Talk:Opinion polling for the next Scottish Parliament election

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

Graphs[edit]

Concerned about the graphs and in particular the lack of any Text on the scales on x and y axis, without this it is difficult to know what weight to give to the coloured lines. A consequence is that they produce a large area of white space at the beginning of the article which rather spoils what is a very neat, clean look. Can we include more information on the scale points, please. Soosider3 (talk) 12:55, 23 March 2023 (UTC)[reply]

Think we need a statement as to what methodology is used to produce the graphs, perhaps a footnote at end of article. Soosider3 (talk) 15:46, 24 March 2023 (UTC)[reply]
I don't understand; could you explain what you mean? Unless you mean that the scale is not clearly visibly; which I admit I've just noticed. EnglishPoliticalPerson (talk) 16:16, 7 April 2023 (UTC)[reply]

Soosider3[edit]

@Soosider3: could you explain your removal of the seat projection section further? The seat projections are not conducted by a blogger, the source Ballot Box Scotland is cited simply for ease as it lists all of the seat projections on there, and it is not an unreliable source. The data is also there which the same as you added for the opinion polls? ErraticDrumlin (talk) 21:46, 17 June 2023 (UTC)[reply]

Seat projection has not been a part of this article until very recently, mainly as it is a rare occasion when any pollster does one for Scotland. Pollsters are the most reliable source we can have and why it is a requirement that pollsters are members of BPC and require publication of methodology and data tables. The same can not be said for a blogger, no matter how good he may or may not be. It seemed your section relies almost wholly on that one blogger and in that context can not be seen as a reliable source. There are several sites out there that also do seat projections, how do we choose which ones to use? what process should we adopt to ensure they provide the highest level of reliability that we should all be aspiring to have in this article.
There is the world of difference between referring to a twitter post, just as we would refer to a newspaper article when poll first becomes available, however this is a holding position until the data tables become available.
I hope that is helpful to you and assure you my only motivation is trying to maintain as high a standard as we can and hope you would share that aim. Soosider3 (talk) 22:14, 17 June 2023 (UTC)[reply]
Okay I hear you, but what about the seat projections that are done not by the Ballot Box Scotland blogger but by actual pollsters? Such as the ones that already existed on the article before I expanded it, referenced to national newspaper sources such as here and here? This justification does not seem to cover the exclusion of this content, and just to clarify, the section was not really 'mine' as it was created by others before me, I just added to it. I see why Ballot Box Scotland seat projections should be inadmissible, so the section could be limited to those projections done by pollsters with a trusted methodology. ErraticDrumlin (talk) 22:44, 17 June 2023 (UTC)[reply]
Good morning
May I suggest having a look at this article https://en.wikipedia.org/wiki/Opinion_polling_for_the_next_United_Kingdom_general_election#Seat_predictions
In particular the methodology used by polling companies called MRP, it is a specific method used and very different from the methods used for opinion polling, note that across the UK there were only 10 of these done in 2022. I am not aware of any being done for Scotland.
The 2 examples you gave are regrettably only a third party projection rather than Polling companies doing the MRP Soosider3 (talk) 07:49, 18 June 2023 (UTC)[reply]
Will with @Soosider3 on this. Unless published directly via a pollster I would not include third party protection - just as we don't with UK polling. I don't recall the projections included to have been via pollsters. EnglishPoliticalPerson (talk) 22:52, 18 June 2023 (UTC)[reply]

Graphs are out of date[edit]

I might be wrong, but it looks like the graphs are many months old. Can we get an update?

Also: the yellow datapoints are almost invisible on my screen. Larger/darker/outlined for all datapoints would be good: the trend lines are very visible, but they are, after all, not the base data. RERTwiki (talk) 21:22, 5 August 2023 (UTC)[reply]

