of the surveys comes from: they are fortune tellers, water-diviners, spell-casters, psychics, numerologists. They awe and intimidate those who believe that they “were weak at math.” But I tell you all that math is but a special area of common sense, that is more, err, common than in our human genome than people think. The only stupid questions are the ones we do not ask because we do not even know how. I think there is a very deep-rooted anti-intellectual streak in Philippine society that has always favored the “humanities” over the “sciences.”
Yet, what more distinguishes man from the other animals, than the sciences of reading, writing, and ‘rithmetic?
JAMES,
you have not cited any provision in Sigaw’s proposed constitution that supports the propaganda line you are peddling.
Cite at least one provision that will address the problems you cite.
And then, tell me how you can be absolutely sure that anything Sigaw promised will remain in the charter they are selling you when in fact their proposed charter says they can amend anything in it 45 days after ratification.
Like I said, you fell for the old bait and switch because you read the propaganda but didn’t read the proposed constitution.
Here’s some free advice that might save your life – READ THE FINE PRINT IN MEDICINE BOTTLES . It will help you avoid taking poison for your illness.
Mlq3 and djb,
Ya do a show on surveys nd please dissect that “net satisfaction” thing and also touch on what all those “did not respond” numbers mean ib the calculation and interpretation of net satisfaction ratings
Prognosticators and trade experts steeped in their theology of the exact science of the markets regaled the people of Mexico that the magic of liberalization would lift them out of poverty. The fraudsters in the Philippines should take note.
Every numerical and mathematical construct used in measuring Mexico’s economy in the last 15 years after NAFTA looked so good in the theoretical universe. Most especially purchasing power parity. A theory used primariluy by OECD countries that require a predicate before its application.
President Salinas then promised that Mexico would export their tomatos instead of their tomato famers. We all know that today the U.S. is building a fence to prevent the millions of desperately poor Mexicans from crossing the border. Have they not heard of ladders? We must’nt forget the drugs crossing over.
Yet the same theology is now being peddled by the snake oil salespeople in the Philippines. Once again using numeral contructs that are so far removed from Philippine realities on the ground.
Mexico is right next to the largest economy of the world. It’s economy today is so dependent on oil and remittances of its workers in the U.S. It has provided cheap labor to U.S. corporations in Mexico and in the U.S.. But still they cross over into the U.S. They are partly responsible for keeping minimum wages in the U.S. low. Unfortunately for them China’s and India’s wages are even lower.
Now the theologians are having second thoughts about their orthodoxy. Their numerical constructs somehow do not have any meaning and bearing on most of the people of Mexico. The results have been disapointing to the theoreticians and disastrous for a large percentage of Mexicans.
“That neo-liberal story may be true. But it is an excuse. It may not be true. Having witnessed Mexico’s slow growth over the past 15 years, we can no longer repeat the old mantra that the neo-liberal road of NAFTA and associated reforms is clearly and obviously the right one.” B. de Long
http://www.project-syndicate.org/commentary/delong51
http://www.project-syndicate.org/commentary/delong18
A wordlwide revolution in the next few generations is in the offing. Finance and labor markets are being globalized on a scale unseen before. Either the Planet will become a Planet of Slums or otherwise most of us will not be around to see. So far Marx’s prediction is coming true. The globalization of exploitation and oppression must come first on so many levels.
Weak states will disappear or be swallowed up. At the level of environmental degradation the Philippine islands are going through, water and arable land is being consumed faster than it is being renewed.
Now even the planet is probably at risk.
mlq3/djb, please proceed with a show on conducting surveys. One issue that a lot of people find difficulty in understanding is the minimum number of 1200 as a statistical sample to represent millions of people as a population.
the senate is worth it; without a senate GMA would be able to pass her ID, her cell phone registration, maybe even conscription. the small cost (1/10 of 1% of the national budget) is well worth it. I agree there could be better senators, but that means electing the CJ, and Sonia Roco, and not Lacson, Honasan and their ink
STATISTICS variant of melvinsky question:
If you need 1200 to capture opinions of millions of metro-Manila voters; 1,200 to capture opinion Ilocano region; 1,200 for ‘rice-granary’ region; 1200 to capture Visayas, 1,200 to capture Mindanao, and eleven (maybe forty-three) to capture the feelings of OFW’s… why again is 1,200 rock-solid as representative of all the Filipinos?
