Jump to content

Talk:Opinion polling for the 2015 Spanish general election/Archive 1

Page contents not supported in other languages.
From Wikipedia, the free encyclopedia
Archive 1

About candidacy votes

Some surveys' data shown in the table may not reflect the exact data shown in that survey main page. As a , I just beg you to refrain from keeping changing those: it is intended for them to show like that. The answer is simple: those surveys' vote percentages are calculated over candidacy votes, and not over actual valid votes. Candidacy votes are all those votes won just by parties, while valid votes are the result of adding blank ballots to candidacy voting' count. Let me explain with a visual example:

2011 results:

PP = 10,866,566
Candidacy votes = 24,015,425
Blank ballots = 333,461
Valid votes = 24,348,886 (= 24,015,425 + 333,461)

% of PP votes over Candidacy votes = ( 10,866,566 / 24,015,425 ) * 100 = 45.25%
% of PP votes over Valid votes = ( 10,866,566 / 24,348,886 ) * 100 = 44.63%

As it can be seen in the example, the difference between both systems of calculating % is rather substantial: 0.6 points in this case. Now, let explain this using the data for the latest Celeste-Tel survey:

PP = 4,590,000
Candidacy votes = 15,123,000
Blank ballots = 379,000
Valid votes = 15,502,000 (= 15,123,000 + 379,000)

% of PP votes over Candidacy votes = ( 4,590,000 / 15,123,000 ) * 100 = 30.35%
% of PP votes over Valid votes = ( 4,590,000 / 15,502,000 ) * 100 = 29.61%

This is why the 29.6% data is used over the 30.4% one. The first one is the one calculated including blank ballots, and thus, the one comparable with all other surveys and with official election results. The latter, however, isn't comparable. The change is, thus, intended.

Official election results are calculated over valid votes instead of candidacy votes, and thus, surveys using data calculated over CV rather than over VV have been recalculated so they resemble values much more comparable with the official election results. This is already explained in several places of the article, yet there is still people ignoring those notes who keep changing the data back to the original values without explanation. So, please, if the issue is to be discussed, let's discuss it here. But stop changing the data without notice. The "Notes" section, and the talk page by the way, are there for something. Thank you. Impru20 (talk) 21:22, 9 April 2013 (UTC)

European election polls

European election polls are mixed together with general election polls. Although extrapolation can be made out of these polls, these are not strictly applicable to general elections.

Regards — Preceding unsigned comment added by 46.25.35.92 (talk) 21:23, 8 June 2014 (UTC)

The only European polls that had been included are two Sigma-2 polls which the own pollster specifically compares with other polls it made for the general election. That is why they're included, as it uses them indiscriminately for both types of elections (Sigma-2/El Mundo has done this in the past, as well). I believe there is no other European poll included here. Cheers. Impru20 (talk) 21:31, 8 June 2014 (UTC)

Reliability of some surveys

I think surveys such as the one from http://todaslasencuestas.blogspot.com.es/ should definitely not be considered as a reliable source and included in the article. It is claimed to be done by some sociology students as an "experiment" and a "hobby", with a sample of only 500 people, and the methodology is not explained at all. Anyone can open a blog claiming to have done a survey, that doesn't mean it should be added. I won't edit the article as I am not experienced Wikipedia editor, but I think it should be deleted.

I should agree with this. I added it because it was referenced to in Electomania's website, but looks like it has not been referenced anywhere else and maybe has a dubious reliability. Proceeding to remove it now. Impru20 (talk) 23:27, 23 November 2014 (UTC)

Impru20, Termómetro Electoral is not a poll or a polling agency, it's just a sociology blog. They got the data from readers who answered on the blog and based on previous polls. We don't know the number, the method or anything. I think it should be removed. TeddyBear01 (talk) 20:23, 3 February 2015 (UTC)

From here it says it's an online poll, made between 23 and 28 January, asking 554 people using an online methodology (seems that the poll's data has been published in several days, that seemingly being the reason of the technical sheet not being present in the vote estimation page, since it was published previously).
It actually seems to be a poll (the sociology blog even going to say that "it is the first time" they use their online polls to get a vote estimation), and offers data on other questions as well as leader ratings, which they seem to have done for some time. The same blog published vote projections that, indeed, were not polls (they were added here but later removed by myself when I noticed that fact), but this one seems to be one. The fact of it being a sociology blog and not an actual pollster is not that relevant, since in Spain it is somewhat frequent for third parties such as blogs or newspapers themselves to make their own polls and vote studies, independently of pollsters. Impru20 (talk) 20:52, 3 February 2015 (UTC)

