Jump to content

Talk:Opinion polling for the 2015 United Kingdom general election/Archives/2014/September

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


Achieving two tables

The RfC above was closed and User:S Marshall put forth a plan for two tables: one focusing on the four larger parties (Con, Lab, LDem, UKIP) and another with a longer list of parties, including the Greens, SNP, PC and BNP. I think this is an eminently sensible way forward. How are we going to make this work? Bondegezou (talk) 10:40, 27 August 2014 (UTC)

I hate to be a stick in the mud but firstly that wouldn't be reflective of the way the polls are published as mentioned above but also if we change it for this article we must change it for this articles preceding articles as well, the status of Plaid and the SNP has not changed. We need to be consistent between articles (that doesn't mean they must be uniformly identical but it does mean the criteria for inclusion should be broadly the same - at present they are). Furthermore there is no more of a reason now then there ever has been before in the last 10 years to add to add the Greens or the BNP to the table, their support bases are broadly the same as they ever have been in the medium to longer term....5% isn't a magic number or threshold. The BNP achieved 5%+ in opinion polls during the last Parliament, should we have considered changing the format then? Any format changes need to be consistent with preceding articles, even if that means making retrospective changes. In the medium to longer term and looking at trends there is nothing to suggest the SNP, Plaid, the Green and the BNP have a stronger case to be added to the polling tables for this parliament then they do the previous 2 or 3 Parliament's Polling tables.
Furthermore in a practical sense I don't understand how you would do this and why would we duplicate information within the same article. I can see why people are eager to show some form of compromise and inclusion (I get that) but it simply would not be a reflection of our reliable sources, the political reality on the ground and it would not be consistent with various Wiki policies (already been outlined above in previous closed discussion - were going round in circles here) Owl In The House (talk) 21:29, 27 August 2014 (UTC)
We had an RfC. You, Owl, made your views clear. The RfC came to a conclusion. I'm uncertain what the value is in re-opening the discussion. I think we need to move on because consensus does not support the status quo. So, how do we move on? One way is to follow the suggestion put forth by User:S Marshall. Bondegezou (talk) 10:30, 28 August 2014 (UTC)
S Marshall T, could you please acquaint yourself with Wikipedia policy on this matter, it is referenced above, consistency is a very important part of Wikipedia policy. I am not against change for the sake of it. Besides I don't really see how the current version doesn't have sufficient support. Its perhaps worth re-looking at who was in favour of the previous changes and that they're reasoning was contrary to reliable sources and indeed several Wiki policies. By all means lets hear sensible alternatives (and make sure they are applied consistently) but there is no urgency or indeed necessity to move away from the current version. Owl In The House (talk) 22:33, 28 August 2014 (UTC)
The above claim that having a second table "would not be a reflection of our reliable sources" is not true. As I quite clearly demonstrated above, half the polling companies (i.e. reliable sources) regularly report on the polling data for the Greens, BNP, SNP and Plaid in their first table. The other half all include them in subsequent tables. The information is available, and there is no reason not to include it; I really don't understand the desperation to avoid painting the full picture for readers. Number 57 22:52, 28 August 2014 (UTC)
I can't really understand what you do actually want, Owl. First, the discussion was already closed, consensus being already reached about whether to add the Greens or not to the main table (which was even voted for) and resulted in a 'No for the time being, let's wait for future developments' response. That said, ever since several users have proposed several alternatives to that scheme, to which you not only oppose (which by itself is an actually very respectable opinion), but also claim it to be impossible to do because, allegedly, Wikipedia policy says otherwise. First, I already discussed Wikipedia policies with you, stating how some of your claims were not correct. You did not counter me in some of those points, yet you still claim that others' ideas somehow go against Wikipedia policies. Sheltering into Wikipedia policies in this way in order to defend your arguments is not right, foremost of all, because you seem to be misinterpreting some of them (I don't know if, as of currently, there is a policy stating that articles should be strictly consistent between themselves (i.e. the Opinion polling article for the 2010 general election should have exactly the same structure that the 2015 UK election opinion polling article)) and also because you seem to be ignoring WP:RAP and WP:IAR, which are also Wikipedia policies. Secondly, you are bringing forward details about an already closed discussion into a new, different one (that is, making a separate table), which is also not right. Again, I tell you that it is right to opine and to be against others' views because yours don't match them. But that is one thing, and another one is your current stance, which seems to be strictly against adding anything else to the article as it is now and responding to others in kind. Please, be reasonable. Impru20 (talk) 00:52, 29 August 2014 (UTC)
