Talk:Opinion polling on Scottish independence/Archive 5

Archive 1

←

Archive 3

Archive 4

Archive 5

Proposed updated leading text

I'd propose to replace the current paragraph:

"Polls in the main table, using the same question, still show systematic differences between different polling organisations. The other tables reflect different ways questions around independence can be asked, which may produce different results."

with these two:

"Polls in the main table, using the same question, show systematic differences between different polling organisations. This is sufficient to lead to polls carried out at the same time by different organisations to show statistically significant differences. The body of polls from one organisation may give a completely different picture to those of another. Further, it is open to clients to select their pollster based on these differences.

The other tables reflect different ways questions around independence can be asked, which also produces different results."

I will leave this for a few days to gather comments before editing. RERTwiki (talk) 16:51, 13 July 2023 (UTC)

Broadly agree, and recognise that this is something you clearly feel strongly about. However we need to be wary of assigning motivations to clients, especially as this could be assigned to just about most of the pollsters. Variation from mean I think demonstrates this most clearly. FON do seem to be getting a lot of the focus here but perhaps can we seriously say that other pollsters such as Redfield & Wilton or Yougov are much better? and if so only by degree. It seems to me that it is the nature of polling that different companies produce slightly different results, this phenomenon is well known and recognised by many. The relative low number of polls conducted on Independence may be throwing this into sharper relief.

Perhaps something to guide the readers, that it is best to generally compare a poll with polling from the same company.

"Polls in the main table, using the same question, can show systematic differences between different polling organisations, therefore it is probably best to compare a poll with those previous ones from same pollster" Soosider3 (talk) 18:12, 13 July 2023 (UTC)

I’m less certain about Redfield and Wilton but YouGov are (currently) in a completely different league to FON when it comes polling. Dunk the Lunk (talk) 21:38, 13 July 2023 (UTC)

Personally, I’m against the existing wording and the proposed wording. On this specific page, which is about presenting data, I think we should absolutely minimise the potential to lead the reader into any form of interpretation, especially before they’ve even got to the tables. Perhaps a better way might be just be a quick sentence to link the reader to the section on ‘Potential for inaccuracies’ on the ‘opinion polling’ page as well as the ‘selection bias’ page which appears to be a potential issue with FON polling. Dunk the Lunk (talk) 21:31, 13 July 2023 (UTC)

I like the sentiment behind Soosider's suggestion: this is what we have to do to understand a poll's results. How about:

"Polls in the main table, using the same question, can show systematic differences between different polling organisations. Therefore to discern trends it is helpful to compare a poll with previous results from same pollster. The other tables reflect different ways questions around independence can be asked, which also produces different results."

I have to disagree with Dunk the Lunk: this is a minefield, and I think it's not responsible to invite people to walk into it and read the list at face value as comparable data.

As regards redfield/yougov/FON : yes it is both a matter of degree, but also somewhat a zero sum game. If one pollster is biased one way from the average, that will tend to bias everyone else the other way. I largely agree that there appears to be a no bias to Redfield, a small no bias to Yougov, and a Yes bias to FON and IPSOS. The Yougov bias might be definitive even though small because the sample of Yougov polls is larger. I haven't really done enough analysis to make those views firm conclusions.

As before, back in a while to see what people think. RERTwiki (talk) 10:13, 27 July 2023 (UTC)

I do understand your point. I do just worry about leading to much but if you and soosider are happy to agree to suitable text I won’t delete it! Dunk the Lunk (talk) 17:47, 29 July 2023 (UTC)

Just went for it, in the end I think we're in agreement. RERTwiki (talk) 11:39, 31 July 2023 (UTC)

Shortening preamble

In an attempt to reduce 'white space' and scrolling to get the actual polling information, I would suggest amending the preamble as below

"Opinion polling on Scottish independence has continually being carried out by various organisations since the 2014 Scottish independence referendum. Polling conducted before the 2014 Scottish independence referendum can be found here. Most polling companies listed here are members of the British Polling Council (BPC) and abide by its disclosure rules. The main table includes primarily those polls which ask the same question as the 2014 referendum: "Should Scotland be an Independent Country?". Any variations that might have an impact on the poll result, such as excluding 16- to 17-year-old voters, are recorded in the 'Notes' column. Polls using the same question, may show systematic differences between different polling organisations. Therefore to discern trends it is helpful to compare a poll with previous results from same pollster. "

Thoughts and suggestions appeciated Soosider3 (talk) 11:05, 7 October 2023 (UTC)

I don't see any need to change this. See above. RERTwiki (talk) 16:50, 8 October 2023 (UTC)

Feeling conciliatory this morning, I've shortened the preamble as much as I can while retaining the meaning. Hope this is helpful. I think the result is very similar to your paragraph above, Soosider. RERTwiki (talk) 11:21, 18 October 2023 (UTC)

White space and scrolling to get to polls

I had been somewhat concerned at the amount of scrolling required to get to the meat of the article - the polls, with the preamble, Index and Graphs there seemed a long way down to the polls. The preamble is probably as short as we can get it, the index is required for navigation of such a large article. Therefore I looked at the graphs and think reducing them to a thumbnail might be the way to go, it still has them in a prominent place and gives reader option to click in for more data while shortening the effort to get to the polls. Therefore I have amended there layout to that of thumbnails - which certainly shortens the trip to the polls Soosider3 (talk) 11:00, 15 September 2023 (UTC)

I’m normally on the mobile version so it personally it doesn’t bother me-the sections are all collapsed when I open the page. If the graphs are reduced to thumbnails that might depending on how they interact in the mobile version? Dunk the Lunk (talk) 15:21, 15 September 2023 (UTC)

Just checked and the thumbnails make no difference to mobile view, Id like to say it was down to foresight and planning but it was a fluke. Sometimes better being lucky than good Soosider3 (talk) 07:39, 16 September 2023 (UTC)

😀 Dunk the Lunk (talk) 11:34, 16 September 2023 (UTC)

Have to say I kind of disagree: the thumbnail is almost invisible.

I think that on the mobile version, it's marginally better (more easily visible) without being a thumbnail, and doesn't really impact navigation, as all the sections start collapsed.

On my PC, I'll have to click further to see the charts if they are thumbnail. Also, the click takes me to the page for the object, not an image.

I think the former layout was better for people reading the article on a PC, slightly better on a mobile.

For editors scrolling down on a PC to the table, the new format saves a click.

I think readers are probably more important than editors (unless there is evidence no b***** else reads this page!).

I'd vote to return to the old format. What do you think? RERTwiki (talk) 09:51, 22 September 2023 (UTC)

I am far from an expert on publishing to this site, but like you had noticed that clicking on thumbnail opens object. So have looked at a few other sites that use graphics similar and they seem to open image with an x that gets you back to article, could it be something about how you are publishing these or the format used to publish.

Have a look and see what you think

https://en.wikipedia.org/wiki/Opinion_polling_for_the_next_Scottish_Parliament_election

https://en.wikipedia.org/wiki/Polling_in_Scotland_for_next_United_Kingdom_general_election#Polls Soosider3 (talk) 10:41, 23 September 2023 (UTC)

Also, the format of the table changed when the graphs were changed to thumbnail, putting the table in a smaller space to the left of the thumbnail, making most entries two rows, and...increasing the scrolling to get into the table... RERTwiki (talk) 10:06, 22 September 2023 (UTC)

Sorry but changes have reintroduced the amount of scrolling for readers, which means a lot of white space before getting to meat of article. I am not at sure if graphs require such a prominent position? Perhaps we should fully reinstate them but after the polls section? Soosider3 (talk) 10:29, 23 September 2023 (UTC)

Two reactions to that. First is that I think the graphs carry a great deal of information in summary form, and should be up-front. They've worked well for years, and I don't really think they should be demoted below the table. The graphs if I recall correctly were introduced because the text was prone to drifting into narratives about trends completely unsupported by the data, and the graphs really prevent that.