Agreed, not updated since early May. I would be happy to help but have no idea how they are done but would like to learn, alternatively we could do a new graph using something like dataWrapper. Discussion and agreement on methodology would be a good thing to allow us to spread the load. Perhaps @Englishpoliticalperson could spare the time to talk us through it Soosider3 (talk) 07:14, 6 August 2023 (UTC)[reply]
Good Morning
Following on from above I have had a go at creating Graphs using datawrapper, but firstly a few points
1 These are potentially a start - not necessary an end point
2 They are very straightforward and make no attempt at Regression etc
3 I like there simplicity - where you dont have to have any understanding of graphs - yes they are messy but then so are polls
4 I like that we can choose to give access to data (consistent with Wiki philosophy) and direct links to Social Media
5 I like that being open allows any to update graphs - removes a bottleneck and spreads the load
Any how - have a look and see what you think
https://www.datawrapper.de/_/f1fiN/
https://www.datawrapper.de/_/hxa0R/ Soosider3 (talk) 06:09, 14 August 2023 (UTC)[reply]
I think these are amply sufficient for the case when the usual editor hasn't got time to do the charts. Just put them in. 217.46.134.126 (talk) 12:08, 14 August 2023 (UTC)[reply]
I ought to have said, I actually prefer the DataWrapper charts. They are bolder and clearer, and I'm not a big fan of Loess (the recent part of the curve is volatile). RERTwiki (talk) 12:40, 14 August 2023 (UTC)[reply]
Okay have published graphs for both votes
However due to limitations of Wikipedia(i think) have not been able to publish the graph as it is in datawrapper and only been able to publish a static image. I may include the link to datawarapper so that folk can get the whole thing complete with data used.
In using publishing wizard it has defaulted to creating a thumb of image, not sure if this is an improvement or not but it certainly reduces the amount of white space and gives prominence to the polls. Think I will leave this for moment and give other opportunity to make comment. Soosider3 (talk) 05:47, 15 August 2023 (UTC)[reply]
Just had a look on mobile and graph looks fine Soosider3 (talk) 08:32, 15 August 2023 (UTC)[reply]
Thanks for putting these together. I'd be happy to produce graphs with trends like the ones I do for Opinion polling for the next United Kingdom general election and a couple of other pages. Dot-to-dot lines can be quite visually cluttered compared to trendlines and can obscure long-term trends rather than illuminate them. Ralbegen (talk) 18:06, 15 August 2023 (UTC)[reply]
sorry all; I've been away the last 2 months and so hadn't been able to update the graph. I'll update the graphs today and put them back up. I think it's advisable to keep them LOESS as they're clearer and neater - and as a visual representation they're more useful. I'll darken the yellow data points to account for @RERTwiki concerns. EnglishPoliticalPerson (talk) 06:58, 24 August 2023 (UTC)[reply]
Appreciate the offer, however perhaps it would be best to get comments from other editors before reverting to previous approach.
Current approach has IMHO a number of advantages not the least is that anyone can do the updates, it is open and transparent, the graphs seem to be neater, tidier and much easier to see, and LOESS may not be to everyones taste. Soosider3 (talk) 08:15, 24 August 2023 (UTC)[reply]
I support the graphs having LOESS trendlines over dot-to-dot lines. While this does restrict who can currently update the graph, the code used to generate the graph for Opinion polling for the next United Kingdom general election had been uploaded to User:Ralbegen/Opinion poll code, so uploading his code for generating a similar graph for here doesn't seem to be out of the question.
@Ralbegen has also expressed a to upload the code to GitHub after cleaning up the code to have it be in the form of a package. Once this is done, it could be further built upon to allow anyone to pull the code and data, and follow some basic instructions to generate an updated graph.
But in the meantime, having the graph be generated similarly to how it's done for Opinion polling for the next United Kingdom general election is fine to do for now. Danyaal99 (talk) 10:34, 3 September 2023 (UTC)[reply]
Good to see you back. I think if you maintain the page you should do it as you wish. Loess is a battle where I’m used to being a voice in the wind: but it’s ok except at the tail of the curve. The current interim chart has graphic faults as well, though the common update might be an advantage.
The trouble there is that any tool we use has to be as good as excel, which will never happen. RERTwiki (talk) 08:26, 24 August 2023 (UTC)[reply]
Fair enough
May I ask that
1 the data points on axis are clearer
2 Some short statement as to methodology used (one that lay person could understand) Soosider3 (talk) 08:32, 24 August 2023 (UTC)[reply]
Agreed 217.46.137.182 (talk) 18:03, 24 August 2023 (UTC)[reply]
To be fair; I can upload the dataset to a github or something of that regard. All I'd need to do on my end is generate the new graph; this does spread the load and so long as contributors remind me I can implement the updates as frequently as needed. For the data points I'll increase the alpha; and perhaps change the colour to be darker for the SNP. I've always preferenced the lighter data points but I understand why this would be controversial. As for the methodology bit that should be more than simple to do. My main opposition to a dot-dot diagram as used currently is largely that; contrary to what's been said it's quite clutersome and doesn't really present the data in the most understandable way even if it's the most true representation as it gives too much weight to each individual point; while it's not too bad at the moment; as soon as an anomaly becomes present it'll be an absolute eyesore. EnglishPoliticalPerson (talk) 20:38, 2 September 2023 (UTC)[reply]
That sounds like a sensible way forward, especially if we can spread the load Soosider3 (talk) 09:40, 3 September 2023 (UTC)[reply]
Is it possible to publish them as a thumbnail like at present? Although I came across that by accident I think it is preferable, reduces a lot of white space at beginning of article and presents the main topic of article the "polls" in more prominent position Soosider3 (talk) 09:43, 3 September 2023 (UTC)[reply]
I've swapped the graph's over again; I've adjusted it to take into account peoples concerns here. Any other issues; don't hessitate to mention them here. I've included all the data and files in a github available in the summary box on both of the files so if anyone has any knowledge of R they can make those changes themselves if they wish + can generate the graph if they wish. The spreadsheet is Excel format so any can edit that and they should be able to update the file on github! EnglishPoliticalPerson (talk) 16:31, 4 September 2023 (UTC)[reply]
Many thanks for that, very much appreciated. Had a look at info on gitfile and seems fairly straightforward (famous last words) So process is
Update Excel Spreadsheet
Update file on github
It should pull through to wikpedia. Soosider3 (talk) 17:50, 4 September 2023 (UTC)[reply]

