Monday, November 2, 2009

Lead Philippines Survey: Do Internet Users Influence the 2010 Surveys?

This is an e-mail survey being conducted by Lead Philippines to determine the influence of internet users to the 2010 surveys conducted by two most popular polling agencies. The results of this survey will be posted at www.leadphil.blogspot.com. The survey question is as follows:

Were you ever interviewed by or a respondent to either Pulse Asia or Social Weather Station (SWS) in any of their 2010 Philippines Presidential Surveys?

E-mail your answer (either YES or NO) to lead.philippines@yahoo.com

Also, to help in this survey, please forward this e-mail to as many internet or e-mail users as possible. Thank you for participating.

LEAD PHILIPPINES

21 comments:

justine said...

Hi, I received your email through Yvotephilippines@gmail.com. the member organizations are asking about the objective of this so-called survey, and how is this going to be used.

maybe you can help me answer this question so that we can forward it to our members.

many thanks,
Justine Castillo
YPS Program Coordinator

SRTC-TD said...

This is a rather stupid question to ask in the internet... Internet surveys are not representative ... The SWS and Pulse Asia surveys are conducted among 1,500 to 2,000 respondents and designed to scientifically represent the national pulse... Do you honestly expect to get a lot of these respondents in the internet? (and more so, among those who are within your reach)... Whatever you gather from your survey is likely to be having coverage biases... and what do you intend to show ... that the SWS and Pulse Asia surveys are not representative? (and do you think internet surveys are)...

Arnel B. Endrinal said...

Hi Justine, thank you for your question/comment. The purpose of this survey is (as stated) to determine the influence of internet users, if any, to the results of the surveys made by PulseAsia and SWS.

Of course, since this is not a scientific survey, it is doubtful whether we can come up with any conclusion. Nevertheless, we will post the data gathered and it is for everyone to determine whether there is any value to them. Thanks.

Arnel B. Endrinal said...

Hi SRTC-TD, Thanks for visiting this site.

If you are saying this survey is not scientific, then we are in agreement.

If you are saying that we just intend to show that SWS and Pulse Asia are not representative, we are not sure how you came to that conclusion.

If you just want to say this is stupid, then that is your call. It would have been better though if you tell us who you are so that you can at least fairly say you are open to the same kind of scrutiny. Thanks.

sjsanjuan said...

I do not understand the intent and objective of this survey. You said that it intends to determine the influence of internet users, if any, to the results of the surveys made by PulseAsia and SWS.

Several points --

1. Do you know how many respondents in the PA and SWS surveys are internet users?

2. What is your methodology in parsing the data? What is your framework in appreciating the information to be collated?

3. Who is behind this survey? Is this a commissioned online survey?

4. You said that you do not expect to come up with a conclusion, considering the unscientific nature of this survey. So what gives? Why the survey in the first place? Is this a trending/conditioning tactic?

Arnel B. Endrinal said...

Hi sjsanjuan,

Thanks for your comments.

As mentioned, this survey is not scientific. Obviously, those who will respond to this survey are also the same people who share the same interest in the topic. Perhaps some know Lead Philippines enough to warrant a response.

There are no methodologies to be employed here nor do we intend to produce or claim a scientific analysis out of the data gathered.

This is not a commissioned survey, but only an informal survey by Lead Philippines to as far as where it could reach via the internet. And no, we do not know if there are internet users who were respondents to PA and SWS surveys.

Are we trying to condition minds? It is your call. But you have to find out the result first and learn more about Lead Philippines before you can make that call.

Meanwhile, I am sure everyone realizes that even scientific surveys may be used to condition the minds of the people. I just hope that everyone is as fervent as you are in looking/analyzing at the results of various surveys, scientific or not.

Jose Ramon Albert said...

My apologies if it seems a default name SRTC-TD (the Training Division egroup site I created was made the name) was attached to a previous post... My name is Jose Ramon Albert. I earned a Ph.D. in Statistics way back in 1993.

My apologies also if I described the question as "stupid"... as I simply can not find any more appropriate word for this kind of question, nor the results of whatever data collection that you may generate...