Second is that we ought to get to the bottom of why you're having problems with scrolling. I have no issues with the current page either on Windows 10/Chrome or on an iPhoneSE. What exactly is it you're seeing? RERTwiki (talk) 16:11, 4 October 2023 (UTC)

As regards the first point, I agree the graphs do carry a great deal of information, however IMO they are there to support the polling information not to usurp it.

I had hoped we had found a reasonable compromise when they appeared as a thumbnail, however it would appear that is not sufficient for your point of view, which also appears to be frustrated by the glitch that makes them appear in the manner they now do ie directly to source rather than as a pop up. far from being an expert but it does seem that your graphs are distinctive in having this glitch and am assuming that is due to how you are publishing them- could you explore that avenue? and advise if you can make progress with it.

Re Second point, not exactly sure who the "we" you refer to is. However it is very clear that there is at least 2 screens of scrolling to get to the actual polls, polls that are the very reason for the existence of this article and really should be the first thing available to any reader.

My understanding is that wiki require preamble and index

There needs to be a preamble - but perhaps it could be shortened,

There needs to be an Index - this one could only be shortened by removing or merging some of the other topics.

There is no requirement to have a graph in such a prominent position. I still think the idea of a thumbnail is the best solution available. Soosider3 (talk) 11:21, 7 October 2023 (UTC)

Firstly I'm still puzzled by your suggestion that scrolling is an issue: I genuinely don't understand.

On a phone, all sections are compressed initially. The format of the graphs has no bearing on the amount of scrolling. And the tables are reached with one or at most two flicks on a finger. I don't see an issue there.

On a PC in full screen mode, the tables are a single click from the top.

I don't see any problems that need to be addressed. Further, I think that the utility of the graphs has been greatly decreased by putting them in as thumbnails, and I'd like to reverse that change and go back to where we were a few weeks back. (I did investigate the publishing format, but haven't yet found a way to get from excel to a more wiki-friendly format.)

Can you give more context on what difficulties you are having with 'scrolling'? RERTwiki (talk) 16:59, 8 October 2023 (UTC)

Hi. First, I've looked for a way to convert my PDFs to something which will expand better from a thumbnail. No luck so far.

Second, I disagree with your implicit description of the use of this page. I don't think people come here for a table of polls with minimal preamble. People come here to understand the current state of polling. To that end, the graphs are vital: nothing else gives any idea of recent trends, historical context, or discrepancies between polling organisations. Without that information, individual polls can be spun out of context into an exciting story, when the reality is one of punctuated stasis, for want of a better term.

The lack of context, and the flaws that flowed from it, were the reason for the introduction of the graphs - I think around 2017, but might be wrong.

For the moment, I'd like to put them back, perhaps a little smaller, as centered rather than thumbnail, and continue the discussion and search for viable alternatives to pdf. RERTwiki (talk) 11:35, 18 October 2023 (UTC)

Aye I tend to agree, that the current format has fallen between 2 stools. Neither the better clarity of original or a reduction in white space/scrolling before the polls.

Its not ideal but the original format is preferable to whats there at moment. Soosider3 (talk) 21:07, 1 November 2023 (UTC)

Opinion Poll Graph

The current version of the graph tracking opinion polling is illegible to colourblind people (like me) because the colours used for Yes and No appear identical. I don't know what colours they're supposed to be, but could either one be changed to something obviously different, like blue, black or deep red? Eoghankll (talk) 16:32, 1 November 2023 (UTC)

We have been trying different ways to present the graphs and tried a few different layouts, that may have exacerbated your issue. As someone who has a defect in my colour vision I fully understand. I recall earlier this year there was a talk topic about the colour in the article but at that time had not looked at the graphs. Sorry to say but I think the topic rather petered out.

The graph author is probably going to reinstate an older clearer version of the graphs, could you give us feedback if that is any easier for you to make out.

Other thoughts would be to maintain the colours but make the lines different, perhaps having one dotted and with different shapes for teh cardinal points. @RERTwiki Do you have any thoughts on this? Soosider3 (talk) 21:14, 1 November 2023 (UTC)

Seems like making the charts visible to people who are colourblind would be both worthy and do-able. I’ll take a look at next update.

Ideally any colour-scheme change would be reflected throughout the article, where green and red are used ubiquitously with the same meaning as highlight colours.

I guess there will be some advice available on choice of colours with this in mind. I’ll have a look, try something, and put up my reasoning for comment. RERTwiki (talk) 22:30, 1 November 2023 (UTC)

There was some discussion re colour vision defects back in May

This application is useful for checking page against various forms of colour blindness https://www.color-blindness.com/coblis-color-blindness-simulator/ The key appears to be around the Hue of the colour and having secondary indicators, such as Text, shading or Cell out line. Let me know if I can help with anything Soosider3 (talk) 06:07, 2 November 2023 (UTC)

Glad to see this is getting some traction. There is also a section on the Commons about suitable colours (probably elsewhere too)

https://commons.m.wikimedia.org/wiki/Commons:Creating_accessible_illustrations Dunk the Lunk (talk) 07:32, 3 November 2023 (UTC)

I had thought of yellow (yes) and blue (no) which have obvious associations. The link seems to show these are Ok for colour-blind, and specifies exact tones. I’ll try that. I’ll avoid patterns for now, since its very hard to make those visually neutral. RERTwiki (talk) 21:09, 3 November 2023 (UTC)

I tend to think that no matter what colours might be used someone somewhere will find an association, as long as we are trying to tackle this issue then these should really be of secondary consideration. If you could share the exact tones (hexadecimal I think) then I'll have a look at transposing these into body of article, that might be a bit trickier as we use shades a fair bit, but at least we have a starting position.

Rather suspect we might have several options and had thought that perhaps at appropriate time we create a new temporary section at bottom of article to test them out and ask readers views. There are so many variables to contend with both technological ( ie monitor variations and graphic cards) as well as the wide variation of colour impairments there can be. Soosider3 (talk) 11:40, 4 November 2023 (UTC)

Had a look at the wiki page re colour vision

Thought occurred that rather than introduce new colours and possible open new avenues for disagreements we use the Blue Green and vermillion on the wiki colour palete, its close enough to what has been long established but moving away from traffic lights (ish)

Would be fairly easy to do using find and replace

I have temporarily added test section at bottom of article (only did a few rows to give us a feel) where I have changed the colour of Column headers and lead, these seemed to be the main areas where colour vision defects would be most likely to have an impact. The other shading seems to hold up fairly well and perhaps we could live with it.

Whats folks thoughts? Soosider3 (talk) 14:37, 4 November 2023 (UTC)

Blue for Yes and Red for No would be really good. It would line up with the map used in the Scottish independence referendum, 2014 article and suit the respective colours used by Yes Scotland and BetterTogether. AlloDoon (talk) 14:35, 7 November 2023 (UTC)

That's an interesting idea all the more so as it is a rationale position to take, I will try and do some work on it this afternoon by finding the hexadecimal for both Yes and Better together and trying them out in the test area Soosider3 (talk) 10:42, 8 November 2023 (UTC)

OK. I updated the chart, switching to blue-green and vermillion from the original colours. It's the smallest change which might do the job and has been suggested above by Soosider. Would be nice to have feedback from EOGhankll.