Should we be including Dont Know/undecided where ascertained they will vote[edit]

Can we have a discussion on whether to include Dont know/undecided where they have been identified as likely to vote but at this time are unsure who for. It seems to me that excluding them gives an inaccurate picture of VI I would suggest we include a column for those undecided Voters Other views? Soosider3 (talk) 19:01, 6 September 2023 (UTC)[reply]

We should follow reliable sources, taking into account what pollsters and media do, which in the UK is not to include undecided voters in headline results. Calculating our own results would not be appropriate. Ralbegen (talk) 21:23, 6 September 2023 (UTC)[reply]
Not sure I completely follow your logic here.
Firstly there is no model as to how data is presented, we should be presenting data in as full and complete manner as possible.
As regards polling for this article, there is good and reliable data produced by the pollsters that identify those likely to vote but undecided/don't know, in particular from Redfield & Wilton who have become the most regular pollster on Scottish Politics. This is data that helps give a fuller picture of VI - should we really be limiting reader access to this by failing to record it in article?
As Redfield & Wilton have emerged as most regular of pollsters in Scotland it throws up another couple of queries, like opening a column for Reform, assumed this has not been done previously as not all pollsters include them in Scottish Polling. Soosider3 (talk) 03:28, 7 September 2023 (UTC)[reply]
We report the headline results as presented by the pollsters, media sponsors, and most other coverage. Wikipedia, including polling articles, summarises reliable sources and isn't (and shouldn't be) an innovator in how to present these polls. In America, pollsters and media typically report figures without undecided voters removed, and the Wikipedia polling articles for American elections follow them. If readers are interested in additional results from the polls included here, they are all (or should be!) linked and available to peruse in full. We are not limiting reader access to these, or cross-tabs by region and age, or the various other results available in full polling tables that aren't routinely headline results from pollsters or reported by secondary sources. Ralbegen (talk) 19:26, 7 September 2023 (UTC)[reply]

Reform[edit]

It would appear that Pollsters in Scotland are now regularly including Reform in their questions on VI, should we include a column for them? There returns are small but no smaller than Alba or Greens Soosider3 (talk) 09:48, 2 December 2023 (UTC)[reply]

Scoop Polls[edit]