In any of the formal and non formal education activities I give on Research Methods, and Statistics, I cite the importance of identifying a research question, goal, as the starting point of a research undertaking... And the importance of having a sound methodology to be able to answer the research question.. Otherwise, garbage in, garbage out.

Internet surveys like this will simply not get anywhere... Even if you manage to get any results, it is pointless to analyze them... because of "coverage errors", i.e. biases in the respondents, and other non-sampling errors (arising from the question wording)...

So I would strongly advise that you discontinue this activity, as it is simply a useless enterprise... Even if you manage to get 10,000 people to answer your question... The results will simply be totally meaningless...

If you continue this activity, then you make yourself and your organization the subject of
scientific scrutiny that will only destroy your
reputation for perpetuating something which
will likely be misused and abused for possibly
political ends.

Yours in the Practice of Sound Statistics,

Jose Ramon G. Albert, Ph.D.

Arnel B. Endrinal said...

Hi Mr. Jose Ramon Albert, Ph.D,

Your apologies are accepted. I find it surprising though that a Ph.D did not afford you to find a word more appropriate than stupid.

We now see where you are coming from, especially now with your tamer language.

No, we are not doing a research so we are not using any "research methodologies". If we are, then we would have hired the services of a statistician. Indeed, we plan to post in this blog the answer to the question we sent out. It would be up to the reader to decide whether the data is garbage.

It is OK if you would like to subject this "unscientific" survey to your "scientific" scrutiny. We just wonder what kind of methodology such scrutiny will use. Also, we think you gave too much credit to this "unscientific" survey for it to be used by anyone for their political gain. We think politicians are using too much already of "scientific" surveys for political gain to even take note of our small survey.

Meanwhile, it is obvious you have a disdain for unscientific surveys, and thus would like to discourage any participation. Have you tried telling that to mainstream mass media organizations who everyday make use of SMS surveys and their results as if they are gospel truth? I suggest you do that. It will keep you busy for a while.

Yours in the practice of Sound Judgement,

Jose Ramon Albert said...

For your info, we have attempted to advise media to put a disclaimer about their use of unscientific surveys... There is nothing wrong with coming up with descriptive statistics on non-probability samples... What is absolutely is when people are going to infer that the results of non-probability samples can be extrapolated to a population. They can not... Is there information from non-probability samples, yes, absolutely... In some cases, there are simply no ways to come up with probability surveys...

I don not have disdain for non probability samples, but for misuse of them... and that is why I caution you that your statistical activity will only give you and your organization a bad light... If you wish to proceed, then go ahead.
But be warned, garbage in, garbage out.

Any research, and data collection activity, presupposes some clear objective. I did not find any in yours.

Cheers!

better said...

srtc-td,

what's the fuss over this survey anyway. it's an informal survey. heck, i can start one of my own.

i am particularly interested in knowing the results. i have been wondering why despite my being in existence for close to 38 years now i still haven't come across anyone who has ever been interviewed by sws, pulse asia, or any other survey firm. nor have i met anyone who knows anyone who has been interviewed.

if you don't like the survey don't participate. it's that simple.

may i ask are you in any way connected to pulse asia and/or sws?

Arnel B. Endrinal said...

Hi Jose Ramon Albert,

Thanks for your comments again and cautionary warning.

You said you attempted to advise media? Well try to do it some more, as they are still at it everyday.

Meanwhile I am sure you have noted that even "scientific" surveys are used for trending and/or mind conditioning purposes. Maybe it is not really on the survey, but on the people using them?

Also, I am sure readers of this blog are sophisticated enough to know what is garbage or what is not and can decide for themselves. Or they won't be here in the first place..

Hi Better,

Thanks for your comments. Now we know your age!

better said...

arnel,

yeah... that's give or take 3 to 4 years. i'm actually 25 in dog years.

Jose Ramon Albert said...

