SWS poll: 79 percent want GMA impeached

[explorer]

from: http://www.abs-cbnnews.com/storypage.aspx?StoryID=15940

Seventy-nine percent or Filipinos wanted President Arroyo impeached over alleged poll rigging and bribery, according to a survey released Saturday.

The third-quarter survey by the pollster Social Weather Stations had 1,200 respondents and was made just before the House Committee on Justice voted to dismiss all three impeachment complaints against the President.

"Anti-Arroyo feelings ran very high in the last few days of the recent hearings of the House Committee on Justice," said the survey, which had a 3-percent error margin.

The poll found 79 percent of the respondents wanted Mrs. Arroyo impeached because of the prevailing sentiment that her telephone conversation with an election official was tantamount to cheating.

"Most Filipinos [believe that Mrs. Arroyo’s] admitted phone calls to a [Commission on Elections] official amounted to instructing him to cheat in the 2004 election and were not merely meant to protect her votes as she claimed in her June 27 apology," the SWS said in a statement.

Presented other exit options for Mrs. Arroyo, 64 percent said they wanted her to resign, and another 51 percent said she should be removed through a people-power revolt.

The results of the survey disputed claims by 158 lawmakers, in voting to throw out all three impeachment cases against the President, that their position reflected the sentiments of their constituents.

Refusing to accept defeat, the advocates of impeachment said they would resurrect the case by filing a motion for reconsideration once Congress resumes session on September 19.

Rep. Alan Peter Cayetano of Taguig-Pateros said the proimpeachment group needs only one congressman who voted to dismiss Oliver Lozano’s complaint as lacking in substance to file the motion during the plenary session.

He said once the "yes" congressman files the motion, it will be seconded and put to a vote. But the impeachment advocates need 27 more endorsers to complete the 79 to be able to send the complaint to the Senate. Cayetano said this is allowed under House rules.

The proimpeachment lawmakers decided not to file a case before the Supreme Court to question the House’s dismissal of the impeachment complaint. Instead they will bring the case directly to the people through demonstrations and rallies.

The 236-seat House voted to dismiss the cases against Mrs. Arroyo at the end of a 23-hour session on Tuesday.

[IceColdBeer]

Sabi wag daw paniwalaan ang mga survey

[roninblade]

san kaya nila kinuha yung respondents nila?
1,200 people represents the entire country?
i dont think so. this is a stupid press release.

[explorer]

san kaya nila kinuha yung respondents nila?
1,200 people represents the entire country?
i dont think so. this is a stupid press release.

all reputable survey companies in the country are in that range of respondents in making their survey and most of their survey results as proven in the past have only less than 5% error baseed in the past elections. if you don't know how the survey works, then i understand why you made a comment like that.

[DoctorDoom]

all reputable survey companies in the country are in that range of respondents in making their survey and most of their survey results as proven in the past have only less than 5% error baseed in the past elections. if you don't know how the survey works, then i understand why you made a comment like that.

So can you explain how this survey works and give us some mathematical analysis on how this statistics is supposed to be accurate?

Unless you can tell us the mathematics on how this survey works, then I understand that you are just someone who will readily believe in any kind of propaganda you read in the newspapers.

[mazdamazda]

all reputable survey companies in the country are in that range of respondents in making their survey and most of their survey results as proven in the past have only less than 5% error baseed in the past elections. if you don't know how the survey works, then i understand why you made a comment like that.

So in that case, you will also believe the SWS Presidential Election Exit Poll stating the GMA won fair & square over FPJ (by a margin of around 1M votes)?

51% wants a people power? If that is the case - where the heck are those people? Nilalangaw ang People Power monument.

[squala]

79% may not be an accurate figure, but for sure, more than half want the President out. It's a sad thing the plenary voting depended on the congressmen's personal choice and not their constituents'.

[DoctorDoom]

It's a sad thing the plenary voting depended on the congressmen's personal choice and not their constituents'.