I had introduced the section for Scoop polls as this seemed a sensible way to present the data they produce, given that the pollsters themselves have stated not to compare them with other Yougov Polls, including them in the main table does exactly that. A look at any article that includes them shows they are clearly well out of line with other main stream polling (by about 8%) I would argue that is further evidence of the wisdom of YouGovs advice. A secondary point in support of a separate section for Scoop polls is to do with a significant difference in the number of polls conducted at UK and Scottish level. In November 2023 there were 34 UK polls in Scotland 2, October 38/6, September 31/3. The impact of this is that at UK level a poll at significant variance from the average, will have little impact at UK level but with such few polls in Scotland it would have large impact on the averages and by that the Graphs. This would present to readers a misleading view of what is actually happening at polling. Creating a separate section allows readers to view the data in the context stated by YouGov ie to compare only with other Scoop polls, this adds value to the article. I have and will continue to revert any removal until such time as there is a coherent rationale to mainstream them, so far that appears to be maninly the work of an anonymous editor with no history of involvement in other articles. Soosider3 (talk) 14:42, 6 December 2023 (UTC)[reply]

This very same discussion took place at Talk:Opinion polling for the next United Kingdom general election#Scottish Election Study poll and you were repeteadly rebuked for your behaviour. The same rationale applies here. It's you against everyone else here, it's the "coherent rationale" the one you are fighting against. It's clear you don't have a consensus for making this change, neither here nor at Opinion polling for the next United Kingdom general election. I warn that if you keep on your edit warring behaviour as you seem to announce, you'll probably be blocked by some passing admin, so I'll strongly advice you to stop attempting to enforce your particular view of things unilaterally. Show there's a consensus for your edits. So far, you haven't. Impru20talk 16:03, 6 December 2023 (UTC)[reply]
Nope, my issue is that their is no coherent rationale for including Scoop polls in main table, their is a view that flies in the face of stated view from pollster, that creates a misleading view for readers, I am happy to continue to rail against that sort of daftness Soosider3 (talk) 16:11, 6 December 2023 (UTC)[reply]
There is a coherent rationale and that has been explained to you at Talk:Opinion polling for the next United Kingdom general election#Scottish Election Study poll. You not liking it is not a reason to ignore all other users not agreeing with you. What you are doing constitutes edit warring, which is prohibited in Wikipedia. Also, Wikipedia is not a place to right great wrongs: if you perceive there is something wrong but many other users don't agree with you, maybe you should consider that it's you the one that is in the wrong. If you keep with this behaviour, admin intervention will have to be sought. Impru20talk 16:21, 6 December 2023 (UTC)[reply]
Lots of people have patiently explained to you why you are misunderstanding what YouGov said. Everybody else seems to understand perfectly that the SCOOP polls form a separate series to the main YouGov polls in much the same way as different pollsters' polls are different series, and that data users should be aware that a change from poll in series A to a poll in series B could as easily be due to methodology as either an underlying change in voter sympathies (or random noise). Series A and B could be Survation and Redfield and Wilton, or they could be YouGov's main polling series compared to the polls they do for SCOOP.
In any case, it's not constructive editing to ignore a consensus on one page only to pursue your own position, representing a minority of one, on other pages with fewer eyes on them. Ralbegen (talk) 16:27, 6 December 2023 (UTC)[reply]
We discussed this matter at Talk:Opinion polling for the next United Kingdom general election#Scottish Election Study poll and reached a clear WP:CONSENSUS. The same logic applies here. Bondegezou (talk) 17:19, 6 December 2023 (UTC)[reply]
By what measure do you or anyone else decide that discussion in one article should be applied to a separate and different article? Soosider3 (talk) 10:48, 7 December 2023 (UTC)[reply]
By what measure do you decide to ignore a consensus in one article and attempt to override it in other articles, taking advantage of fewer attention being put on them? If your edits found opposition from other users in another article, it is all but logical that similar edits will likely find opposition in other articles. Thus, a cautious approach should have been taken. But here we are not even speculating: you met opposition from other users in this very same article, yet you chose to ignore them and push your edits forward anyway. Impru20talk 11:18, 7 December 2023 (UTC)[reply]
By what measure? By the expression of views of different editors, demonstrating a WP:consensus. It's currently 3:1 against you here. I have not seen anyone ever agree with you on this point, Soosider3. Have some humility and acknowledge that for once. Bondegezou (talk) 10:58, 8 December 2023 (UTC)[reply]
stepped back for a wee while and reflected on situation. To my mind it boils down to one simple thing, and that is facts v opinion. I realise I am the lone voice here, but that does not make me wrong, merely outvoted.
Fact is Yougov stated unequivocally not to compare with other YouGov polls, those voices raised against that are expressing their opinions, as they are entitled to do, however opinion is not the same as facts. Reinterpreting meaning doe snot change the fact, disassembling doe snot change facts. Including polls directly into article flys in the face of the statement from YouGov as without a mechanism to differentiate they will be directly compared to other yougov polls, especially for anyone not aware of the difference, there being no mechanism to advise and inform the reader of anything different. The folk who will produce the graphs will include them again reinforcing the point I have made.