DJB, thanks for the feedback, most specially for spotting the unconscious switch in my reasoning from an argument about inputs i.e. changes made to parts of the Constitution (of Liberty, Govt and Sovereignty) to an argument about outputs i.e. the effect of a change in one part on the other parts of the Constitution. I believe Justice Puno also made the same slip-up when he tried to justify that with the change in ‘the constitution of government’, “The constitutions of liberty and sovereignty remain unaffected…”, as in this passage, he has made a switch to discussing the effects of such a change. To be consistent with his original line of reasoning, he should have remained silent or at least stated that any effects on the other parts of the constitution are irrelevant, which i believe is closer to your line of reasoning (re: ‘nontrivial’ changes coming from a PI) as well.
What this inconsistency makes me aware of is that aside from considerations of ‘degree’ (Carpio’s heuristic) versus ‘scope’ (Puno’s ‘partial change vs. total change’ formula), both of which deal with inputs to the Constitution, there is also the separate matter of the resulting effect of these changes which can only be evaluated on a case to case basis.
Presumably, as Melvinsky has suggested, an improved version of RA6735 can specifically address both the input and output sides i.e. the kinds of changes allowed to be made as well as limitations on the possible effects of these changes to the rest of the Constitution.
Given that there is no hard and fast rule for determining in advance how a change in one part of the Constitution can affect the rest, and the deferring to Abe’s concern with the dangers of judicial activism, i believe the enabling law should adopt the Puno interpretation with some limitations. The limitations have to do with safeguards against paradoxes that may arise from reflexivity, specifically self-destruction of the People’s initiative when applied against itself. This means that any proposed change via the PI should leave Article XVII ‘Amendments or Revisions’ untouched. Any change in this section should be considered a revision.
UP Student, according to DJB’s Oct 13, 2006 weblog post (‘Innumerate Survey Reporting By Social Weather Stations!’), the margin of error of +/- 2.89% is computed by getting the reciprocal of the square root of the sample size (i.e. 1/(1200^0.5). (The size of the population of Filipino voters is not a relevant variable in this case.) If you have a bigger sample size, then the margin of error decreases, but that would make the survey more expensive. For example, if the sample size is 10000 respondents, then the margin of error is +/- 1%.
DJB.. interesting point-to-ponder that you raised:
“A broad majority of people in the U.S. believe that O.J. Simpson killed his wife. A jury of his peers found him not guilty.” The court of public opinion says guilty; the court system said ‘not guilty’.
I was one of those who, based on all the stuff I had heard via media, thought OJ was guilty. The sum total of all my deliberations about OJ-guilt was probably 30 minutes of serious thinking, spread over 6 months; versus the jurors which were sequestered and had to sit through days of testimony and deliberation. I also suppose many of the stuff I gathered via media were tossed out (stricter rules of evidence to protect the innocent).
Hello guys,
A friend of mine and I share a same view on Mathematics. Being in fields that only occassionally deal with the higher levels of mathematics; we have the opinion that it is a subject that is certainly not easy to go back to. It is not an easy subject to grasp again by just reviewing the books.
Having said that, I take my hat off to those who still excel in it whether still practicing or not.
BUt does the SWS mean that the amendments and the petition are one and the same?
To be strict about it; the petition will contain the amendments but the amendments shown may not be inside a petition.
[…] What seems quite striking is that Chief Justice Panganiban has decided to join the majority to pay lip service to “people power†by effectively emasculating its implementation through people’s initiative. I am referring to the majority’s requirement that when it comes to complex amendment amounting to revision a deliberative body is demanded for the purpose. Reading this requirement into the constitution of sovereignty is amendment by judicial fiat at the expense of people power. Non-deliberative body is incapable of “informed judgment”? Certainly like many political blogosphere habitués, Justice Tinga, alluding to the nocturnal council and the covert rule of the “demigods” that produced an elitist American Constitution, was not convinced in this regard, and rightly so, by paying tribute to the exchanges in the Internet blogosphere, among other public spaces in the Athenian mold.