I entirely support the Marshall Plan. It seems a sensible way forward. (Anyway, as I see it, WP:LOCALCONSENSUS and WP:PERFECTION argue against Owl's interpretation of consistency.) Bondegezou (talk) 15:13, 29 August 2014 (UTC)
Being bold and to move discussion forward, I've tried to start the new table with detailed results, including all parties that the polling organisation gives details on. I'm lousy at tables, but it's a beginning! Bondegezou (talk) 15:23, 29 August 2014 (UTC)
Good work. I've formatted the table to give the others their correct colours and solve some width issues. Number 57 17:39, 29 August 2014 (UTC)
I first want to point out this sentence from my previous comment in this thread: "We need to be consistent between articles (that doesn't mean they must be uniformly identical but it does mean the criteria for inclusion should be broadly the same - at present they are).". - It is disingenuous for other editors to say that I have said otherwise, clearly comments suggesting I have said otherwise should obviously be disregarded. My objections are on the basis of criteria for inclusion. I actually have nothing firmly against the new table at the bottom of the article, it doesn't seem to detract too much from the main body of the article. That said it does look a bit out of place. Will somebody be going through every single poll and adding that to the table? If so, good. Also if someone is prepared to do that then why can't they also do it for the previous article? Like I say the status/polling significance of the SNP, Plaid, Greens and BNP hasn't changed since then.
When I say it wouldn't be a reflection of reliable sources, I am of course referring to the piece of data the polls themselves highlight, obviously the pollsters publish lots of information but they do provide their own summary and it is this that gets published and highlighted by reliable sources. That said I am not rigidly against having a table like the one suggested at the bottom of the article, those who think I am saying otherwise, could you please get that in to your heads. I was also taking into account that we tend to avoid repeating and duplicating information, especially within the same article.
All that said, I don't find anything overly objectionable about the suggested table provided that this is a change that is done properly. Indeed if done properly I would consider it an improvement to the article but if it is not done properly and done in a way just to please or include certain individuals or view points then frankly I'm against any change. What do I mean by that?
1. The table would need to cover every single poll since the general election and not simply start from today (or whenever), any change would need to be retrospective. This was the same basis that UKIP were added to the main table and yes consistency is important.
2. The same table would need to be added to at least the previous polling article, (again for most of these parties the grounds of inclusion/their status and significance simply haven't changed).
3. This table should be one single table, not one for each year, it is important that this table is a "scroll down for further details" table and does not become the main focus of the article because that would not be an accurate reflection of reliable sources. Yes they publish both Headline polls and the full data tables but it can't in any way be argued that the full data tables are the main focus of reliable sources. It should be a case of making extra information available to those who want it, in a way that reflects the coverage of reliable sources, not the wooly arguments we saw in the previous discussions of "inclusion" etc.
4. I also wonder whether it's worth adding Respect to this new table, I don't have a firm view either way on this, I'm just putting it out there.
If these points are met then I am in full support of the table
May I also just restate that I have never been rigidly against change, just the proposals that had been put forward, it is worth noting that the current suggested table has never been put forward on this talk page, it was boldly added. I want to be constructive here and I feel my above comments on the table its self are constructive. Indeed in some ways I am suggesting we should go further with it but like I say we should only do this if we are to do it properly, or we shouldn't do it at all. Owl In The House (talk) 12:37, 30 August 2014 (UTC)

Detailed poll results

I have cut and paste this from the main article to the talk page as this is still ongoing. Thank you for creating the table, we now have something solid to work with, now that I know what you're actually proposing and where you propose to put it, I actually think it's quite a good idea, provided the three points above are addressed. Owl In The House (talk) 13:16, 30 August 2014 (UTC)