Only relying on consensus without giving proper weight to stated facts is in this example an expression of a fundamental weakness in editing Wikipedia, where opinion is apparently given equal weight to facts. Soosider3 (talk) 02:20, 9 December 2023 (UTC)[reply]
Why is the inclusion of SCOOP polls in the graphs such a concern of yours? Polls of all methodologies are included in the data tables and graphs, this is to allow people to compare the trend over time and to observe methodology differences. The differences between the Scoop series and normal YouGov polling are no different to the differences between YouGov and pollsters, and it’s widely known that it’s unwise to compare pollsters to each other, yet they’re still presented on the same table. There’s a reason we don’t have a table for all individual pollsters, it makes it even more confusing for the reader as all pollsters have methodology differences. It’s not up to you to decide what will and will not confuse the reader, they need to do that for themselves and form a balanced opinion, which excluding a set of polls for no reason other than a misunderstanding of the rationale behind the methodology doesn’t help with. Myself and numerous other editors, not just on this page but on others, have reverted your edits and explained why we have done so, yet you continue to wage this edit war despite being the sole editor out of several who seems to have a problem with this current setup. Please do not unilaterally impose your view on the article when talk pages on several articles have explained to you why you’re wrong. GamerPerson12345678 (talk) 12:43, 10 December 2023 (UTC)[reply]
It is noticeable that you fail to actually answer the thrust of my argument and revert to questioning motives, other say etc, none of which produces a single piece of verifiable evidence to support there view, I would suggest you read my last post. Yougov statement is a fact, one that I think you should be giving cognisance to. Every single post to the contrary basically asserts they know better than the actual professional pollsters, it really is that simple.
It boils down to opinions being given the same weight as stated facts by YouGov thereby creating a false equivalence, of wiki editors stating they know better than YouGov about YouGov polls, it really is rather laughable. Soosider3 (talk) 19:09, 11 December 2023 (UTC)[reply]
By that same logic, we should not be putting say Ipsos polls in the same table as YouGov polls due to known methodological differences, which would defeat the entire purpose of articles such as these. Nowhere in YouGov's statement does it recommend excluding them from polling aggregates, of which this article is. Please note the fact that every editor in this section bar yourself is interpreting YouGov's statement in a certain way. GamerPerson12345678 (talk) 16:32, 20 December 2023 (UTC)[reply]
lets not go round again, my position re scoop remains the same Yougov state not to compare with other yougov polls, introducing them into table, imho, invites readers to do exactly that especially as there is no facility for a note to give readers appropriate information. I am well aware I am the sole voice with this view, however that doe snot make me wrong merely outvoted by by concensus.
I note you reverted my figures on this poll, I would like to know why? my data is taken directly from the published data where as yours come from a blogger BBS, I would suggest that tables from teh actual poller are the best and most reliable data and would ask that you actually read the tables before reverting again. Soosider3 (talk) 17:40, 20 December 2023 (UTC)[reply]
The numbers direct from the pollster do not add to 100% (likely due to don't knows) and need to be adjusted to match the other polls. I believe this is what BBS did. Eastwood Park and strabane (talk) 20:37, 20 December 2023 (UTC)[reply]
Nonsense , we do not adjust the numbers from pollsters we record them exactly as the pollster reports them in their data tables. The scoop polls do not record likelihood to vote and captures data including not voting, it's why their numbers are so different from other YouGov polls which is probably why YouGov said not to compare them. In no way do we 'adjust' pollsters numbers, mind you it would be helpful if we had a notes column that allowed some context to be indicated. As for adjusting do we do that if 16/17 years olds are not recorded, do we do that if likelihood to vote is not included- of course we don't and neither we should. Report what pollsters find, just the facts and nothing else Soosider3 (talk) 20:47, 20 December 2023 (UTC)[reply]
As a matter of practice these articles include headline figures with DK/won’t vote removed. I’d appreciate if you came to terms with the fact that you’ve been constantly outvoted on this issue, along with the substance of the SCOOP polls, and stopped the constant edit warring. This is not a productive use of editor time and I’d guess that passing mods would take a dim view of your actions. GamerPerson12345678 (talk) 02:48, 22 December 2023 (UTC)[reply]
Had a further look at the table, there are rows that don't add up to 100% which may well reflect the variety of manners that pollsters report data, but I also see ( after a few random samples) examples where the numbers do not match those of the pollsters. Have folk actually been 'adjusting' the pollsters data to match the table? If so this is unacceptable, we do not adjust data to fit table we adjust the table to match the data. Also had another look at the BBS article you referred to, interestingly he is very clear that he was following Yougov advice and would not include scoop polls in his tracking but would report on them separately, he also stated in article that it was difficult to make exact comparisons because the pollster only reported % and not actual numbers. I think you need to be very careful not to be cherry picking the bits you prefer. Perhaps we need to look at the layout of the table, I am all for brevity and a clean tidy look but perhaps we need a few more columns to capture more accurately the variety of ways pollsters present their data, including some of the minor parties reported on, Don't knows(undecideds) but also a note column to highlight variations. That would give us scope to capture the data without feeling we had to adjust it, done properly it would give readers a better more comprehensive look at polling information. Soosider3 (talk) 21:32, 20 December 2023 (UTC)[reply]