Simply because the CONSTITUTION has mandated how the voting should be done. The CONSTITUTION did not say that a REFERENDUM is needed to impeach the president.

[cyberdoc95]

This is just something I don't get on Philippine Surveys. I'm not quite sure if I would beleive these surveys. Kung ang election returns nga sa Pilipinas ay wala takot na dinodoctor maski na they can be criminally charged for doing such, how much more on surveys na ganito.

How reliable would the survey be and be reflective of the true sentiments of the entire filipino population? And still, tuloy pa rin sila ng karereport. I don't see anything bad about it but will it really make a difference when we all know how these surveys are done.

While I beleive that surveys should be done, something in this survey is just inherently flawed. 79% is a big number,by the way, and if the President sees that only 21% is actually in favor of her staying in power, then she should leave at once. BUT that doesn't seem like happening. WHY? Well for one, the Palace does not beleive in these surveys and clearly for the same rationale as I have.

SO, all I can say is: SO WHAT if 79% want GMA impeached? Obviously, not even the House of REP is convinced that she should be!

[jvm13]

79% is not precise, but at least it gives us a picture of what most filipinos feel.

[DoctorDoom]

79% is not precise, but at least it gives us a picture of what most filipinos feel.

Statistics is one of the most abused fields in science. People always love to play with numbers to convince other people. Anyone who is clueless on how statistics is actually performed will readily believe any kind of number or percentage - sink and sinker.

[empy]

So can you explain how this survey works and give us some mathematical analysis on how this statistics is supposed to be accurate?

Unless you can tell us the mathematics on how this survey works, then I understand that you are just someone who will readily believe in any kind of propaganda you read in the newspapers.

Statistics is one of the most abused fields in science. People always love to play with numbers to convince other people. Anyone who is clueless on how statistics is actually performed will readily believe any kind of number or percentage - sink and sinker.

:bwahaha: these are some pretty condescending statements...do YOU understand how public polling surveys work? or are you trying to say that SWS is dishonest and is working off their own agenda?

[boydapa]

it uses random sampling ksi kaya from the point of view of mathematics valid sya... therefore it represents the whole population

[pissword]

hahhaha.. nung sa thesis namin, to get yung ratio ng respondents or tinatawag na sample size..by the way human resource management po yun course ko.. 1 way of getting the sample size is by using Slovin's formula which is

n = N/(1+N(esquared))

where in:

n = the sample size
N = the population size (85M)
e = margin of error (either 0.01 or 0.05)

pag margin of error is 0.05 300+ pax lang nga ang kelangan para sa buong bansa eh hehe pero if 0.01 at least 10K pax ang kelangan..

now ang concern ko lang naman po is i dont think pwede nilang gawin ito sa buong pilipinas, cluster sampling ata dapat, parang per region masmaganda hindi yung 1200 representing buong bansa na...

[empy]

it may sound strange to some of you, but 1200 surveys can be a representative sample of a population of millions as long as the sample is completely randomly generated and free of bias. as long as you are perfectly aware of the margin of error (typically +/- 5% as stated above), you can draw perfectly valid conclusions.

trust me, i do this for a living

it's a worldwide industry standard to use a few thousand samples to poll an entire population...look at Gallup polls as an example, as well as all the polls conducted by other news agencies such as CNN, USA Today, etc. who all follow the Gallup methodology.

the one thing that concerns me about philippine polls is that survey samples may become skewed because your population has differing levels of access to mass media, the internet, and telephones...and as such, you may have some demographics that are not fully represented. but any statistician who knows what she is doing can correct for this.

[DoctorDoom]

it may sound strange to some of you, but 1200 surveys can be a representative sample of a population of millions as long as the sample is completely randomly generated and free of bias. as long as you are perfectly aware of the margin of error (typically +/- 5% as stated above), you can draw perfectly valid conclusions.

trust me, i do this for a living

it's a worldwide industry standard to use a few thousand samples to poll an entire population...look at Gallup polls as an example, as well as all the polls conducted by other news agencies such as CNN, USA Today, etc. who all follow the Gallup methodology.