On the other hand, speaking for a majority in Lambino, Justice Carpio, who in the main managed to avoid the highbrow scheme that had attended the machinations of the facts and the law by the majority in Javellana, prove to be as straightforward as Concepion and as biblically deferential to the fundamental law as a civil law judge – well, if only he had steered clear not from Santiago but from resolving the amendment/revision question.
In a different breadth, aside from the blunt admission by Carpio that “there can be no fixed rule on whether a change is an amendment or a revision,†the precedential law in the Philippines appears to be against his “green to red spectrum†formulation; indeed, even in Javellana there was well-neigh unanimous concurrence (or in fact the matter was simply taken for granted) that revision was permissible notwithstanding that the 1935 Constitution only spoke only of amendments that, were it approved and ratified in a plebiscite, would form “part of this Constitution†according to its express language. […]
CVJ: Thanks for mentioning the bit of knowledge that every intelligent reader of public opinion polls has come to understand: the size of the random sample used in a survey DETERMINES the margin of error and should be the first thing one asks about any survey. Thanks to melvinsky for asking it.
But what is this “margin of error” thing?
Well first imagine that you have 1000 marbles, 755 white and 245 black. Clearly the exact proportion of white to black balls is 75.5 percent.
Now suppose you draw at random just TEN marbles and count how many are white and how many are black. Suppose you find 7 whites and 3 blacks. The best you can say is that the proportion of white to black balls is 70 percent. If another random sample contains 8 balls, you say 80 percent is your estimate. But you see, with just one draw, you can never hit the EXACT proportion of 75.5% because your random sample size is TOO SMALL. It does not afford you the arithmetic precision required to get the exact proportion.
So now you increase your size to say 100 marbles. One time you draw a sample and low and behold there are 75 whites and 25 blacks. Now your estimate improves in precision. Your estimate is 75 percent white to black. But you still cannot get the exact real percentage of 75.5 percent unless you use all 1000 marbles in your sample.
Same principle applies when you have millions and millions in the target population.
In SWS surveys, they pick 1200 respondents, and as the simple formula shows, the famous margin of error or plus or minus 3 percent is really a rounded up number from plus or minus 2.89 percent, the reciprocal of the square root of the random sample size of 1200.
The Margin of Error is the price we pay for the small sample size relative to the large population. It represents an UNCERTAINTY in any raw statistic we get, like what percentage of people are for or against chacha.
The Second bit of knowledge intelligent poll readers know is that these “errors” and “uncertainties” in the raw statistics COMPOUND and ADD to each other whenever we try to get tricky and start adding and subtracting the raw statistics to produce things like the Net Satisfaction Rating (which subtracts one statistic from another: percent satisfied minus percent dissatisfied.) Indeed the likely error or uncertainty in the NSR is plus or minus 6 percent or twice the error in each of the components. That is why NSR is largely a useless and erratic statistic, except when viewed on the scale of years worth of data.
MB,
When the SWS recently announced that the President’s Net Satisfaction Rating “inched up” from minus 14 percent to minus 11 percent, that was like saying, after finding 8 white balls out of 10, following a first draw of 7 out of 10, that “the number of white balls has inched up to 8” — a patent absurdity. But why? Because using our formula, the margin of error for a sample size of 10 is plus or minus 32 percent or plus or minus 3 marbles! What that means is with a sample size of 10, you cannot believe any change that doesn’t involve at lest 3 marbles.
Likewise, you shouldn’t believe any change in NSR that doesn’t involve at least a 6 percent change. Of course you are allowed to monitor the apparent statistic over time, But bear in mind, if you subtract the NSR from one quarter from that of another quarter, the error in that result will be plus or minus 12 percent! Making it largely useless.
By the way, as you already know, the polling firms INVENTED the Net Satisfaction Rating as a means of hiding the embarrassing proportions of UNDECIDED that are subtracted away when you take the difference.
It is when statistics are actually inconclusive that most of the people can be fooled by the propagandists and spinmeisters.
UP Student,
The case of OJ is said by some to be a case of JURY NULLIFICATION. The jurors in that case are accused of knowing he had killed his wife, but they still found him not guilty. The reason was that as the reigning icon of black people’s success in a white world, OJ ought not be dashed to pieces as a murderer of a white woman. But recall that he lost the civil suit and is said to have been impoverished by the judgments in it. He never ran through airports for car rental firm commercials again. But was Justice done? You decide.
