Fake Pollster Chronicles: HEY! This one's on our side! (UPDATED)

0


So we know readers come to Caracas Chronicles looking for the finest in Fake Pollster news so, today, we have a special treat: apparently some guys on our side are getting in on this game!

Meet the (either self-parodically or incredibly-cluelessly named) Instituto Delphos, who apparently think calling up 100 people on the phone and asking them what they thought of Monday night’s debate constitutes a proper poll.

Turns out 16.4% of their sample thought HCR won the debate…and if you’re wondering how a poll with a sample size = 100 can have fractions of a percent in their responses, you’re already smarter than the people these jokers are trying to bamboozle.

Once more, there’s no Google trail suggesting Instituto Delphos is a thing. They don’t even meet the abysmally low International Consulting Services standard of providing a warm body journos can quote by name. Still and all, Globovision reports their “results” utterly acritically, without even a throw-away line about how this polling institute is new in the field.

Say it with me now…por eso es que estamos como estamos…

Update: Joke’s on me – turns out Delphos is the unserious arm of a serious polling company. I still think this poll is a methodological wreck, though. It’s not just that with a sample-size = 100 you get a 9.8 point margin of error on a poll where the gap between the candidates is in the 2-4 point range, it’s that asking people to self-report whether they watched the debate and expecting them to tell truth is the kind of thing that would (and should) get you an F in an Intro to Research Methodology course anywhere…

1 COMMENT

  1. This isn’t about fact checking this is about having the minimum of common sense to think “hey I’m a political reporter…is it weird that I’ve never heard of this firm before?” I mean, come on people!!

  2. Delphos is a legitimate pollster. Run by Felix Seijas Jr. (Felix Seijas Sr. runs IVAD). Sometimes pollsters dont have the money or the client, and yet they still have the interest and they run small polls, with small sample sizes and questionnaires.

    100 is better than 0. And I am guessing the fraction of the % is due to a filter question that reduced the sample to a number that when broken into a percentage produced decimals. As in: “¿usted vio el debate?”.

    Tomalo con calma Quico…

    • Chamo, I’ll update the post but all I can say is that if the relationship between Seijas and his kid appears to be a lot like that between Ravell and his kid – la segunda generación salió jodedora…

    • I happened to meet Felix Seijas Jr several years ago at OCEI, now INE. He was smart, fresh from a graduate degree in statistics or demography in the UK.

      The person I met does not look like making that kind of mistake, not even as a start up first job.

      Is that a case of “tirate esa encuestica ahí y asi arrancas, vas haciendo el nombre y despues te damos algo mejor”?

  3. I stand corrected. The 16,4 comes from an average of all the responses people gave when asked something along the lines of “en una escala del 1 al 20 como evaluaria a ….”

  4. I’m with the Globo=FAIL theory, kliq’s theory notwithstanding. Here’s my evidence:

    “50% conocía sobre la realización del debate”
    “el 60% de las personas que sabían acerca de la realización del debate”
    Consistency, please?
    Not to mention “El 70,6% de los encuestados piensa que debe haber otro debate.” kliq’s idea doesn’t explain that one. Delphos could be legit, but this anonymous reporter has no clue, and the editor should be ashamed.

    • No, sorry, a phone survey with N=100 and asking people to self-report whether or not they watched the debate and then what they thought of it more or less defines the firm as in-no-way-legit. It’s absurd!

      • Or, rather, reporting it as a real poll is absurd. I run an absurd little poll on this site too, just for fun. Delphos is running something that’s closer to my just-for-fun little poll than to real public opinion research. It’s reporting it AS public opinion research that’s tantamount to fraud.

        • I said Delphos could be legit. I didn’t say the poll itself was anything resembling a good idea. I agree with what you are saying, but Globo’s report still seems like the worse offense here.

    • 1. Re-read that sentences. “60% del 50%” means 30% of the total sample saw it on tv.