I'm an engineering consultant and I do statistics for a living too.

If you are a statistician yourself, you probably understand that "random sampling" is just an English term. You must disclose the level of randomness and your sampling scheme to reflect how random is your "random". Without this information, the margin of error becomes irrelevant to the conclusion.

My point is, in order for a statistical report to be believable, you must disclose the details on how the statistics was conducted and not rely on mere reputation. Atleast you should also mention that the results were audited by a third-party auditing firm (if you are open to be challenged by your peers and not hide from the vagueness of english terms).

When statistics finds its way into a news report, the reader is expected to take everything sink and sinker. No questions asked. Especially if the result of the survey is aligned to your opinion, you are naturally bound to defend the results.

This news report simply lacks in the details that could help it become more believable. Again, we are simply expected to rely on their so-called reputation and expect to believe whatever garbage statistics they tell us.

[empy]

I'm an engineering consultant and I do statistics for a living too.

If you are a statistician yourself, you probably understand that "random sampling" is just an English term. You must disclose the level of randomness and your sampling scheme to reflect how random is your "random". Without this information, the margin of error becomes irrelevant to the conclusion.

My point is, in order for a statistical report to be believable, you must disclose the details on how the statistics was conducted and not rely on mere reputation. Atleast you should also mention that the results were audited by a third-party auditing firm (if you are open to be challenged by your peers and not hide from the vagueness of english terms).

When statistics finds its way into a news report, the reader is expected to take everything sink and sinker. No questions asked. Especially if the result of the survey is aligned to your opinion, you are naturally bound to defend the results.

This news report simply lacks in the details that could help it become more believable. Again, we are simply expected to rely on their so-called reputation and expect to believe whatever garbage statistics they tell us.

uh, ok. any reputable survey has an open-book policy on its methodology and sampling schema. anyone can audit these polls to his heart's content.

it seems the problem here is that you don't think this company has any kind of credibility and its methods are suspect. i don't know anything about the company either - do you have any reason to believe that they haven't adhered to industry-standard survey methodology, or that they have some kind of hidden agenda?

[IceColdBeer]

This is the same firm showing GMA leading and winning the election and also used by the administration to tell the people that they didn't cheat as what the survey shows.

Now that its survey shows negative result eh lumalabas hindi sila credible hehehe.

[DoctorDoom]

uh, ok. any reputable survey has an open-book policy on its methodology and sampling schema. anyone can audit these polls to his heart's content.

You got it exactly right. It's a standard practice that we "presume" these things. This is the reason why quality-minded employers perform actual background checks on the references you write down on your resume, because most infos written on a resume are "spiced-up" or downright dubious. Some people are bold about their claims, hoping you will not take the time and effort to verify it. If someone actually did verify, then they will back out from their claim. If you read various opinions in the newspapers, many have actually doubted these surveys, they just don't want to bother challenging them maybe because it's just a waste of their time.

it seems the problem here is that you don't think this company has any kind of credibility and its methods are suspect. i don't know anything about the company either - do you have any reason to believe that they haven't adhered to industry-standard survey methodology, or that they have some kind of hidden agenda?

I don't want to speculate. My point is that statistics and mathematics in general are very easy to abuse, because most people don't want to be bothered by the "nerdy" stuffs on how you came up with the figures.

There is a good proverb which illustrates this point:
" Tell a man there are a billion stars in the galaxy, and he will believe you;
Tell a man that a bench has wet paint, and he will have to touch it just to make sure ".

Moral of the proverb: If a man is incapable of challenging your claims, he will most likely believe you.

[pissword]

yes sir M54 powered yun lang din ang concern ko kung equal ba ang pag gather ng samples sa lahat ng region ng Philippines.. as i stated above, kapag 5% ang margin of error kahit as low as 300++ samples lang pwede na, mataas na ang 1200PAX na ginamit nila hehe dati d ko din siya maintindihan