Tale of the 2004 SWS Exit Poll
If you thought the distinction between REVISION and AMENDMENT was hard, wrap your brain around this one: PRECISION is not the same as ACCURACY.
When tailors measure waist size in inches, they typically use a tape measure that has tick marks separated by 1/16 of an inch. Any apparent change in your waste size will be recorded as some multiple of one sixteenth of an inch because a smaller change cannot be read precisely. Thus we say that a typical tape measure has a precision, or a margin of error, of one sixteenth of an inch.
Now suppose I want to impress my tailor so I breathe in real deep as he measures my waist size. He still measures with the one sixteenth precision of his tape measure, but he will get a completely INACCURATE reading.
In the same way, the Margin of Error we have been talking about in an SWS survey, is really the “reading precision” of the SWS poll. The tick marks on the SWS tape measure are 3 percent apart.
But any opinion poll result, can be wrong for reasons that have nothing to do with the random sample size, or any statistical factor. For example, take the 2004 SWS Exit Poll (which had 5000 respondents and precision better than 1% Margin of Error). It claimed GMA had won in NCR. It was 100% inaccurate. Comelec’s official result showed that FPJ won by a landslide here. (Yet the SWS 2004 Exit poll predicted the actual results in ALL regions of the country, except NCR & ARMM, to within the advertised Margin of Error!)
I actually met personally with Mercy Abad, the head of Trends MBL, which collects the survey data for SWS, and a representative of the SWS, in 2004 after the elections to call them out on this. They had no explanation. And when they tried to snow me with “it was just a glitch” I noticed they got very shifty eyed.
Fact is this: the Palace did not want nor could it afford for the worldwide headlines on the day after the 2004 elections to be that the opposition candidate FPJ had just swept the National Capitol Region (NCR), according its most prestigious polling organization. No. Mercy Abad and Mahar Mangahas went on nationwide TV, late on May 11, 2004 to announce to the world that their exit poll showed GMA had swept NCR! The news was carried by IHT immediately, the Washington Post, and others.
The statistical probability of that Manila exit poll being accurate, given that every SWS poll up til then showed FPJ winning Manila was one in sixty-six quintillion.
That to me was the nadir of SWS in an otherwise brilliant run since the 80s. I personally believe there was hanky panky with the random sample or the questionnaire data and that the security leak occurred not at SWS but at Trends MBL (an organization most people never heard of but which collects all the pollsters data!)
People aware of this grand anomaly in the SWS exit poll of 2004 have never gotten a satisfactory answer from SWS, much less Trends MBL.
They will be in action again in 2007!
does sws conduct reliability and validity tests of their instruments (i.e., questionnaires) before setting to capture the opinion of its actual sample?
DJB and MLQ3,
Please do have a show on surveys and statistics 101…..
The series on parliamentary governments did helped a lot on letting us people understand the issue,and for the critics that the show(The Explainer) is not for everyone…who said it was?
I forgot, I nowwork at nightshift…does the explainer have a replay on weekends?
djb,
kudos for doggedly pursuing the issue of the 2004 elections survey controversy. The thorougness in planning and the ‘presicion and accuracy’ of the execution of the grand conspiracy to assure GMA’s victory was/is impressive.
Malacanang poltical operator Tommy Claudio had given assurances that they are getting ready for 2007.
DJB… OJ-money is like Marcos-money….. Your facts are off regarding what happened to OJ after the liability lawsuit. OJ is still living a life of exceptional comfort despite the fact that in Feb 1997, OJ was found liable for Nicole Brown’s death. Yes, $33.5million was awarded to Nicole Brown Simpson’s estate and to Goldman’s parents. But OJ’s financial advisers had prepared for this eventuality, and OJ’s assets were large out-of-reach. [Ron Goldman’s parents did collect only about $500,000 through an auction of property seized at Simpson’s former home in Los Angeles — most of that had to be used for attorney’s fees and other expenses.] Simpson’s NFL pension and an annuity were out-of-bounds. Simpson spends his time playing golf. OJ’s position has been consistent through the years : “….If I have to work to pay them, I won’t work. It is that simple. So I’ll just play golf every day.”