      2. 70,6% of the total sample (100) is indeed incorrect. After all, there is no such thing as 0,6 of a respondent. If I had to bet, I would put the blame on an incorrect reading of the survey report by the journalist. It happens all too often.

      I don´t question that a phone survey of 100 people is very imprecise, and far from representative of the venezuelan population. Still the report does not make any assumptions of the sample being representative of the venezuelan population.

      Its a simple and humble phone survey. With a small sample size. Más nada.

      And yet it does serve some analytical purpose. Given what we know of the venezuelan population, its class breakdown, penetracion de telefonos fijos, 100 people surveyed by phone is objectively better data than no data at all.

      • kliq, adage: what’s worse than no data? bad data. There are statistical tests to determine if 100 out of a population in the millions is worse than no data, and I’d bet 100 is below the minimum sample size for a questionnaire of this nature.

      • OK, I’ll take the blame on reading comprehension (and hide behind the fact it’s my second language) – good point, kliq. But other errors are still there.

        extorres, I’m not willing to blame Delphos for the “bad data,” since for all we know, without this “intrepid” Globo reporter we might never have heard about this. If they never intended to make this public, and were using it for some internal purpose, this is all about bad reporting, and not bad polling.

        • I don’t argue that the reporting may have been a worse offense to professionalism, but I would bet the data is “bad”, even for iternal purposes. Extreme example: suppose you had a bag of 1000 pills, 900 cyanide and 100 vitamin C. You pull one out without looking and swallow it, then conclude the rest of the pills are safe for human consumption. Bad data like this, for any purpose, is worse than no data, which would have kept you doubting at a more appropriate level of caution.

  5. Come on, klig…legit my foot. It could have been Seijas senior himself, it’s crap.
    Is that a pollster? Anyone can be a pollster then. You can call one hundred people if you take half a day off.

  6. Ok this has almost nothing to do with the topic at hand but WTF is Rafael Poelo thinking by writing this http://www.noticias24.com/venezuela/noticia/27645/advierte-que-pablo-es-el-candidato-a-quien-chavez-teme-y-capriles-el-papaya-al-que-prefiere-enfrentar/

    Please read it and tell me what you think, but I can give you a brief summary. Basically Poleo is saying that the candidate that Chavez really fears is Pablo Pérez and that’s why the TSJ ruled the way it did on the COPEI issue and that the new COPEI leadership (Luis Ignacio Planas) is “chavista”. So TSJ ruled like that on COPEI so COPEI ends up endorsing Capriles which is supposedly what Chavez really wants. And that Chávez inhabilitó a Leopoldo López instead of Capriles because he doesn’t fear Capriles but he fears Leopoldo.
    I know I shouldn’t expect much of Poleo…. but… really? REALLY? Why?
    Por eso es que estamos como estamos.

    • Planas has already stated that they’ll keep the commitment to Pérez. As for the rest, you shouldn’t expect anything at all from Poleo in terms of real information or analysis. He is sometimes amusing, I’ll give him that, but his idea of journalism is pushing his agenda, and if the truth interferes, so much for the truth.

      • I believe nobody should take anything our pundits say at face value. That goes for journalists and bloggers. Nonetheless, I can fathom what could Poleo’s agenda be. If his goal is to be persecuted by each and every government, he’s doing a wonderful job.
        Regarding Poleo’s article, if chavismo underestimates HRC and it’s concerned about PP’s popular appeal, HRC’s supporters should be celebrating. HRC will be partially protected by a maracucho lightning-rod and may as well escape the thoughest part of the chavista mud-sliding campaign. Isn’t that a good thing?

  7. Thanks . My candidate is Capriles since April 2010. But I woudn’t care if Pablo Perez was the candidato de la U except that I really don’t think a candidate endorsed by the fantabuloso combo of the guanábana + Poleo could have very strong arguments against Chavez.

    • Venegringa… ALL candidates, including Capriles, will have AD+Copei backing come Feb. 13th.