If we use these as the colours in the tables, We also need to choose 4 other colours: yes/no highlights for over 50%, and yes/no for the margin column. I suggest we cut these down to two, and use the same colour in both cases.

I'll see if I can try some in the test section. RERTwiki (talk) 22:52, 7 November 2023 (UTC)

Good Morning thanks for that, however you appear to have used the Reddish Purple rather than the vermillion, a typo? or different thoughts, if latter please share your thinking. For me the Reddish Purple appears a bit pinkish and begins to be difficult to see Black text on a Red(ish) background.

We seem to be moving towards Yes/No column headers and Lead highlight as being the same, I think that makes sense and simplifies the process, and I am all for simple. Thinking about the highlighting in the cells I begin to think we are making this harder than we might need to, perhaps all we need is to embolden the text in the cell and leave it at that, we have separate columns for either so not sure what adding colour adds, there is also the coloured Lead 2 columns to the right. Soosider3 (talk) 10:39, 8 November 2023 (UTC)

Okay, have wrestled with Test Area as it was behaving very odd and unstable but got there

Have added new table with colours as per Rertwiki as well as amending my original idea Bluish green/Vermillion also in that test table removed background colours for cells inside Table leaving them just Bold

Have included a 3 rd test table which I intend to try our AlloDoon suggestion (but later today as I must get off this keyboard) Soosider3 (talk) 12:19, 8 November 2023 (UTC)

Correct, my mistake, I used reddish purple as similar to the existing red and wrongly called it vermillion. RERTwiki (talk) 15:03, 8 November 2023 (UTC)

Looked at the new test section. I know this is just a personal opinion, but I’m seeing the vermillion as somewhere between s*** brown and dihorroea coloured. Really wouldn’t be comfortable with associating ‘no’ with that colour…!!!

I think the chart colours (blueish green/reddish purple) are both nice strong colours with positive, somewhat a-political associations. Experimenting last night there is quite wide scope for admixing black and/or white with the base colours to keep the same hue and give scope for emphasis colours. I’ll fiddle some more when back at my desk.

I think narrowing the pallet is a good idea, I agree that simple is good. RERTwiki (talk) 15:17, 8 November 2023 (UTC)

Okay. Thats interesting as I see the vermillion as almost an orange and the reddish purple as a fairly lurid pinkie colour. Perhaps just emphasis the issue we are trying to resolve ie not everyone sees it the same and so many technological variations that can all have an impact.

Have added the 3rd test table using the colours suggested by @AlloDoonAlloDoon Was easy finding the Yes Blue the Bettertogether Red was more variable, but have had a stab at it, at least for the Column headings and lead entries - not sure about it but will test out in Colour Checker thingy Soosider3 (talk) 16:34, 8 November 2023 (UTC)

Arghus, some bot has reverted the test area giving me a wee lecture re using sandbox to practice. I have responded and will revert asap Soosider3 (talk) 18:59, 8 November 2023 (UTC)

Kindly use sandboxes to practice editing without causing disruption to the article mainspace. Thanks and happy editing. Definitely not a bot, lol xP The Herald (Benison) (talk) 19:04, 8 November 2023 (UTC)

Glad to hear your not a bot an easy mistake to make. Could you elaborate as to what disruption it is causing to article and weigh that against the potential benefits Soosider3 (talk) 19:23, 8 November 2023 (UTC)

With good faith assumed for your edits onto the article mainspace, there are certain guidelines to be followed when you are creating and expanding an article. They are mostly enlisted in WP:MOS and MOS:LAYOUT. These must be followed in order to maintain consistency of articles throughout the Wikipedia. Your test edits, beyond the external links subsection, are clearly test edits. Hence, to give the benefit of the doubt, you were given a level 1 warning for test edits and the changes were reverted back to the old version. It is totally okay to learn the editing and syntaxes when you are new to Wikipedia. But, you can use sandbox pages to practice editing. Kindly go through THIS page to learn more about that. Since your previous edits were a deviation from these guidelines, but did not, in a strict sense, constituted vandalism, I labelled them off as disruptive editing, assuming good faith. Do not worry, just follow the guidleines and use sandbox pages to practice editing. Thanks and happy editing. The Herald (Benison) (talk) 22:38, 8 November 2023 (UTC)

Firstly I am not learning editing, a few of us editors were merely trying to find a good way to share the visual changes required to make article better for folk with a defect in their colour vision, to reach consensus before implement such a large scale alteration to the whole article.

A good faith edit of the highlighting colours would entail a change on every line in the polling table which runs to several hundred, a fairly large task, better to have some consensus before expending so much energy. To be clear this is not about changing text or data but rather an overhaul of the look of the complete article, colours, highlighting. However if thems the rules then of course I will respect them. Soosider3 (talk) 22:51, 8 November 2023 (UTC)

There's always new things to learn. I'm still learning after a decade of editing. As I was looking through your contributions, your edits, along with others involved, looked very much like test edits. But please keep those test edits out of article mainspace. Use talkpages and sandboxes for that. Everyone involved in this project can use WP:Teahouse for any queries and questions regarding it. Thank you for taking the initiative, but if you are doing it, do it right. Thanks. The Herald (Benison) (talk) 04:05, 9 November 2023 (UTC)

A number of perfectly valid updates to the page appear to have been lost with this edit back to an earlier version, including an entire poll entry in the main table. Dunk the Lunk (talk) 13:04, 9 November 2023 (UTC)

I have a copy of the table before I changed it, took complete copy of table to give me something to work on in my Sandbox,let me know what's been lost and i'll reinstate. Soosider3 (talk) 15:57, 9 November 2023 (UTC)

You can go through the page history to recover any lost data easily, along with syntax and wikimarkup. Click on the revision and go to edit tab. Then copy and paste on to your sandbox and resume editing. The revert was made on to the last stable version with no test edits beyond the external links section. Toodles. The Herald (Benison) (talk) 16:51, 9 November 2023 (UTC)

or perhaps you could have just deleted the offending section? Dunk the Lunk (talk) 17:28, 9 November 2023 (UTC)

Colour Vision amendments

Started new section to hopefully pull our thinking together, here is my understanding of where we are at We need to change some colours/highlighting to improve access for folk with a colour vision defect. During the course of that discussion we seem to have agreed on three specific tasks 1 The colour of the Yes/No column header 2 Use the colours from 1 for highlighting in Lead column 3 Felt we were moving towards only having embolden in the cells with in the table ie no colour or shading

Had a look at using colours from Yes and Better Together, while logical it looked as if we were only going to swap one problem for another, it seemed that we would still leave some folk unable to see them clearly.

Using the Bluish Green for Yes seems acceptable to most and is completely good on the colour checker as far as viewability goes for Colour Deficit folk

WE need to reach some consensus of what colour to use for No, it would appear we have identified 2 options, either Vermillion #D55E00 or Reddish Purple #CC79A7, if we can agree on this then I believe we could go ahead and make the changes in the main table, which I would be happy to do.

My tuppence worth is that the Reddish Purple is rather garish and would prefer the Vermillion Soosider3 (talk) 23:15, 8 November 2023 (UTC)

Have amended main table in article to be more colour vision friendly, taken my lead from previous discussion. Coding was a bit of a pain as there were several shades used for both Red and Green, at the very least the coding re colour and highlighting is now consistent.