DJB… I wish you put down the actual numbers when you described the 2004 election results for the NCR region. This sentence (especially the ‘100% inaccurate’ part) is quite a challenge to decipher: “It(SWS exit poll) claimed GMA had won in NCR. It was 100% inaccurate. Comelec’s official result showed that FPJ won by a landslide here.”
If SWS reported 45% for GMA (based on exit poll) from a survey-sample of 400 where 20 was discarded (illegible or some other reason), and Comelec-actual gave GMA 38%, how do you get “100% inaccurate”, is it?
A google-search on the expression “sample size margin of error” returns a number of websites on surveys and sampling. The web sites do not mention GMA; they have less histrionics; the “cold-blooded” reading material may get boring, but the science will be more robust. For example :
http://www.publicagenda.org/polling/polling_error.cfm
http://www.raosoft.com/samplesize.html
[cvj… I actually know a bit about the answers to the questions I raised. My statistics is good enough to be instructor at college- and MBA-level.]
For one, I always take surveys with a grain of salt. As someone said, “There are lies, damned lies, and statistics.”
If ever, I take survey results as mere indicators, but there is nothing absolute or sacred in their accompanying statistics.
To quibble whether one million, less than one million, or more than one million read the proposed amendments before they signed the petition is senseless hair-splitting, because no matter what the real figure is or was, the fact is that not all of the 6.3 million signatories (certainly not the dead whose names appeared on the lists) were given the opportunity to read the proposed amendments in full. This omission was fatal since it didn’t satisfy the constitutional requirement. And this is the only thing that matters, not the exact number of those who actually read the amendments.
Maybe Justice Carpio does not have the precise figure, but neither does Lambino nor SWS. Who cares? The incontrovertible fact is that Singaw didn’t comply with the constitutional requirement, and that’s all that matters.
By the way, are survey results admissible in court as evidence? Somebody please educate me.
Shaman.. on “are survey results admissible in court as evidence?” Of course… at least on TV. But kidding aside, “the Bystander” — atty erwin james — can cite Philippine law — article and section. As for Missouri:
(11) The results of statistical examinations or studies, or of audits, compilations of figures, or surveys, involving interviews with many persons, or examination of many records, or of long or complicated accounts, or of a large number of figures, or involving the ascertainment of many related facts, shall be admissible as evidence of such results, if it shall appear that such examination, study, audit, compilation of figures, or survey was made by or under the supervision of a witness, who is present at the hearing, who testifies to the accuracy of such results, and who is subject to cross-examination, and if it shall further appear by evidence adduced that the witness making or under whose supervision such examination, study, audit, compilation of figures, or survey was made was basically qualified to make it. All the circumstances relating to the making of such an examination, study, audit, compilation of figures or survey, including the nature and extent of the qualifications of the maker, may be shown to affect the weight of such evidence but such showing shall not affect its admissibility.
Shaman… brilliant question!!! Apparently, survey results are NOT admissible in Philippine courts. If there are no sections in Philippine jurisprudence similar to the Missouri law section that I just mentioned, then the “hearsay rule” (reproduced below) prevents the use of survey results:
5. TESTIMONIAL KNOWLEDGE
Sec. 36.Testimony generally confined to personal knowledge; hearsay excluded.  A witness can testify only to those facts which he knows of his personal knowledge; that is, which are derived from his own perception, except as otherwise provided in these rules.
*** In other words, the presentation of survey results is legally questionnable as there is no opportunity to cross-examine the interviewees on the oral or written statements that they allegedly made.
Precisely! Thank you, UP student.
So, DJB, where’s the fly?
UP Student, thanks for the links. The ’20 Questions Journalists Should Ask About Poll Results’ is handy. So it means that population size *is* a factor in deriving the margin of error for a given sample size, although as the raosoft link says, “the sample size doesn’t change much for populations larger than 20000”.
Good to have a Stat teacher around. (I myself got a 1.0 on the subject in College, unfortunately that mark was obtained in DLSU which has a grading system opposite to UP’s.)