___________________________________________Cut and Paste (Start)___________________________________________________________
Detailed Poll Results
The tables above show poll results just for the four largest parties. Detailed poll results are given below:
Date(s)
conducted
Polling organisation/client Sample size Cons Lab Lib Dem UKIP SNP Plaid Green[n 1] BNP Others Lead
27–28 Aug Populus 2,006 35% 34% 8% 13% 4% <1% 5% <1% <1% 1%
________________________________________________Cut and Paste (End)_________________________________________________________
I have restored the above content. As per WP:PERFECTION, further editing to improve the table can take place on the actual article. As per WP:CONSENSUS, only one person, or at least nocturnal bird, has shown opposition to this addition (and even s/he now calls it "quite a good idea"), which was the result of an RfC, so it seems to me entirely within policy to keep this material in the article until such time as consensus should change. Bondegezou (talk) 09:37, 31 August 2014 (UTC)
Fair enough, if we see editors starting to make some inroads into backdating this change within the next 48hours, I'm agreeable that there is consensus for it staying, I understand there is a lot of work to do on it, so patience is required. However, I maintain my above comments and that this table should cover every single poll in this article for it to be appropriate. I hope those who argued the case for it will put just even more effort into implementing it, otherwise it will have to go. Owl In The House (talk) 10:15, 31 August 2014 (UTC)
I do not feel that your attempt, Owl, to impose an arbitrary deadline of 48 hours is useful or in keeping with how Wikipedia works. It would be nice for the detailed results to cover every singe poll in the article, but if that doesn't happen or doesn't happen quickly, that doesn't change the situation. I refer you again to WP:PERFECTION. We had an RfC. No-one other than you has raised any objections to the RfC result or subsequent editing to put the recommended changes into effect. Myself and three other editors have stated their support for this approach and/or got on with making the changes. You have made your views clear, but I would urge you to now respect a WP:CONSENSUS decision. Bondegezou (talk) 18:10, 31 August 2014 (UTC)
It was not an arbitrary limit for this to be completed by, it was time for people who wanted this change to show that they do intend to backdate this. User:Impru20 has shown that signal of intent by starting to backdate and has made a great start. I hope you didn't think I meant it must all be done in 48 hours...a) I don't have the authority to say that. b) its not realistic and c) thats not what I wrote, I said that I understood a bit of patience would be necessary. All I wanted to see was a sign of intent that what has been agreed Unanimously will be implemented to the full. There is no disagreement here, however, if we get to the point of being in the run up to the election and whats been unanimously agreed hasn't been implemented then I will consider starting a discussion to have the table removed, if it is incomplete then theres not much point in having it. I'm sure it won't come to that though, User:Impru20 has made a great start. Owl In The House (talk) 20:53, 31 August 2014 (UTC)
I refer you again to WP:PERFECTION: there is no reason to remove the table simply because of incompleteness. That's a pretty core principle behind how Wikipedia works. Bondegezou (talk) 21:52, 31 August 2014 (UTC)


It seems a little odd to include SNP and Plaid Cymru without noting that they are only possible to vote for in their respective nations. Their proportion of the total vote would likely be far higher if the individual results were shown for Scotland and Wales respectively. Adam Cuerden (talk) 23:12, 2 September 2014 (UTC)

We are including every party that each poll lists. We could add a footnote to explain about the SNP and PC. Bondegezou (talk) 09:42, 3 September 2014 (UTC)
I think it would be more helpful to note it above the table, in one sentence. I don't think that we need to belabour the point, but stating, clearly, that those two parties are limited by geographical area puts them in context. Adam Cuerden (talk) 10:50, 3 September 2014 (UTC)
I've done a footnote, to match that done for the Greens. I have no objection to those, now, three notes being changed to text above the table.
By the way, a small point, but perhaps we need to be clear on when to put "<1%" and when to put "0%". I've put "<1%" when the pollster found at least one person saying they would vote for that party, but they reported the number as less than 1% (often just marked as "*" on results tables). However, if a pollster found no-one saying they would vote for a particular party, then "0%" would be appropriate. Bondegezou (talk) 13:39, 3 September 2014 (UTC)
During checking my cockup on the 0 vs * issue, I've found it impossible to distinguish between 0 & "rounds to 0" in yougov polls. As <1% does imply more than 0, perhaps some sort of note would be appropriate. Iliekinfo (talk) 17:56, 3 September 2014 (UTC)
Done. Impru20 (talk) 18:13, 3 September 2014 (UTC)
In all honesty, I think that a paragraph before the table, contextualizing it, would be far, far better. Remember that not all readers of the article will necessarily be from the UK, and I do think the results are somewhat misleading if you don't know, for example, that Wales only has around 5% of the population of Great Britain, and Scotland just a bit over 8% - that greatly recontextualizes the figures - getting 3-5% support overall when only 8% have the choice is a very different thing than the uncontextualized data; and we don't have individual sections or, so far as I can see, article on this subject in Scotland or Wales, so I think it matters. Adam Cuerden (talk) 18:05, 3 September 2014 (UTC)
I'm all for specific polls of Scotland or of Wales being included on an appropriate page, just as we have polls on specific constituencies or on sets of marginals. And if you want to re-write what is currently in footnotes into text before the table, do give it a go. Bondegezou (talk) 20:17, 3 September 2014 (UTC)