Used Bluish Green #009E73 and Vermilion #D55E00, removed all colour from the cells within the table and left the lead emboldened.

Given that our test area was taken away from us I believe we need to move this issue forward with good faith edits. Soosider3 (talk) 13:32, 9 November 2023 (UTC)

The colours are really jarring. I think we should go for blue/red and keep it consistent with the 2014 Scottish independence referendum colours. Well done for all of the effort though Soosider. AlloDoon (talk) 14:08, 9 November 2023 (UTC)

I also think we should retain boxes being coloured for yes/no AlloDoon (talk) 14:09, 9 November 2023 (UTC)

You had suggested that previously and I had looked for the Yes blue and Bettertogether red but found they failed the colour blindness test thingy.

I think the table looks better and clearer with no colour in the cells, the table data is only 4 columns wide, it really does not require further emphasis than to embolden the lead. Soosider3 (talk) 15:08, 9 November 2023 (UTC)

Had closer look at table using colour tester, a couple of observations

1 There may be an issue in using Black text on Red background (No emphasis in cells especially for absolute lead)

2 The shading highlight for Yes lead is so pale as to be almost invisible (that's across a number of colour vision defect types)

3 If and I say if we are to retain the colour highlighting in the cells then we require to look at the 4 shades we will require for this test them for task in hand and then implement them Soosider3 (talk) 16:07, 9 November 2023 (UTC)

Thanks for all the effort Soosider, the table is coming on very nicely. It's certainly not easy to get something that's not jarring, but meets the aim of being friendly for all users! Could the reds be lightened further so there is a greater contrast with the black text, and perhaps the blue overall lead darkened slightly so there is a greater contrast between the two blues? Would that work? Dunk the Lunk (talk) 17:25, 9 November 2023 (UTC)

I was thinking something similar will have a look later this evening but wanted to check others thinking first Soosider3 (talk) 17:29, 9 November 2023 (UTC)

What colour tester are you using Soosider? I tend to use https://colororacle.org/ and the newly added colours work for all major forms of colour blindness. AlloDoon (talk) 21:51, 9 November 2023 (UTC)

Absolutely agree there has been significant progress, its definitely coming together if nothing else we have tidied up the coding and the various shades used throughout the article.

I've been using on online tester at https://www.color-blindness.com/coblis-color-blindness-simulator/ Having said that I have also been looking at the article on other devices in particular my iPhone, it seems to reinforce something I said earlier that the variations may not be all down to Colour vision problems and that the technology can play a role too, ie graphic card, monitor and software can all impact on what is actually seen. For example what I can barely make out on my desktop is clear on my phone, that rather suggests to me that we try to stick to the advice given about colours to use that are best for colour vision rather than what might be more aesthetically pleasing.

With that in mind today I will be looking at the 4 shades we need for the highlighting in cells (someone in background mutters about only using bolding and no colours) and see how I get on. Soosider3 (talk) 09:35, 10 November 2023 (UTC)

Had a look and did some further reading on colour vision etc a couple of things kept coming up, that Red/Green is the worst for colour vision and that 100% the best is black/white. So if we needed it we have been completely correct to move away from previous Red/green, but we are unwise to rely on too much colour. Therefore I think I am going to take a stance on no colour in cells, it is unnecessary and makes article less accessible. I can live with the Blue/Red (perhaps a slight change in the hue of both) as column headers and Lead but we do not require any more than that, for heavens sake the table is only 5 columns wide why would we need it, only 3 of the cells will ever have emphasis every row will have 2 of these 3 cells emphasis - that's overkill. Soosider3 (talk) 11:01, 10 November 2023 (UTC)

I would agree with that. I think we need to balance accessibility for colour blind users with what is accurate for Yes/No and somewhat visually workable for most users. The current colour scheme currently works should be visible for >99.9999% of users, with the only exception being those with Achromatopsia. AlloDoon (talk) 11:07, 10 November 2023 (UTC)

Taking out the coloured boxes is bad and inconsistent with other opinion poll articles in my opinion AlloDoon (talk) 17:35, 10 November 2023 (UTC)

Taking out the coloured cells improves the visibility of the data to a wider range of readers which is surely the whole point of the exercise. Soosider3 (talk) 18:12, 10 November 2023 (UTC)

YouGov / SCOOP Poll

I cannot understand why YouGov's recent SCOOP poll is being excluded from this article when YouGov are a member of the BPC and the poll is listed on Professor John Curtice's website. @Soosider3 could you please explain the rationale.

I would welcome others opinions on this. AlloDoon (talk) 14:06, 9 November 2023 (UTC)

The fact a YouGov/SCOOP poll from November last year which had Yes ahead by 1% has been on the table means the burden is on you to prove why the poll should be excluded. AlloDoon (talk) 14:41, 9 November 2023 (UTC)

The reason is simple Yougov themselves have said that because of different wording and none use of their usual weighting re VI then scoop polls should only be considered against other scoop polls and not standard polls. See this from yougov re there Feb 20023 scoop poll (also not in our article) see very last paragraph. https://yougov.co.uk/politics/articles/45269-nicola-sturgeons-legacy-according-scots

Perhaps the solution is that we have a section especially for scoop polls, at the very least we should remove Nov 2022 scoop poll for consistency Soosider3 (talk) 15:06, 9 November 2023 (UTC)

Using slightly different methodology to regular YouGov polls does not invalidate the poll. The note states: However, the voting intention results used slightly different wording and did not include YouGov’s standard turnout weighting and so should not be directly tracked to other YouGov voting intention figures as they are not identical. Instead, they should be tracked to other SCOOP voting intention polls conducted by the Scottish Election Study (latest wave conducted in November 2022). The latest YouGov figures voting intention poll using its standard methodology (17-20 February) can be tracked to our latest poll conducted 23-26 January. This difference in methodology may have contributed towards the difference in figures, though to a large extent this is likely to be just the normal sample variation within the margin of error. AlloDoon (talk) 16:16, 9 November 2023 (UTC)

The statement could not be clearer or more specific, Yougov state "......so should not be directly tracked to other YouGov voting intention figures as they are not identical. Instead, they should be tracked to other SCOOP voting intention polls conducted by the Scottish Election Study " that could not be clearer I am baffled as to your position on this matter. Soosider3 (talk) 16:24, 9 November 2023 (UTC)

I have to say I totally disagree with this. Of course this poll should be included (I assumed it had got lost in the heralds blanket revision (what a patronising person btw). Just because YouGov are using a different methodology from usual doesn't invalidate it - YouGov would surely also say that their polls can't be directly compared to any other pollsters, would they not? The only thing that bothers me so far is I haven't seen a published link yet other than on What Scotland Thinks (which has different figures to what has been used on here). Dunk the Lunk (talk) 17:17, 9 November 2023 (UTC)

I also agree that we should strongly favour including all polls, particularly from organisations which have previously been included. If we exclude a particular poll we seem to be acting on whim. If we edit history for consistency we seem to be from the ministry of truth.

The range of results from apparently honest polling methods isn’t narrow enough for us to exclude a poll based on nuances of how weighting has been done or questions asked. (And we already have a protocol for dealing with Non-standard questions).

So please, keep it in. RERTwiki (talk) 09:09, 10 November 2023 (UTC)

Even when the pollster concerned clearly and specifically states that scoop polls should not be compared to their other polling? that would be misleading to readers many of whom will not have your experience in interpreting and understanding polls. Soosider3 (talk) 09:39, 10 November 2023 (UTC)

Read the note again. The pollster says: "This difference in methodology may have contributed towards the difference in figures, though to a large extent this is likely to be just the normal sample variation within the margin of error."