DJB, thank you for the informative discourse on sample size in a statistical survey. The next most asked question and least understood is how to select the respondents as part of the sample size. What source records and how do you define and select the respondents?
cvj… I did not even notice that “20 questions to ask about poll results”. To anyone else interested, here is the link:
http://www.publicagenda.org/polling/polling_20q.cfm
Sample-size is only one of the 20 questions to ask. Four other questions to always have in the back of your head as you read a poll result:
(Number 2) Who paid for the poll/why was it done? …The important issue for you as a journalist is whether the motive for doing the poll creates such serious doubts about the validity of the poll results that the results should not be publicized.
(Number 4). How were those people chosen?
(number 5) What area … or what group were these people chosen from? Remember, the use of a scientific sampling technique does not mean that the correct population was interviewed.
(Number 13) What questions were asked?
On sensitive questions–such as abortion–the complete wording of the question should probably be included in your story. But at the very least, you must have the exact wording as you are preparing the story.
Sometime ago, I made the point that in most, if not all “statistical” surveys concerning GMA’s approval rating or about her policies and actions, the poll results consistently follow the voting pattern in the 2004 election. It appears that the same people who did not vote for GMA (some 65-70% of the voting population) but who divided their votes among FPJ, Lacson, Roco, Villanueva, and two other “also-rans” thereby giving GMA a winning plurality of 30%, are still anti-GMA.
While GMA enjoys a solid backing from her supporters, the utterly divided opposition could not gain any headway in their quest for unity, hampered by conflicting agenda and dubious motivations of its recognized leaders. So far, there is no evidence of substantial migration to and from either camps (except, of course, the Hyatt 10, Cory Aquino and Drilon, among the more notable ones, but who hardly made a dent on GMA’s presidency).
As long as GMA could stay the course of the nation’s economy (which by most account is on the road back to health), she should be able, at least, to maintaim the status quo until 2010. If the nation see a real economic turn-around as a result of her effort, who knows what the people would do to keep her in power and thereby ensure their future well-being.
UP Student,
Whoever said he takes his surveys “with a grain of salt” needs to realize that exactly the same scientific methods are used by Quality Control departments of, for example, food and pharmaceutical enterprises. If they took that attitude, we’d all be dead from food poisoning or the side effects of drugs that don’t meet numerical specs on purity or composition.Only when people don’t want to bother with figuring out what the statistics really mean have to take them with a grain of salt.
Shaman,
Surely you do not mean to imply that proponents of a people’s initiative have to PROVE that every single signatory has read and understood the initiative petition. Not even the Supreme Court said that. They use the SUFFICIENCY OF FORM as the requirement that at such a perusal was at least possible. But the germane point is that it is supposed to be the Comelec that determines such facts, not the Supreme Court. Of course the SC is also allowed to deal in facts and to make conclusions from them.
The fly in the ointment is that SWS data proves their alleged FACT wrong. That doesn’t mean their Ruling was wrong, only that the Court can no longer defend its decision SOLELY on the INDUCTION they reached that “not more than one million could have seen” the initiative petition.
Imagine na lang what you would really say if the subject of the initiative were a proposal that YOU support. What are the reasonable requirements on you as a proponent. Metaphysical proof that every single one of 12% of the voters have read and understood your amendment?
Remember: sauce for the goose is sauce for the gander. one day the people really will have a bright idea. I bet you change your tune then.
Bencard,
She hasn’t been impeached in the Lower House because there isn’t a Senate to convict her. Yet. If they had impeached her in 2005 or 2006, they’d all be dead after she was acquitted. They may be a bunch of moral dwarves in the House, but they’re not stupid. But all that will change once there are 16 hanging judges in the Senate. Senator-Judges, that is. It’s got nothing to do with economics, morals or anything else. It’s just who wins and who loses to most of these folks.
manuelb,
these are all paranoia, its really hard to deal with this problem.
and tell you what. it might not save your life at all if you take to heart what is “finely” written in the medicne bottle. that’s because no medicine is without side effects and these effects are finely written precisely because they are not the expected desired effect and could occur only idiosyncratically or in insignificant, negligible toxicity therapeutic ratio!
pretty much like the laws itself.