I have added incomplete & outdated tags as this section is now both. Would it not be better to complete it elsewhere and import it when complete? Is this possible? One problem is double-entry of data, I for one do not have the time or inclination to enter polling data twice, one in the 2014 summary table and once in the detailed table. In practice, completion of the detailed table is slow (not surprising, given the amount of work involved) and it brings down the quality level of the whole page.Saxmund (talk) 11:34, 14 September 2014 (UTC)

It is the nature of Wikipedia that it is often incomplete: see WP:PERFECTION. It is not Wikipedia policy to hide away material while it is being completed because doing so discourages the editing required to complete it! Bondegezou (talk) 14:44, 14 September 2014 (UTC)
WP:PERFECTION, "Also, redundancy within an article should be kept to a minimum (excepting the lead, which is meant to be a summary of the entire article, and so is intentionally duplicative)." Iliekinfo (talk) 02:20, 15 September 2014 (UTC)
Yes, redundancy should be kept to a minimum. In this case, the conclusion of lengthy debate on the subject and an RfC (see above) was that the way forward was to use two tables, i.e that two tables is the minimum degree of redundancy required to satisfy a desire for a table focused on the main parties and to include all the data on smaller parties that reliable sources make available.
The RfC closure is still recent and, indeed, has already been reviewed. It seems inappropriate to me to re-open that discussion so soon, although I note that some are unhappy with it. I feel we should try and make the 2 tables plan work (within existing policy and guidelines). If, at some future point, consensus is that this approach is not working (or, at least, there is no consensus that it is working), of course we can re-visit the plan. Bondegezou (talk) 10:44, 15 September 2014 (UTC)
I wasn't involved with the discussion, but having read it a few times:that was not the conclusion. That was given as an....apparent "ruling"...by someone uninvolved in either discussing or implementing the changes, it didn't actually feature in the discussion at all, let alone as a conclusion. If I miss the next time this is 'officially' debated, have it on record that I'd somewhat prefer the article to be the "detailed" table only, but only if the detailed table is complete, otherwise the "simple table" only. "I feel we should try and make the 2 tables plan work" Why? What for? What is the actual argument in favour of two tables? Iliekinfo (talk) 17:03, 15 September 2014 (UTC)
I admit I didn't mention it at the time (I went for the status quo) but I think it would have been preferable to have moved to a detailed table after the next General Election which after all is no more than 8 months away. We could then have started a new set of results in a more detailed format, without having to retrofit 4.5 years' worth of detailed data.Saxmund (talk) 17:34, 15 September 2014 (UTC)

An RfC is meant to be closed by an uninvolved editor. That's how they work. Bondegezou (talk) 18:14, 15 September 2014 (UTC)

Yeah, but https://en.wikipedia.org/wiki/Wikipedia:Closing_discussions#How_to_determine_the_outcome.
I'm all in favour of Saxmunds "next time" proposal btw. Though more precice rules about who makes the table or doesn't (Respect?) might be best decided ahead of time...Iliekinfo (talk) 18:17, 16 September 2014 (UTC)


Cite error: There are <ref group=n> tags on this page, but the references will not show without a {{reflist|group=n}} template (see the help page).