There is no reason to suggest the poll is not compliant with standard BPC weightings, the questions and methodology were just very slightly different than their other polls, as is the case across BPC members. AlloDoon (talk) 10:38, 10 November 2023 (UTC)

Respectfully, this is not the first time you have picked on a poll that is better for No and tried to have it removed from this article when there is clear precedent for the poll. AlloDoon (talk) 10:39, 10 November 2023 (UTC)

Good Grief. You can twist and ignore context as much as you like, you can fall back on your too often default position of ad hominem, however the FACT remains that YouGov have categorically stated that the poll is not comparable with their other VI polls due to different methodology and wording. It really is that simple. I have therefore have created a new section for the scoop polls which seems a sensible place for them to be. Soosider3 (talk) 12:03, 10 November 2023 (UTC)

The fact of the matter is there is no basis for this whatsoever. Read the note again. They have said there is very minor differences in methodology and wording to other YouGov polls (as how there are differences in methodology between BPC members) but the resulting difference is expected to be very minor and the poll is listed on John Curtice's website, who is the head of the BPC, and also verified by Ballotboxscot. AlloDoon (talk) 13:03, 10 November 2023 (UTC)

Okay

Lets step through this.

Here is the actual article from Yougov website in February https://yougov.co.uk/politics/articles/45269-nicola-sturgeons-legacy-according-scots

This is a quote from that article, last paragraph "However, the voting intention results used slightly different wording and did not include YouGov’s standard turnout weighting and so should not be directly tracked to other YouGov voting intention figures as they are not identical. Instead, they should be tracked to other SCOOP voting intention polls conducted by the Scottish Election Study (latest wave conducted in November 2022)."

The key sentence is the one about "so should not be directly tracked to other YouGov voting intention figures as they are not identical. Instead, they should be tracked to other SCOOP voting intention polls" This from yougov themselves and written by their head of data journalism.

That is my rationale for not including them in main table and opening a new section to capture there information.

I think you will find that Ballotbox Scotland that you mentioned above takes a very similar view to mine. Yes report them but dont compare scoop to other yougov polls.

Genuinely I struggle to follow your rationale, to me it appears straight forward and unambiguous, we would be in error to treat the scoop polls as ordinary polls and therein run the risk of not giving readers as clear and consistent view as we might. Soosider3 (talk) 14:56, 10 November 2023 (UTC)

Been doing some digging, there appear to have been 7 SCOOP reports carried out since Dec 2021, at this time I have not been able to get to teh data to ascertain how many of them asked the IndyRef question. If this is the case then there is data we need to get and get it into our article (we can perhaps hassle about where exactly when we have a clearer picture)

This came from noticing that Ballotbox Scotland referred to a Jun 23 scoop, and I went hunting. Could find little data on Scottisg Election Study site, nor on Scottish Opinion Monitor other than a link to their github page where I can see index for several SCOOP reports but have so far been unable to access them.

I have now reached the addled stage but if anyone feels like it the links I have so far are

http://scottishelections.ac.uk/scoop-monitor/

https://github.com/ScottishElectionStudy/Scottish_Opinion_Monitor/

I will message Ballotbox as well

Don't do anything daft while I'm away Soosider3 (talk) 16:10, 10 November 2023 (UTC)

Again, pollsters have slightly different methodologies, that does not invalidate the poll. AlloDoon (talk) 17:36, 10 November 2023 (UTC)

That would depend where the data is captured, placing in the main table would be misleading a section of its own would not. Soosider3 (talk) 18:16, 10 November 2023 (UTC)

I still don’t agree with this. Surely YouGov would also say their polls aren’t directly comparable with polls from other companies but we have them all listed together. And I would guarantee that over time pollster’s methodologies change and develop. So a poll from 2023 may not be directly comparable to one from 2014. At best this should be a note to say Scoop polls are different to other YouGov polls. Dunk the Lunk (talk) 17:34, 10 November 2023 (UTC)

Again the company themselves have made the statement "However, the voting intention results used slightly different wording and did not include YouGov’s standard turnout weighting and so should not be directly tracked to other YouGov voting intention figures as they are not identical. Instead, they should be tracked to other SCOOP voting intention polls conducted by the Scottish Election Study " I really don't see what the fuss is, by al means let's capture them in their own section but to ignore YouGov own statement is just daft. Worse it goes against what YouGov specifically states we shouldn't do. Soosider3 (talk) 18:08, 10 November 2023 (UTC)

But the table isn’t a tracker of YouGov standard polls. It’s a list of all polls conducted by BPC member companies (or other agreed criteria such as Lord Ashford) and which ask the standard question. Which the Scoop polls surely do. Dunk the Lunk (talk) 21:24, 10 November 2023 (UTC)

My that is some feat of mental gymnastics, the article tracks polls, it is is prime purpose to capture record and by doing so track polling for Independence. In doing so it offers these polls for comparison with each other, broadly speaking BPC polling follows a set script, but when a BPC member states not to compare Scoop polls with these we should ignore that, why? Do you seriously suggest we should ignore the professionals, the very ones who do the polls, that would be a rather arrogant stance to take and I am sure it is entirely unintentional. Soosider3 (talk) 05:51, 11 November 2023 (UTC)

By this logic we would need to separate out all polls by company and methodology. If one YouGov poll isn’t comparable to another YouGov poll how is it comparable to an IPSOS or Redfield, or Find Out Now poll? For example it’s clear that YouGov and IPSOS use very different methodologies that arrive at different answers (YouGov broadly favouring No, IPSOS broadly favouring Yes)-these companies are almost certainly both ‘right’ within the constraints of their different methodologies. The September Find Out Now poll was published with a whole host of different weightings (ie different methodologies) of which we agreed to choose one, but one that probably isn’t the same as the previous FON polls. But we didn’t separate it out. And finally, What Scotland Thinks, run by the head of the BPC, hasn’t separated it out on its poll tracker: https://www.whatscotlandthinks.org/questions/how-would-you-vote-in-the-in-a-scottish-independence-referendum-if-held-now-ask/ Dunk the Lunk (talk) 09:48, 11 November 2023 (UTC)

More mental gymnastics, it feels a bit desperate as you try to spin unrelated issues to justify your position. I don't intend to get drawn into your deflections, other than to say how it shows the paucity of your logic, strip it away and you are basically saying you know better than the author of the report, don't compare with other you gov reports compare with scoop reports. Soosider3 (talk) 11:37, 11 November 2023 (UTC)

I'm afraid to say the one doing mental gymnastics here is you Soosider.

Different BPC have slighty different methodologies. The YouGov poll for SCOOP notes that it uses a slightly different methodology than normal but the impact on this compared to regular YouGov polls is expected to be minimal. John Curtice, head of BPC, lists the poll on his website. There's no reason to exclude the poll from the main table of polls, all of which are not directly comparable in terms of methodology - for example many exclude 16-17 year olds - but include everything else together. If you can provide a coherent reason why the poll does not follow required BPC weightings please provide proof of this. AlloDoon (talk) 15:23, 11 November 2023 (UTC)

Again, selective quoting, this is not about methodological minor difference, if it were we would merely place a note in the appropriate field and not count it into the graph. Yougov state not to compare to their usual polls. That's not me doing mental gymnastics, it is a clear unequivocal statement from the pollster. Answer why they would state that but you somehow think they didn't mean it.

Scoop is actually a tracking poll with an established series, from Scottish Election Study.