DJB, you may be right about the 16 hanging senator-judges, but I think if ever GMA would be removed involuntarily, it would have to be by a Parliament through a no-confidence vote. I think the senators’ days are numbered. The senate is right now the most expensive, cost- ineffective, unproductive and high maintenance agency of the government – a virtual trapo enclave.
DJB, I was the one who said that I take surveys with a grain of salt. Perhaps, I should have qualified it by saying opinion surveys.
Sampling of opinions is vastly different from sampling of inanimate objects. They cannot be compared. Opinions can be influenced by the way the q
DJB, I was the one who said that I take surveys with a grain of salt. Perhaps, I should have qualified it by saying opinion survyes.
Sampling of opinions is vastly different from samling of inanimate objects. They cannot be compared. Opinions can be influenced by the way the questions are worded or structured, by the way the respondents understand the questions or even the words in the questions, etc. In other words, opinion surveys can even be manipulated.
I’m not saying that surveys are useless. I’m saying that I take them simply as indicators to certain directions. On the particular question of exactly how many signatories read, much more understood, the amendments of Lambino’s initiative, quibbling about the exact number is, to me, irrelevant because the indication is that not even the majority read the amendments. I hold that everyone who signs an initiative upon petition must read and understand what he is signing for. How can anybody take an initiative, or take the lead, when he does not even know what he is espousing? The presence of signatures in a people’s initiative upon petition is not merely form, but there should be substance behind those signatures, and the substance is that each and every one of the signatories has read what is being proposed. Otherwise, if the signatures are to be taken to be merely form and devoid of substance, then signature-taking can be manipulated, as it had been manipulated by Lambino. The SC looked for proof that at least each and every signatory was given the opportunity to read the full text of the amendments by (1) printing the full text of the amendments on the sheets to be signed by the people, or (2) attaching a copy of the full text of the amendments to the sheets. On this score, Lambino was found to be dismally wanting, even if SWS may prove that more than one million, and not “no more than one million” read the amendments.
james, please don’t let font size be your guide in determining risk. ‘fine print’ in medicine bottles is mostly due to space constraints. As Andy Grove (the legendary Chairman of Intel) said, ‘Only the paranoid survive‘.
Bencard, i can’t see how you can say that of the Senate and not say the same of the House. Both the Senate and the House can be said to made up of ‘trapos’ (maybe not 100% of them, but certainly more than 50%). Given that trapos look after their own interests over and above that of the public, our protection rests on them not being able to conspire as a group.
To the extent that trapos in one chamber coalesce into a herd because of common interests and ‘group think’, our only safeguard rests on the trapos of the other herd who have their own interests to protect. Far from ideal and unfortunately not foolproof, but i’m afraid that’s what we have to settle for given that our leaders are an amoral bunch.
cvj, at least the House of Reps was able to pass a substantial number of bills, in between the distractions caused by some trying to impeach the president. The senate? I think it has conducted more investigations (in aid of legislation kuno) than bills actually passed. Personally, I think if there will ever be a Parliament, I would like to see more members of the present House(the 50% non-trapos you referred too) there, or new faces from other regions, e.g., Mayor Robredo of Naga, Cebu Gov. Gwen Garcia, Gov. Cruz of Bulacan, Mayor Duterte of Davao, and a few others who are real achievers, rather than “puro porma”. Of course, these people are executives, not lawmakers, but I think they will do a good job in any capacity just the same.
From: The Desk of Dr. Jude Borley,
Manager,
International Commercial Bank plc
First-Light, Kaneshie Branch,
Accra Ghana.
Attn: Dear,
I hope this may not constitute sort of embarrassment to you, I know you don’t know me very well, but First I must solicit your confidence in this transaction, this is by virtue of its nature as being utterly confidential and top secret. But I am assuring you that all will be well at the end of the day.
I have decided to contact you due to the urgency of this transaction, as we have been reliably informed of your discreetness and ability in transaction of this nature. let me start by introduce my self properly to you, My name is Dr.Jude Borley, I am the regional manager of the International Commercial Bank of Ghana first light kaneshie branch in the Western Region of Ghana.
I came to know you in my Private Search for a Reliable and Reputable Person to handle this Confidential Transaction. I am 45years of age and married with 3 lovely kids. It may interest you to hear that I am a man of PEACE and INTERGRITY; I only hope we can assist each other.