BETTER:

THE "fuss" is about having proper information in the public domain. Not everyone is trained enough to understand what statistics mean...
Even teachers of college statistics courses are not quite teaching Statistics concepts properly…. Many textbooks on college statistics have been found by a study of the Philippine Statistical Association to be erroneous… So even if people are “sophisticated” enough to learn the use of the internet, that does not necessarily mean people understand Statistics… There is a whole Science to collecting and analyzing data.

The title of the topic suggests that the results of the survey will be descriptive of the Influence of INTERNET USERS in the Phiilippines in 2010 surveys… The use of informal internet polls will not make the results representative.. The only way to ensure representativeness of such a target population is to use chance methods… But there is “self selection” bias here, as the only ones who will participate are those who will be industrious enough like yourself to go though the survey.

And to answer your question about affiliation, I work for a policy think tank of the Philippine government. I am not affiliated with Pulse Asia, although the Fellows of SWS elected me into their fold (but this is more an honorific title and I am not required to defend the results of any activities of SWS… That is the responsibility of the officers of SWS). My own research interest is largely around poverty statistics generated by the National Statistics Offfice household surveys, and education statistics generated by administrative reporting systems of the Department of Education. My reaction here is purely out of a concern for the possible misuse of the results here. Too many “informal” surveys are floating around… We should understand there is a science to the collection and analysis of data. Having earned a degree in Statistics, I feel it my duty to request this survey (and many surveys in media to be discontinued, or at the very least, provide a disclaimer about the survey not being represenative, in the public interest).


BETTER:

THE "fuss" is about having proper information in the public domain. Not everyone is trained enough to understand what statistics mean...
Even if you have gone through a college course in Statistics, it is very likely that
you may have misunderstood what surveys
mean... I have been involved even in "training"
teachers of statistics... many of whom are not quite teaching statistical concepts properly..
and this is why we have been giving training
courses to statistics teachers themselves...
(in order to prevent the blind from leading
the blind).

Cheers

better said...

jose ramon albert,

no big deal. i'm still wondering though where all those survey respondents are, really?

in addition, why is it that whenever politician x commissions a survey the results always ends up favoring him. and if politician y commissions his own he also gets results favoring him. the same goes for politician z. is there science here or is this just simply a case of giving the one who's paying for the service the results he wants?

don't get me wrong i have nothing against statistics and i don't doubt that it is a scientific field. i do however question the "politicial" think tanks and survey firms that use statistics in rather strange ways precisely because of the observation i stated above.

i'm just a layman and in my view, which i back with common sense, (i'm just saying because it might not be apparent) arnel's survey is straightforward and brilliantly simple.

"Were you ever interviewed by or a respondent to either Pulse Asia or Social Weather Station (SWS) in any of their 2010 Philippines Presidential Surveys?"

i, for one, would really like to know the answer to this. please don't deny me that by asking arnel to discontinue his survey.

now correct me if i'm wrong. i suspect your reaction to this survey is one that is rooted on a desire to protect the "integrity" of the surveys conducted by sws and maybe even the other survey firms.

finally, would you oblige me by telling us how arnel could improve (not discontinue) this particular survey centered on the question:
Were you ever interviewed by or a respondent to either Pulse Asia or Social Weather Station (SWS) in any of their 2010 Philippines Presidential Surveys?

again, as i said, this is really important to me.

btw, SJSANJUAN, you may also show us how to do this properly if you want.

better said...

"now correct me if i'm wrong. i suspect your reaction to this survey is one that is rooted on a desire to protect the "integrity" of the surveys conducted by sws and maybe even the other survey firms."

correction: replace "integrity" with "credibility"

Ben Kritz said...

Dr. Albert, I can appreciate your academic concerns and I agree that it is important to view the results of this as well as any other survey in the proper context. Nonetheless, opinion surveys are always subjective no matter how competent the data analysis, and so even the ones produced by 'professional' organizations must be taken with a very large grain of salt. Unfortunately, when the subject is an election campaign, and especially when the campaigners are providing the potential survey respondents little in the way of concrete ideas upon which to form quality opinions, taking it with a very large grain of salt is the best we can do.