The correct place for their data is in their own section. Soosider3 (talk) 11:13, 12 November 2023 (UTC)

So why is What Scotland Thinks, run head of the British Polling Council, including it on their tracker? Dunk the Lunk (talk) 19:04, 11 November 2023 (UTC)

WST often a bit tardy in doing the actual work on their site, data links etc so I suspect we will have to wait and see, in meantime I will email them and ask. By the same token Ballot Box Scotland (mentioned earlier in discussion) takes the view of not comparing to other polls, yes report on it but separately. Soosider3 (talk) 11:16, 12 November 2023 (UTC)

Show another example where the polling company has stated not to compare a piece of their work with another piece of their work........I can wait. Soosider3 (talk) 11:28, 12 November 2023 (UTC)

I think you are completely misunderstanding who their statement is aimed at, which is political scientists, academics and students who are likely to be using the data for direct analysis and studies eg https://d25d2506sfb94s.cloudfront.net/cumulus_uploads/document/x4ae830iac/YouGov%20–%20GE2015%20Post%20Mortem.pdf, or Journalists/bloggers who may be saying something like ‘support for independence is up or down by X% points since the previous YouGov survey’. It’s not for the likes of a basic table which exists to list all polls asking about Scottish independence. Once again we are in danger of failing to present all data/information to the user, but selectively editing it. Dunk the Lunk (talk) 12:59, 12 November 2023 (UTC)

That's speculation again, it belongs in its own section. In meantime have started tidy up exercise on coding of colours, will move into highlighting later today, this should polish off work to improve article for colour defect folk (well almost) Soosider3 (talk) 13:13, 12 November 2023 (UTC)

The blog for the Find out Now poll for the Alba party in June says ‘This is not a standard hypothetical voting intention poll on Scottish Independence, but it is consistent with the ones we have conducted as recently as…’. Not sure what the heck they mean by its not a standard hypothetical voting intention poll as they don’t elaborate any further, but this says they’ve done something different from normal, it shouldn’t be directly compared to previous ones but is ‘consistent’ with them. Dunk the Lunk (talk) 13:31, 12 November 2023 (UTC)

Aye that was a weird one, which was also compounded by it being a smaller poll Soosider3 (talk) 13:42, 12 November 2023 (UTC)

They also did something weird for the poll in Sep by introducing a new weighing, fortunately breakdown data was also done for more usual weighing and we settled on that. Wonder if we are likely to see more of these sort of polls that wander away from the 'norm' of the last 10 years or so perhaps it is inevitable as the whole thing politically twists and turns over the years. Soosider3 (talk) 13:47, 12 November 2023 (UTC)

Have had very useful email exchange with whatscotlandthinks, they clarified their position. Having checked the guidelines they were of the view that whereas Voting intention for Westminster/Holyrood in SCOOP polls was different enough to NOT be comparable with mainstream yougov polling, the same issue did not apply for polling in SCOOP polls on Independence and therefore they believed it was okay to report and compare, although obviously there is no weighting for likelihood to vote.

I accept there view on this matter, which means we might have a bit of backfilling to do with previous SCOOP polls that may have been missed Soosider3 (talk) 15:56, 21 November 2023 (UTC)

Colour Vision Amendments II

I believe we have made substantial improvements in this area, well done everyone, not only are tables more readable for folk with a colour Vision Defect but a consequence has been an improvement in the coding with consistency in Colour used in Column Header and Lead Highlighting Column. Still some work to do in the highlighting within the tables, these are in 2 main areas 1 Huge variety in coding and hues used across the article tables 2 The 4 shades used in table to highlight, lead and absolute lead need to be agreed I would prefer to avoid us getting into reverting back and forward, so thought I would share my thinking - so far The lead for No shade seems to be okay The absolute lead for No is too dark The lead for Yes is too pale The absolute lead for Yes is too pale With that in mind I am working on finding better shades and will share these at the bottom of main table when I have proposals ready. Hopefully we will reach agreement that would allow me to sort out the coding. As always would welcome any thought. Soosider3 (talk) 13:49, 13 November 2023 (UTC)

Here is the current colouring of the table with improved coding:

Date(s) conducted	Polling organisation/client	Sample size	Should Scotland be an independent country?			Lead	Notes
			Yes	No	Undecided
29–30 Oct 2023	Redfield & Wilton	1,092	45%	50%	5%	5%
20–25 Oct 2023	YouGov/Scottish Opinion Monitor (Scoop)	1,244	40%	49%	11%	9%
5 Oct 2023	Labour win the Rutherglen and Hamilton West by-election
5–14 Sep 2023	Opinium/Tony Blair Institute	1,004	45%	41%	9%	4%
15–21 May 2023	Ipsos MORI/STV	1,100	51%	45%	4%	6%

Below is an example of the table where the colours are adjusted for blue to mirror red:

Date(s) conducted	Polling organisation/client	Sample size	Should Scotland be an independent country?			Lead	Notes
			Yes	No	Undecided
29–30 Oct 2023	Redfield & Wilton	1,092	45%	50%	5%	5%
20–25 Oct 2023	YouGov/Scottish Opinion Monitor (Scoop)	1,244	40%	49%	11%	9%
5 Oct 2023	Labour win the Rutherglen and Hamilton West by-election
5–14 Sep 2023	Opinium/Tony Blair Institute	1,004	45%	41%	9%	4%
15–21 May 2023	Ipsos MORI/STV	1,100	51%	45%	4%	6%

Or alternatively, red mirrored from the blue colour:

Date(s) conducted	Polling organisation/client	Sample size	Should Scotland be an independent country?			Lead	Notes
			Yes	No	Undecided
29–30 Oct 2023	Redfield & Wilton	1,092	45%	50%	5%	5%
20–25 Oct 2023	YouGov/Scottish Opinion Monitor (Scoop)	1,244	40%	49%	11%	9%
5 Oct 2023	Labour win the Rutherglen and Hamilton West by-election
5–14 Sep 2023	Opinium/Tony Blair Institute	1,004	45%	41%	9%	4%
15–21 May 2023	Ipsos MORI/STV	1,100	51%	45%	4%	6%

My preference would be to mirror blue from a lower saturation of our existing red colour:

Date(s) conducted	Polling organisation/client	Sample size	Should Scotland be an independent country?			Lead	Notes
			Yes	No	Undecided
29–30 Oct 2023	Redfield & Wilton	1,092	45%	50%	5%	5%
20–25 Oct 2023	YouGov/Scottish Opinion Monitor (Scoop)	1,244	40%	49%	11%	9%
5 Oct 2023	Labour win the Rutherglen and Hamilton West by-election
5–14 Sep 2023	Opinium/Tony Blair Institute	1,004	45%	41%	9%	4%
15–21 May 2023	Ipsos MORI/STV	1,100	51%	45%	4%	6%

Or use something similar to the existing table. Thoughts? To my eye the blue looks deeper saturated than the red even though it isn't, but it is an improvement on the deeper saturated red we have at the moment. AlloDoon (talk) 10:04, 14 November 2023 (UTC)

I am looking to transition the Brexit articles to use the standard wikipedia colour coding used in US referendums. Example below of 2022 California Proposition 1.

AlloDoon (talk) 10:10, 14 November 2023 (UTC)

Good work, you really have got the bit between your teeth on this subject and we have come along way on this matter. Initially was going to suggest a pause to reflect as I began to feel we might be getting caught up in the techy stuff and in danger of losing sight of our original purpose, however on reflection lets press on.