I have packaged a financial transaction that will benefit both of us, as the regional manager of the International Commercial Bank; it is my duty to send in a financial report to our head office in the capital city Accra at the end of each year. On the course of the last year 2005 end of year report, I discovered that my branch in which I am the manager made Eleven Million, Five Hundred and Fifty Thousand Us dollars [$11,550.000.00] which our head office are not aware of and will never be aware of. I have since place this fund on what we call SUSPENSE ACCOUNT without any beneficiary.
As an officer of the bank I can not be directly connected to this money, so this informed my contacting you for us to work together, so that you can assist and receive this money into your bank account for us to SHARE. The sharing of the fund are thus: 30% for you 50% for i and the remaining 20% for any legal expenses we have acquired in this transaction.
Note! there are practically no risk involved this transaction, is 100% risk free and legally blinded, it will be bank to bank transfer, all I need from you is to stand claim as the original depositor of this fund who made the deposit with our branch so that our Head office can order the transfer to your designated bank account.
If you accept this offer to work with me, if you find this proposal suitable for you, do furnish me with the following information’s: Your full Name, Contact Address, Tel/Fax Number, Date of Birth, Marital Status, Nationality, Occupation and Position. To enable me fix your information’s in our account data file, so that you will be recognized in our bank as one of our customers before the onward transfer will be made.
I will appreciate it very much. If this proposal is acceptable by you,Please Kindly reply me through my alternative email address box at : [email protected], do not make undue advantage of the trust we have bestowed in you. And how we can achieve it successfully. Please feel free to ask me any question.
Best regards,
Dr. Jude Borley
From: The Desk of Dr. Jude Borley,
Manager,
International Commercial Bank plc
First-Light, Kaneshie Branch,
Accra Ghana.
E-mail:[email protected]
Tel.:+233243601243
Attn: Dear,
I hope this may not constitute sort of embarrassment to you, I know you don’t know me very well, but First I must solicit your confidence in this transaction, this is by virtue of its nature as being utterly confidential and top secret. But I am assuring you that all will be well at the end of the day.
I have decided to contact you due to the urgency of this transaction, as we have been reliably informed of your discreetness and ability in transaction of this nature. let me start by introduce my self properly to you, My name is Dr.Jude Borley, I am the regional manager of the International Commercial Bank of Ghana first light kaneshie branch in the Western Region of Ghana.
I came to know you in my Private Search for a Reliable and Reputable Person to handle this Confidential Transaction. I am 45years of age and married with 3 lovely kids. It may interest you to hear that I am a man of PEACE and INTERGRITY; I only hope we can assist each other.
I have packaged a financial transaction that will benefit both of us, as the regional manager of the International Commercial Bank; it is my duty to send in a financial report to our head office in the capital city Accra at the end of each year. On the course of the last year 2005 end of year report, I discovered that my branch in which I am the manager made Eleven Million, Five Hundred and Fifty Thousand Us dollars [$11,550.000.00] which our head office are not aware of and will never be aware of. I have since place this fund on what we call SUSPENSE ACCOUNT without any beneficiary.
As an officer of the bank I can not be directly connected to this money, so this informed my contacting you for us to work together, so that you can assist and receive this money into your bank account for us to SHARE. The sharing of the fund are thus: 30% for you 50% for i and the remaining 20% for any legal expenses we have acquired in this transaction.
Note! there are practically no risk involved this transaction, is 100% risk free and legally blinded, it will be bank to bank transfer, all I need from you is to stand claim as the original depositor of this fund who made the deposit with our branch so that our Head office can order the transfer to your designated bank account.
If you accept this offer to work with me, if you find this proposal suitable for you, do furnish me with the following information’s: Your full Name, Contact Address, Tel/Fax Number, Date of Birth, Marital Status, Nationality, Occupation and Position. To enable me fix your information’s in our account data file, so that you will be recognized in our bank as one of our customers before the onward transfer will be made.
I will appreciate it very much. If this proposal is acceptable by you,Please Kindly reply me through my alternative email address box at : [email protected], do not make undue advantage of the trust we have bestowed in you. And how we can achieve it successfully. Please feel free to ask me any question.
Best regards,
Dr. Jude Borley