The polls should be included in the main section in my opinion, but should probably be differentiated from other YouGov polls, as per what YouGov said about them. Possibly add a note to them, or naming them as YouGov (Scoop) rather than just YouGov? Eastwood Park and strabane (talk) 00:06, 9 December 2023 (UTC)[reply]
In addition, the most recent Scoop poll numbers appear to include don't know or some other response - adding all the percents together gets you to 73%. Edit: This appears to be the case for all of them. Eastwood Park and strabane (talk) 18:33, 10 December 2023 (UTC)[reply]
Yeah, imagine that SCOOP polls are different from the usual yougov polls, who'd have thunk it, well appart from Yougov who issued a statement along with the data - remember that!!! oh what a tangled web we weave ........... Soosider3 (talk) 06:57, 21 December 2023 (UTC)[reply]
Calculating percentages with undecided and refused removed feels within scope of WP:CALC to me, and if a secondary source has done it all the better. Including a separate table with them removed is not a profound methodological difference and there is no need to relitigate your misunderstanding of a YouGov statement which has been discussed to death already. Ralbegen (talk) 18:13, 21 December 2023 (UTC)[reply]
That is some stretch Wikipedia:CALC is a subsection of No Original Research, there is a stronger argument that what you actually do is original research and therefore has no place on wikipedia. WP:CALC is for simple straightforward calculations " Basic arithmetic, such as adding numbers, converting units, or calculating a person's age, is almost always permissible." What you do is far beyond what any person could regard as 'Basic Arithmetic'
It is also noteworthy that there has been no discussion in the article on your preference, none what so ever, in fact it is noticeable that you seem to have taken it on yourself to 'amend' the numbers with little if any explanation.
An ordinary reader coming to this article would have no indication of what is actually on display within the table and would most likely take the data presented as being an accurate record of the polling companies data.
Wikipedia's policy of just the data is what we should be following, if your idea is to prevail then we need to have a very clear statement right at beginning of article as to what is happening and why and a method of indicating when these measures have been applied. This would be the least required to ensure transparency. Soosider3 (talk) 13:02, 22 December 2023 (UTC)[reply]
It does mention however that routine calculations aren't considered original research, provided there is consensus. If excluding don't knows for consistency is routine for this article then calculating numbers to do so wouldn't necessarily break WP:CALC. Eastwood Park and strabane (talk) 19:11, 22 December 2023 (UTC)[reply]
It does not mention routine it states basic arithmetic, what you do is far beyond what any reasonable person would consider basic arithmetic. It therefore breaks would:calc Soosider3 (talk) 19:46, 22 December 2023 (UTC)[reply]
I would say that including don't knows as a separate category wouldn't be unprecedented, such a category exists at Nationwide opinion polling for the 2024 Republican Party presidential primaries and Opinion polling on Scottish independence. It would probably need to be consistent across UK election articles though. Eastwood Park and strabane (talk) 19:16, 22 December 2023 (UTC)[reply]
It would be I think a more sensible way to present the data, however we would need some clarity about don't know. There seems to be potential 2 groups, those who have said they will vote and are undecided and those who don't know. If we included a column for notes that would allow identification of many of of the variables such as no 16/17 year olds, no weighting for likelihood to vote and any other issue, perhaps it would clarify for the reader what they are actually seeing. It would feel to me like a better way of going rather than trying to shoehorn data into the table we get the table to reflect the at times wide variety of how pollsters do their work. Soosider3 (talk) 19:54, 22 December 2023 (UTC)[reply]
Addition and division are usually considered basic arithmetic. Secondary sources doing those for us helps too. Partisan VI polls in the UK are routinely reporting with undecided and refused respondents removed as the headline and it would be inappropriate to move away from this practice as a general rule, and using secondary reporting of poll results that provides figures in the usual format is perfectly acceptable. Calculating results with undecided and refused voters removed is straightforward and unambiguously produces the displayed results that editors can reasonably do, though when editors do it rather than from a secondary source it would probably be worth including a note about it :-) Ralbegen (talk) 20:01, 22 December 2023 (UTC)[reply]
That is not correct and actually flys in the face of what WP:CALC actually says, manipulating the data in this manner most definitely makes it original research which maybe why WP:Calc is a subset of original calculation. Now I am sure we will have some mental gymnastics to make your desired out come fit, but the fact remains you choose to use secondary and less reliable sources of data over a higher and more reliable data source and compound that by not making it clear and transparent what you are actually doing. I think I could generate a lengthy list of WP guidance that goes against. Soosider3 (talk) 05:15, 12 January 2024 (UTC)[reply]