My preference had been for no colouring in the cells within the table as black text on white is 100% visible to colour vision defect folk. Recognising that this may be a minority view and looking at your work I think the last sample you posted is preferable (assuming that Headline Yes Blue is lightened to be the same as that in Lead column) That seems to my eye a clean crisp colour palette and one that I could live with.

Happy to do the leg work and sort out coding could you share the coding for the colours of the last table and I'll get onto it. Soosider3 (talk) 13:10, 14 November 2023 (UTC)

Many thanks soosider, I am also happy with the last table. AlloDoon (talk) 14:25, 14 November 2023 (UTC)

AlloDoon, great job with the table colours, I also prefer the final table.

Sole nit is that I would make the column header colours the same as the margin colours. To my eye the red might be the same, the blue looks very different. A palette of 6 colours is probably better than one of 7 or 8.

I'll change the graphs to use the new colours next time. I'm minded to use the level highlight colours from the yes/no columns, but will choose whichever of the palette seems best when I get a look at it. RERTwiki (talk) 15:13, 21 November 2023 (UTC)

Tables looking much better now, have already reduced palette to 6, so Yes/No column headers are the same as those used in Lead highlighting. The darker blue had an issue of contrast when tested against Monochrome. Have trawled through whole article to ensure there are only these 6 colour palate used, hopefully not missed anything. Soosider3 (talk) 15:49, 21 November 2023 (UTC)

Thanks Soosider and AlloDoon for the work, I lost track of who was doing what.

The colours are indeed the same as in the column headings, I guess I just need a new monitor or better eyes! RERTwiki (talk) 16:58, 25 November 2023 (UTC)

I've updated the charts using the highlight colours from the yes/no columns.

Any thoughts let me know. RERTwiki (talk) 18:30, 25 November 2023 (UTC)

They look fine, good work Soosider3 (talk) 11:59, 27 November 2023 (UTC)

FindOutNow

I've taken a look at the many, many polls conducted in 2023, and there is only a single poll not conducted by FindOutNow that has shown a Yes result. A single poll out of dozens and dozens. FindOutNow seems to consistently be inflating the Yes result in its polls. The impartiality of the pollsters has to be called into question too considering they are quite clearly pro-separatism and somehow seem to keep churning out polls that are completely at odds with every other mainstream poll being conducted around the same time. 2.99.68.104 (talk) 04:49, 8 July 2023 (UTC) BE sock. Mutt Lunker (talk) 12:37, 5 August 2023 (UTC)

No, none of the polling companies listed here (which are all members of the BPC or abide by their rules) are biased or in favour of one side or the other. Any accusation of impartiality by any of the pollsters listed here is false. It serves them no benefit.

The methodologies of pollsters do produce different results. Most appear to be pretty neutral but Find Out Now do produce results with the most extreme variation from the mean of any pollster. Their method for getting a sample is very different to other pollsters and they are also much cheaper which has the potential for more ‘exploitation’. For example, there is currently a lot of chat in the pro-independence blogosphere about how FON dont weight for 2014 (which it’s argued is 9 years out of date). However it conveniently ignores the fact that Redfield & Wilton don’t weight for 2014 either and they seem to produce results favouring No. For example this twitter post by someone who actually commissioned a FON poll which appears to me to be factually incorrect: https://twitter.com/Celebs4indy/status/1677264079818194945. The point here is that the poll company itself is not inherently biased, but that their results can be used in a biased way. All this has been discussed previously on this page-perhaps you should read this before commenting? Dunk the Lunk (talk) 06:31, 8 July 2023 (UTC)

"Find Out Now do produce results with the most extreme variation from the mean of any pollster."

And they also happen to be openly pro-separatist and thus biased. You don't find that strange at all. That the pro-separatist outlet just happens to keep churning out polls showing support for the cause it supports a good 10 points higher on average than every other poll conducted around the same timeframe.

If there was an outlet that was openly British nationalist and it kept showing polls showing support for remaining a part of Britain a good 10 points higher than every other poll I'd be raising the same questions.

Something tells me you'd be agreeing with me in that situation, though. So why the ridiculousness here? 2.99.68.104 (talk) 09:39, 8 July 2023 (UTC)

Oh wait you do largely agree with me, I went back and read what you suggested I read. Oh well, nevermind then. Guess there's nothing anyone can do until the BPC excludes them or blacklists them or whatever. Ridiculous it's gone on for this long. 2.99.68.104 (talk) 09:55, 8 July 2023 (UTC) BE sock. Mutt Lunker (talk) 12:37, 5 August 2023 (UTC)

I completely agree with Dunk the Lunk

We have had this discussion and we seem to agree that it is methodology bias rather than any company deliberately bias. Soosider3 (talk) 10:28, 8 July 2023 (UTC)

Oh. Oh hold on. For some reason I was of the opinion that Find Out Now was in some way affiliated with or funded by The National. I was of the impression it was a polling organization working for a pro-separatist media outlet. I appear to be gravely mistaken with that belief.

Alright, well I firmly retract my claims about deliberate bias on the part of the pollsters then, but continue to echo the complaints about methodology bias. 2.99.68.104 (talk) 10:54, 8 July 2023 (UTC) BE sock. Mutt Lunker (talk) 12:37, 5 August 2023 (UTC)

I don't think FON are pro Indy. The trouble is their method is deeply flawed for a controversial issue with thousands of committed activists in the population. Consider ALBA is an organisation whose sole purpose, leaving aside Alex Salmond's ego, is to game the electoral system to inflate pro-Independence representation.

So consider ALBA commission an Indyref Poll from FON. Is it so far beyond imagining that some ALBA supporters (are encouraged to) flood the FON website in the relevant window to distort the poll? I can tell you what lies they would have to tell, but you can probably work it out.

The method is not fit for purpose in this context.

I think we have to carry BPC polls though. It might make sense to strengthen the leading comments about differences in methodology affecting results.

I've written to the BPC about this in general terms, to no response so far. I will do anything else I can think of to get some sensible changes made, which admittedly might not be much. RERTwiki (talk) 16:54, 12 July 2023 (UTC)

Can I suggest we take some time out here, we seem to be in danger of lapsing into some of the less desirable behaviours of social media, with accusations of bias etc etc. We edit a poll aggregator and generally despite the different stances we might have on Independence, we do so in a balanced and consistent manner. Lets not let the toxicity that can exist on the topic, in other places, pollute our efforts here.

Judging by the views this article gets, we seem to have some value for readers, I believe that is because our present system is fairly robust and simple, if pollster is part of BPC we record it and note any variations. As we are more immersed than many in this perhaps we are more aware of Methodology variations that different pollsters have, pollsters are not biased. Suggest we use that phrase "methodology variations" rather than bias - it has too many connotations from other places. Soosider3 (talk) 20:49, 12 July 2023 (UTC)

If I might add something here. I manage the Independent Voices blog and have commissioned two polls from Find Out Now in 2023. They are professional, dilligent and strong on communication around method. Everything is transparent. There is no value in trying to influence polling companies one way or another, FON just happen to be inexpensive and accessible. The big factor in the polling company differences is pretty simple: it's how the sample is constructed. Polls that use the 2014 vote to weight their samples tend to boost the number of No voters in that sample. The No voting intention is strongly correlated to the % of No voters in the sample. Some pollsters have samples that are a poor reflection of the current electorate, for example Panelbase have only 12% who did not vote in 2014 - this figure realistically should be around 27%. Pollsters that have a decent % of new voters and that don't boost the No sample by 2014 weighting tend to show Yes in the lead. Not just Ipsos, also Opinium, Hanbury and FON. Oh, and both Yougov (December 22) and Savanta (Feb 21) have published polls unweighted by 2014 result (both with Yes leads). Given that these polls are funded by the Times and Scotsman unsurprising that they went back to their 2014 weighting! IndyVoices (talk) 15:34, 4 December 2023 (UTC)

I agree with you exactly on the actions we should take: keep tracking all polls to keep people informed.