TeddyBear01 is right Impru20, Termómetro Electoral only use nameless people in Internet for their "Termómetro", but it's a mixed methodology because the "poll" contrast with another real polls (no such the "Termómetro"), include when the different polls use different methodology. Moreover, no body can knew how it's made because they don't publish their real methodology and they don't give any real clue. 150.244.132.78 say you the same, and he is right, Electomanía have a very similar methodology, but the difference it's that Electomanía is public every part of their methodology and without doubts, while Termómetro Electoral no. Electomanía put out a PDF (with 374 people), while Termómetro Electoral no, and Termómetro Electoral use a very lax extrapolation of seats, without forks. Both of them aren't polls, if the same if I do a extrapolation with every poll. You say that they put IDV, but they say: "Intención Directa de Voto de los diferentes sondeos publicados en Enero (PODEMOS - 22,1 / PP - 13,6 / PSOE - 12,9 / IU - 3,9 / UPyD - 3,8 / Ciudadanos - 3,6)", they IDV it isn't a real IDV, is a mixture of IDV's of different polls (two failures). You say that they have a methodology and it's true, but we only can know that it's a mixture of the internet reviews and the different polls (another failures), and no more. 95.122.196.92 (talk) 19:31, 10 February 2015 (UTC)

1. Most pollsters use "nameless" people, so you can't actually use that as an argument (My Word uses a similar online methodology, for example).
2. It does not say anything about it using other polls, but of it using "responses received" and "party evolution" (which is what most pollsters do, actually).
3. Pollsters usually don't publish their real methodlogy, or only publish it partly. If we are to abide by that, most polls would have to be removed, because Spanish legal regulation on opinion polling is practically non-existant outside of electoral periods.
4. You are comparing it to Electomanía, when they are two different things. The "Termómetro Electoral" asks about voting intention ("what would you vote if the election was held today?"). Electomanía asks about "what do you think that the result will be if the next general election would be held today?"). Electomanía is not added here because it is not a true voting intention poll, but rather asking about the respondants' "beliefs". If they asked about vote intention, their addition could be considered, but that's not the case.
5. Again, most pollsters don't publish many of the things you are putting about here (in fact, only Celeste-Tel and some other random poll do publish a PDF).
6. You can't use the "it shows seats without forks" as an argument. GAD3 also shows seats without forks. Sigma-2, when it does publishes seat awardings, usually does it without forks (see 2011 election polls). As does DyS. And Celeste-Tel, while showing forks, does a kinda weird extrapolation of seats (it is technically impossible for a party polling below 5% to get more than 10 seats) but they are published because that's what the source says. Furthermore, most polls do not publish a seat awarding at all. If we are to add polls judging by whether they publish seat extrapolations or not, most of them would be removed. That's not a valid criteria, really.
This said, "Termómetro Electoral" seems a legit poll so far, even asking for other things such as leader ratings and other random questions on Podemos and the such. Maybe in the future, if a decent legal regulation on opinion polls is established in Spain, we could argue about polls' requirements, but as of now we have included polls from:
Opinion pollsters which rarely publish their raw data or methodologies (such as DYM, most Simple Lógica polls, Invymark, some Sigma-2 polls...).
Opinion pollsters which keep constantly changing on their methodologies and data treatment methods (Metroscopia).
Opinion pollsters which are suspected (yet not proven) to not be pollsters at all (Celeste-Tel and NC Report, which have received criticism in that sense in the past).
If we do abide by what you say (which I understand), probably only the CIS polls and few others would be allowed in here, because of the largely unregulated nature of opinion polls in Spain which makes pollsters to often not publish their data. I know this is not an issue in other countries, but in Spain it is like this. In order to allow for most of these opinion polls to be placed in here, we need to have a largely permissive and comprehensive view on the issue. While I'm rather against opinion polls of separate blogs, which need to be more deeply analyzed than others published by media sources (due to their rather dubious veracity in most cases), as long as they do seem legit, do show their technical data and are not clear partisan manipulations, there is nothing preventing them from being there. It is not rare for independent opinion polling political sites to make their own polls (such as Scenari Politici in Italy). But that only should abide, as I said, for "voting intention" polls; Electomanía is not one of these as it asks about a very different thing. Cheers. Impru20 (talk) 19:41, 10 February 2015 (UTC)
EDIT: From how many IP accounts are you talking? I now see that both 150.244.132.78 and 95.122.196.92 seem to originate from the same place in Spain. Though I'm not an expert on this, it is so rare how two different IP users enter here the same day, when almost no one actually edits this page, to delete the same content and using the same arguments... Impru20 (talk) 19:44, 10 February 2015 (UTC)
EDIT 2: Yes, I saw what you have just said about the IDV, and you know: I thought just what you said when I saw it, was something that almost made me to disregard this as an actual poll. But then I actually made the calculations myself and, you know, the "mixed" data of IDVs published in January is not even close to that (i.e. the highest IDV for PP, aside from this poll, and for polls made in January before this one is 13.3%, while this poll shows a IDV of 13.6%. It's mathematically impossible for it to be an average or mixture of polls). It must be probably an error; they have used the plural instead of the singular. A bit rare, but it is not something I have not seen before with other pollsters. Mistakes like this are not uncommon. Impru20 (talk) 19:57, 10 February 2015 (UTC)

Ok I accept that I can be wrong, but I think that it isn't a poll. However, I accept the poll in Wikipedia. About the IP accounts we aren't the same and I don't know where is the other IP, but if it's where I live (where is it exactly same place?), I think it isn't really difficult. Cheers. User:95.122.196.92 (talk) 20:32, 10 February 2015 (UTC).