Thus, I find Arnel's exercise to be no more or less valid than those of Pulse Asia or SWS. As he said, it is for the reader to decide what value to attach to them in making up his own mind. And I'm certain you'll agree that even the "professional" surveys are not immune to "garbage", such as the recent one which asked respondents to pick three candidates -- a stupid (if I may borrow your word) question lacking a clear objective, since voters in the upcoming election will obviously only be choosing one candidate.

So I'd cut Arnel some slack, if I were you. The 'clear objective' of Internet surveys is to encourage discussion, not to display our statistical prowess, and judging from the number of extended comments this post has already generated, I'd say Arnel has done a hell of a job.

Yours in Friendly Discourse,

Benjamin D. Kritz, M.Sc., MBA

AsiaWest said...

A disclaimer should be enough. Suppressing free inquiry harks back to the time of the Inquisition.

Jose Ramon Albert said...

One does not need to conduct a survey to be able to judge another survey... (especially if the "informal" survey will itself have "selection biases")... The SWS and Pulse Asia are just two of many institutions doing surveys... Ibon comes out with its own set of surveys. And so do so many fly-by-night institution. If we are to judge surveys, then we should inspect their methodologies.

BTW, for people's info, a few years back, SWS and Pulse Asia were not the ones doing the survey operations... They were subcontracting another institution: TNS Trends (which is the competitor of another well known market research group--- AC Nielsen).

Pulse Asia and SWS, as far as I know, conduct some surveys on a regular basis --- with some people, institutions funding these surveys. These "sponsors" get their own briefings... Occasionally, some of their surveys are also commissioned by some people, and it is the policy of these institutions not to release the results of these commissioned surveys unless the client agrees to having these released to the public.... I know of some surveys, commissioned by some senator, which showed the senator was not popular... The results of the survey were then used to identify target groups which he was weak with... and in the subsequent election, the senator came out within the winning circle.

I am not a defender of these institutions... But of proper sound statistical survey processes. I do not know well enough the methods of SWS nor Pulse Asia in choosing respondents... All I know is that the people behind them are people of academic repute... I challenge not
just SWS and Pulse Asia, but all those conducting surveys to DISCLOSE THEIR PROCEDURES for selecting respondents; and to subject themselves to a scrutiny by independent reviewers. That would be the best way to determine which among these institutions is conducting their surveys well.

That is what I think would be the best way of determining the soundness of practices... not by having an informal survey to find out if respondents say YES or NO to "Were you ever interviewed by or a respondent to either Pulse Asia or Social Weather Station (SWS) in any of their 2010 Philippines Presidential Surveys?"

Jose Ramon Albert said...

A DISCLAIMER for ANY INFORMAL SURVEY is proper...

But, the question still remains, what use will there be for the results of this survey, even if they may be "interesting" to some of those giving comments here...

If our task is to determine whether these survey groups are doing their job well, we don't necessarily have to do a survey ... (which itself is riddled with problems on coverage, and even on the question itself)... There is, to my belief, no way to improve this survey... and neither is there any meaning to it...

What we ought to do is ASK all institutions that conduct surveys, SWS, Pulse Asia, and other private groups, and even public institutions like NSO, the BAS, the BLES, the BSP, among others to subject themselves to scrutiny by independent reviewers... Only then can we establish the underlying answer to the question that seems to be the at the bottom of this endeavour ---" are the results of surveys being conducted accurate, precise, and meaningful?"

better said...

mr. albert,

but the fact remains this is just an informal survey. it is hardly something to be bugged about.

i do agree though with your assertion that all these survey firms should be subjected to scrutiny by independent reviewers.

Arnel B. Endrinal said...

Hi Everyone,

I am sure SWS has a reason for conducting a survey asking respondents for three presidential candidates instead of just one. As we learned from Dr. Albert, there should be a goal and methodology to achieve such goal in every survey.

I am sure Dr. Albert, will, with his connections at SWS, and with his desire to caution us on the dangers of surveys (I hope not just on informal surveys, by the way) will equally provide us illumination on what that last survey and similar SWS surveys in the past are all about as to their most probable purpose/goals.

Will you, please, Dr. Albert?