Edit wars[edit]

Can we all please end the constant edit war that has erupted on this page over the SCOOP figures? First with whether the polls should be included in the first place, and now with whether or not to include or exclude DKs and refused from the overall figures presented in the data table. Whilst I'm grateful that we've finally ascertained that SCOOP polls are to be included in the data tables, I'd urge all editors to follow the accepted guidelines for Scottish VI tables and to remain consistent with previous polls where VI figures are reported in the data tables excluding participants who voted Don't Know or who refused to answer, as seems to be consensus on this page and on others where SCOOP polls have been reported. GamerPerson12345678 (talk) 22:25, 26 December 2023 (UTC)[reply]

it is regrettable that you choose to see this as an edit war rather than a genuine attempt to have a consistent, transparent approach to capturing Polling data. ps I would be grateful if you could refer me to the "accepted guidelines for Scottish VI tables" you refer to.
What in effect is being recoded in the table is not actually polling as produced by the actual polling companies ( the best source of data) but rather "Headline" figures generated by Newspapers and other media ( a less reliable source) and also generated by some editors. There is clearly consensus for this practice however it throws up some issues.
1 The article is called "Polling for .....", should it not actually be called "Headline figures for..."
2 There is no indication of the true nature of these numbers, it is not un reasonable to assume that an ordinary reader will think numbers are taken directly from Polling Company Tables
3 There is a significant variation on how these numbers are arrived at and no notes field to provide explanation
I would suggest that
A) We introduce a sentence at the beginning of the article to explain what is actually being shown.
B) That we have a Notes column to highlight which numbers are being used and how arrived at. Soosider3 (talk) 14:49, 2 January 2024 (UTC)[reply]

Capturing Details of other[edit]

I have been thinking about how we might capture more detail of others recorded by some pollsters, obviously we could add extra columns but that seemed excessive. I came across this feature on another site and have applied it to the Ipsos Poll others for Constituency poll. Visually it shows % but with a click on 'show' for further details Would welcome views of other editors Soosider3 (talk) 05:22, 8 February 2024 (UTC)[reply]

Retrieved from "https://en.wikipedia.org/w/index.php?title=Talk:Opinion_polling_for_the_next_Scottish_Parliament_election&oldid=1209531332"