It is not whether what we think or the opinion we have, but about what sources say. I don't like to be so permissive on opinion polls, because there is a tenuous red line between what can be considered acceptable as an opinion poll or electoral study (the latter being a figure which has become popular between national newspapers, which adds up to the trouble) and what can't. Usually for other countries which have more formal regulations on this issue, there are more restrictions. But in Spain we have reached a point where opinion pollsters are becoming more of a joke than anything else, publishing the data they wish and not the full results, sometimes missing their methodologies' data, or constantly changing their methodologies as they see fit, or considering some random data they publish as an actual poll... and we don't have the elements to guess whether they are lying or saying the truth because of the Spanish legislation only regulating this matter during electoral campaigns. Frankly, since the 2012 Andalusian election, opinion polls are more focused on influencing political opinions than in actually predicting election results, failing in all elections held ever since. But we must work with what we have and with what sources say, whether we like it or not (sadly). In this case, TE shows technical data and shows reasonable results (i.e. not a wild 'poll' saying 80% would vote for a party) and has some very decent backing, so unless proven otherwise in the future, there is no reason now to justify not having this one but having others from official sources which may either have less data available, or have some doubts about their true veracity.
I'm not saying the place for privacy, but from the geolocate function of Wikipedia I get the same coordinates... but as I said, I'm not an expert on this. Understand that it is very rare for two IP addresses to enter the same day on the same page and edit the same section, using the same "Electomanía" argument, as it is a frequent happening in Wikipedia for a person to use different accounts. Anyway, as per WP:GOODFAITH I believe you unless you had proven otherwise, which you didn't. So, cheers. ;) Impru20 (talk) 21:06, 10 February 2015 (UTC)

About tables

Wouldn't be better to show the tables collapsed by default? It would make the page easier to watch. Asturkian (talk) 11:27, 17 April 2015 (UTC)

External links in Body of Article

Hello, Editors! I've started the process of removing External Links from this article to meet WP:EL, but will need some help from you. First, please post sources as citations/references instead of External Links when adding new content.

If you forget, myself or another editor may revert your edit containing ELs. If someone does, please feel free to revert back *and then immediately* change the ELs into cites/references so you don't have to redo any of the rest of your edit.

We can get rid of the EL issues tag at the top of the page once all of the external links are removed (and converted to cites where appropriate). I've started the process and hopefully others can help out with a few here and there until they're all gone. Thanks, Stesmo (talk) 23:55, 2 August 2015 (UTC)

WP:EL can be subject to exceptions such as WP:IGNORE. In the case of opinion polling, external links are much prefered over citations/references due to the amount of links and the simplicity ELs give to readers. It has been also a custom practice for most election opinion polling articles due to this. As WP:EL is not forcibly required, we can make use of the exception so as to avoid creating unneeded issues. Impru20 (talk) 23:59, 2 August 2015 (UTC)

Convergence and Union

We should have to specify that in some polls after the dissolution of CiU (in June 2015), the poll results put under CiU actually are the percentage polled for Democratic Convergence of Catalonia (Convergencia Democrática de Catalunya, CDC), and that was only one of the two parties that composed the defunct federation. Regards. --Sfs90 (talk) 22:02, 18 August 2015 (UTC)

National "Vote Intention" table has disappeared

Someone seems to have removed the "Voting preferences/Vote intention" area from the National polling section, quite recently, I believe. I do not see it in the history. I do not see why anyone would do this, especially as the Catalonia "Voting preferences/Vote intention" area is still there, along with the Contents item for it. Regards, Endelite (talk) 01:24, 23 October 2015 (UTC)

I removed it as part of a process of table updating ad fitting out. The issue was that each pollster uses a very different method for gathering data, so those numbers may not be really comparable between themselves; this, coupled with the fact that many pollsters don't even publish these dat, or publish it in an incomplete way without fully separating the data between abstaining and undecided. I explained the reason when I removed it, so I'm rather surprised that you didn't find it in the history. Also, most other opinion polling articles for other countries don't even show this data, so I wonder whether it would be, maybe, just "too much" info. I was the editor who added the Vote Intention table to the article in the first place, so I removed it until a better way of showing those figures can be figured out. The info is in the History section, so it won't be lost if it needs to be recovered. The Catalonia section has not been touched yet, so that's the reason it has not been removed from it. Impru20 (talk) 09:03, 23 October 2015 (UTC)