      I know most people would like for AD or Copei to simply disappear. We have the power to do that: le’ts vote for Hugo Chávez in 2012.

      • You are totally right. I was just really shocked when COPEI endorsed Pablo. Because I never thought of him as totally Jurassic, and I think he isn’t … I think it says a lot about him and about the pathetic state in which COPEI is, that they endorsed Pablo Perez, AD’s candidate.
        The other thing is that as we all know, there was widespread rejection of AD and COPEI for at least the 1990s and early 2000s… that’s why our party system collapsed. IF Pablo gets stronger in the polls after their endorsement by these two parties, could be attribute it to the fact that at least in the case of AD, maybe they are not as rejected by the country as they were in the 1990s, in the early 2000s? Maybe they have recovered their image, to some extent?

      • That’s a misleading analogy, GTA. The fact is that AD and Copei, when faced with a choice among several opposition candidates, went for a particular one. After February 13th, they will have no choice but to endorse whoever wins (or maintain their endorsement, in case PP wins).

        You imply that the early endorsement makes no difference because, in the end, whoever wins will be endorsed by AD and Copei. A lot of us think that the early endorsement is an indication of who among the opposition candidates best reflects the AD-Copei way of doing politics and is more poised to advance their interests.

        • I agree, PP seems to reflect a safer choice for AD and Copei (safer for Copei, which does not have to compete with UNT anywhere, unlike AD). But in the end, PP will need PJ and HCR will need AD+Copei.

          As for their interests, they share common goals, that has been drafted throught the Unity platform and the agreements signed by all the potential candidates on Sept. 26th.

        • It’s a simplistic and very biased way of explaining why PP got AD/Copei’s endorsement. Should I infer that HRC got PPT/Podemos endorsement because they think that HRC’s way of doing politics is closer to what they were used to in the not so distant past, i.e. Chávez style? Yes, AD/Copei might have skeletons in their closets, but the other guys – including Borges/HRC’s PJ – are not angels either.
          Yes, there’s probably a divide between old parties (AD/Copei) and the newcomers (Podemos/PJ) because ALL of them want a piece of the cake. It’s the real world not a fairy tale. Old guys don’t like newcomers and viceversa, so it shouldn’t be a surprise when things get messy: newcomers are unwilling to play well with the old guys (Maracaibo, Chacao and Bolívar in 2008) and the old guys won’t give anything for free (Miranda 2012).
          Besides, endorsement is not all about sharing the spoils, it’s also about jumping on the right bandwagon. What’s the point of endorsing a candidate who can’t win it all? We don’t know on what information (polls) AD/Copei based their decision. For you and many other PJ’s fans it is a no brainer that HRC is The Man, but that’s not necessary the case for voters. We can read all the polls we want, but we’ll have to wait until February 12th to find out who’ll be the winner.

  8. At least they are “honest”: they are not pollsters, bur oracles. I wish some other pollsters would finally present themselves as what they think they are: sages.

    I’d hope a University and their Mathematics and Political Science Departments could make polls (it would be good for getting some income in the colleges’s coffers and some real scientific legitmacy out into the public sphere). I’ll have my people call their people… 😉

  9. They clearly call this a survey which is different than a poll. Media often surveys its viewers to determine what they think on all sorts of subjects such as if their female reporters look sexy enough or too much. To be honest this is mostly about selling political advertizing

  10. Here’s another company with some dubious credentials. This consultancy firm analyzed the opposition debate and concluded that Leopoldo Lopez won hands down. http://www.ultimasnoticias.com.ve/noticias/actualidad/politica/evaluan-debate-de-la-mud–cada-uno-se-lucio.aspx
    Who are ODC Consultores? According to his Linkedin profile ODC’s owner Oswaldo Ramirez also worked for the Alcaldia de Chacao between 2001 and 2009.
    No conflict of interest there, then….

    • Great catch!

      Luz Mely Reyes is one of the less stupid political reporters in Venezuela, though – it’s discouraging that she didn’t suss that out…

Leave a Reply