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

Event bars revisited

I have just deleted an event bar which is of the "party leader X says Y" variety which I don't think would have made the cut even before our recent cull of Notable Events. Might be worth discussing what should go in during the election campaign. TV debates (and substitute programmes) I think are agreed. The Scottish table now has "start of official campaign" (which I think ought to mention the dissolution of Parliament as I think that is a bona fide factual event) and the launch of the Scottish Green manifesto. Do we include manifesto launches? And if so which ones go in the GB table and which ones in the "sub-national" tables? (I think the SNP ought to go in the GB one, if we include manifesto launches, as it is polling well enough to have an effect on the outcome of the election). Just thought it worth mentioning as the tables could become rather cluttered. Saxmund (talk) 18:59, 30 March 2015 (UTC)

Hi. Broadly agree with you, User:Saxmund. No to X said Y, yes to debates and debate substitutes. However, I'd also say no to manifesto launches. It's an election, of course parties are launching manifestos, they're not notable events. In terms of Parliament being dissolved &c., we could go with whatever is listed at the Parliament website timetable. Bondegezou (talk) 22:16, 30 March 2015 (UTC)

I support the official parliamentary events regarding the beginning of an election period being placed in the events bars. Any time somebody says something shouldn't be included, but well reported controversies should be included such as "Bigotgate", and major TV appearances of the five party leaders. Launching manifestos should probably be included, since there are many predictable events that we include in notable events, incidental with major party events. The only ambiguity is the place of the SNP, which is big enough but very unclear. Onetwothreeip (talk) 03:20, 31 March 2015 (UTC)

If there is another Bigotgate, I wouldn't include it. It is too much of a matter of opinion what is notable enough, and it is too open to partisan interpretation. What is notable enough to include? Does the Sturgeon memo count? I would stick with fixed political events, and the practice here has been for a while to include set-piece TV debates and similar. To try to make a principle out of it - maybe we shouldn't include anything which is in the control of a single political party, either planned (which would exclude manifesto launches for example) or unplanned (which would exclude Bigotgate-type events). I would also say we shouldn't include any TV events that only include a single party leader. Just some thoughts. Saxmund (talk) 11:31, 5 April 2015 (UTC)

England

Politicalbetting.com now lists a number of England-only polls. Should these be added? 193.62.42.156 (talk) 09:13, 23 December 2014 (UTC)

I would certainly hope so, in the section currently labelled 'sub-national polling'. Please do start a new sub-section! DrArsenal (talk) 11:45, 23 December 2014 (UTC)

Well I have started it. Of the four polls listed in this Politicalbetting post I can't work out how to disaggregate the England figures from Ashcroft and Ipsos-MORI. ComRes was easy, I haven't had time to look at ICM yet, that's next along with the 2010 GE votes. We will now need to look out to see if England figures are carried in GB VI polls.Saxmund (talk) 16:38, 1 January 2015 (UTC)

There is now a table with the four polls discussed by Mike Smithson and the General Election results. There is probably some work to be done to check that no other pollsters are publishing identifiable England data and that none of these four have done so predating December 2014 (Smithson's comments imply not, but are not conclusive).Saxmund (talk) 20:53, 1 January 2015 (UTC)

I think I have now added all polls giving an identifiable England sample going back to Dec 2014 when the existence of England-only data was first flagged up by Mike Smithson. In addition to the polls listed by him (Ipsos MORI, Ashcroft, ComRes (phone), ICM) Survation and Opinium have started giving England figures since then. There is a possibility that some of the original four will have England data prior to Dec 2014 which I will look into if I have an idle moment, but I think it is more important that we show a full data set for recent polls leading into the election. Saxmund (talk) 20:59, 5 April 2015 (UTC)

Column 1 = Date(s) conducted NOT date published

I refer to the placing of the YouGov poll conducted between 1st and 2nd April 2015. Should come before the 7-way Debate in the table since it was conducted (in part) on the 1st, ie. before the debate, even though it was published after. Column 1 is NOT "date published" as some editors seem to think. best, Sunil060902 (talk) 19:04, 5 April 2015 (UTC)

SleepCovo does not yet seem to understand that polls are primarily listed by fieldwork date and not by publication date/time and has been reverted again this morning. I have tried to make contact on his Talk page, and you have invited him to this page, but to no avail. Hopefully he will desist. He has now apparently been blocked for edit warring (I had a warning about it yesterday), and I see he has two previous blocks this year. I suspect we will get a few new people editing this article during the election campaign and they may not be aware of our usual standards and practices. Anyway, I am going to have a day or two off as I spent a long time yesterday updating England polls and a few other bits & pieces. Saxmund (talk) 10:09, 6 April 2015 (UTC)

Just to add, where two polls have exactly the same fieldwork dates I always list by publication date/time, for example Populus (usually released by lunchtime on a Monday) comes before Ashcroft (4pm) and I think this reflects other editors' practices as well (it certainly seemed to be the case when I started editing a couple of years ago and I started doing it because it appeared to reflect current practice) Saxmund (talk) 10:09, 6 April 2015 (UTC)

Better way to sort the table on opinion polls?

Was wondering if there's a better way to sort the table on opinion polls. Currently, if you sort the table away from the initial display (like, by sorting by margins, for instance), you can't go back to sorting the table by dates anymore. That's because the first column has the day and the month together; so if you sort by the first date, you'd get all the 1's together, and 2's and so on. Perhaps we could have another column displaying the month separately? That way, you'd be able to sort by day and month. (I'd do this, but don't know enough Wiki syntax to pull it off :) ) Twotino (talk) 10:19, 10 April 2015 (UTC)

Polling conducted just before 7-way Leaders' Debate on ITV on 2nd April

Someone, for reasons unclear, is insisting on amending this page to show that polling conducted by Panelbase 31 March-2 April and You Gov (for The Sun) 1-2 April show up as having occurred AFTER the 7-way televised Leaders' Debate on ITV from 20:00BST-22:00BST on 2nd April, even though the fieldwork for both pieces of research will either have occurred entirely or almost entirely BEFORE the Debate. I have tried to correct this more than once, but I don't want to be involved in an edit war. Can other users please input here? 11-April-2015, 18-06BST86.27.84.55 (talk) 17:07, 11 April 2015 (UTC)

I must agree with you. A poll conducted within a fieldwork date range that includes days before a given event is obviously conducted before that given event. For a comparison, the 7-way Leaders' Debate would have been held on a fieldwork range of 2 April-2 April (2 Apr to sum up). 1 April-2 April (1-2 Apr) and 31 March-2 April (31 Mar-2 Apr) would be obviously previous to the debate. Cheers. Impru20 (talk) 17:14, 11 April 2015 (UTC)

I agree too. Consistent practice has been to list polls by fieldwork dates, and slot events in the same strict chronology. Someone seems not to understand this very simple concept. Saxmund (talk) 10:54, 12 April 2015 (UTC)

Agreed with everyone else. Bondegezou (talk) 12:24, 12 April 2015 (UTC)

<WP:SOAPBOX comments removed by Dweller. --Dweller (talk) 16:15, 16 April 2015 (UTC)>

SNP

I am aware that the percentage changes for the SNP are not as significant as the seats they will win, but surely a party predicted to come third in the election (by seats) should be included in the main table, particularly seeing how many parties are listed in other countries' opinion polling pages and in light of the Lord Ashcroft poll which put them ahead of the Greens? Thenextpm (talk) 09:48, 22 April 2015 (UTC)

We have a detailed polling section that includes the SNP, but that's not getting much use. They are, of course, covered in the Scotland section. Some polls only report a combined SNP/PC figure, which makes reporting an SNP result complicated. Without the SNP regularly being reported in a standard way by polls, it is fiddly to cover them. I would be happy to see the detailed table updated. Bondegezou (talk) 10:54, 22 April 2015 (UTC)