I'm not accusing any pollster of bias. I am saying that I think the FON method has obvious vulnerabilities which makes it unsuitable for independence polls, though it nonetheless has to be included until the BPC takes action. I don't know enough about other methods to comment on them.

I do think the leading caveat on methodology needs to be strengthened, and will float something below. RERTwiki (talk) 16:35, 13 July 2023 (UTC)

The accusations of bias were on my part and it was due to a misreading/misunderstanding of the data. I thought FON was affiliated with pro-separatist organizations due to the fact they were running the poll in that outlet or something. My mistake. 2A00:23EE:15E8:2A84:7995:D699:FAEA:EAC7 (talk) 19:56, 17 July 2023 (UTC) BE sock. Mutt Lunker (talk) 12:37, 5 August 2023 (UTC)

I have been thinking about the whole postcode lottery/find out now issue. One thing I don’t understand is some (not all) of their polls include 16-17 year olds. How does this work when you have to be 18 to register on their site? Or am I missing something? Dunk the Lunk (talk) 17:50, 29 July 2023 (UTC)

Find Out Now Sep 8-13 Poll

I have added link to data table, however I think we need to have a discussion as to which data table we should link to, FoN at the request of their client, has produced tables that have data on both weighing for or without 2014 referendum vote, it produces a small difference. However they have gone on and ascertained Voting intention, likelihood to vote and turnout weighing as well. These do produce a much larger change in projected outcomes, depending on which combination of weighing is used. If folk could have a look at the 8 data sets produces and share thoughts as to which we should use for main table. I tend to think that the first 2 tables offered are the ones most consistent with other pollsters, however if other weights used were to become more commonly used then we may need to open a sub table to capture them Have a look and let me know what you think https://findoutnow.co.uk/blog/small-lead-for-yes-to-scottish-independence/ Soosider3 (talk) 10:31, 15 September 2023 (UTC)

Hmm it’s a tricky one I’ll agree. Generally I think we should report the headline figures in the main table but in this case FON appear to be reporting two headlines (‘With 2014 vote quotas’ and ‘Without 2014 vote quotas’). Obviously there needs to be the minimum of a note to state there are multiple versions weighted differently. I’m not sure which of the two headline ones would be best to be honest but happy to go with the consensus (unhelpful I know). However, I have to say this just increases my scepticism about the FON polls. It’s like they’re not confident in their own methodology so are throwing a load out and hoping one sticks…

With regard to the notes for 16/17 y/o and 2014 has this been done/checked throughout for consistency? Dunk the Lunk (talk) 15:14, 15 September 2023 (UTC)

I'm fairly sure the 16/17 year olds is highlighted where appropriate, not so sure with 2014. Soosider3 (talk) 16:11, 15 September 2023 (UTC)

Okay, I have checked all polls recorded in 2023 for consistency of weighing, done on a wee spreadsheet that I would be happy to share. Having looked at that I went back and changed the data link for the last FoN poll, reason for that was they had produced 8 different data tables, counting out and in various weighing, they had introduced a weighing that was new to me 'Turnout' this was in addition to 'Voting Intention' Having checked 2023 polls this is first time I can see such a weighing being used so felt it best avoid it. We can go back and look at it again if it becomes a regular occurrence, but for now its a one off and best avoided.

As a summary so far in 2023 there have been 35 Polls recorded that have access to full data tables of these

35 have weighed to 2019 GE

23 to 2021 SGE

25 to 2016 EU Ref

23 to 2014 Ref

28 for voting Intention

29 Include 16+

There is an unclear picture but think that at least some of the variations/trends/changes may be down as much to weighing variations as any thing else. Soosider3 (talk) 07:50, 16 September 2023 (UTC)

This makes sense and I’m happy with it as long as you/others are. I do think there should be a note with this rationale explained, potentially with a link to the FON website page you link too above? Dunk the Lunk (talk) 11:33, 16 September 2023 (UTC)

I'm a little confused as to why you would take out turnout weighting where it's available. Yougov are the only main pollster I know that don't weight for turnout in indyref polls, whcih only serves to give a large number of 'don't knows'. The likelihood to vote figures (as with everything) are transparent and available for scrutiny. The figures align very closely with the actual indyref turnout figure. Plus, I had to pay for the voting likelihood question! IndyVoices (talk) 15:44, 4 December 2023 (UTC)

FocalData Poll October

Was doing some tidy up work on links to data, really struggling to find data tables from this Polling Company, yes I can find published data from the client but we should be linking to best data. Focaldata appear to require an account to be able to access their published polls, which seemed a bit off. Anyone else having any luck with this? Soosider3 (talk) 11:59, 27 November 2023 (UTC)

Looking at the data published on the clients site and looking at the partial tables they published, it is difficult to see where the numbers for both the Yes/No question and the Leave/Remain, they appear to be subsets of about 500 responses and not the 1000+ claimed, this also raises the issue of reporting on what are actually subsets of a full poll.

Information can be found here

clients article https://www.these-islands.co.uk/publications/i389/its_the_way_you_ask_them.aspx

clients published tables on google docs https://docs.google.com/spreadsheets/d/1kkmgMPSSDAOfIgwjowGKJr6aXjowJaGc6fnZaP_O_IE/edit#gid=0

Please have a look and see what you think, to my mind there is something not right here, 4 weeks later and no pollster published data, sub sets masquerading as full reports. Soosider3 (talk) 14:20, 27 November 2023 (UTC)

I had emailed focaldata and got response pointing me to the google docs link, is this normal for a pollster? Have responded asking for more clarification especially around querying if this is their usual practice and why so many questions appear to be missing from the published information. Soosider3 (talk) 19:21, 27 November 2023 (UTC)

That's a couple of days now and no comment, should I take it that silence is agreement and remove this polling data? Soosider3 (talk) 14:43, 29 November 2023 (UTC)

Focaldata are a member of the BPC so their polling should be included in the article. One of the stipulations of being a member of the BPC is that the polling data should be published within 2 working days on their website. AlloDoon (talk) 10:21, 6 December 2023 (UTC)

Good morning, still not able to find data on their website. Any thoughts on data as published by client. These appear to be subsets but unlike previous smaller sample polls no indication that they have weighed for smaller sample, if correct this has to raise questions about the inclusion of this particular poll Soosider3 (talk) 10:39, 6 December 2023 (UTC)

It is acceptable for Focaldata to link to Google docs so long as it contains relevant polling data and sub-sampling. The weighted sample of the poll which you have linked on the independence question is 545 for both Yes/No and Remain/Leave questions. As this sample is weighted and within an appropriate margin of error below 7% it is acceptable to the poll listed in the article with a note explaining the sample size is smaller than 1,000 and has a margin of error above 3% AlloDoon (talk) 16:01, 6 December 2023 (UTC)

The point I was attempting to make was, if the link is not from the pollsters website can we be certain of its veracity, at the very least it raises the suspicion of a doubt. BPC rules very specific re publishing on pollster website. Until now all links to data tables have been from pollsters website and thereby can be assumed that they are the tables the pollsters stand by.

Re numbers of samples, not convinced that these are other than subsets of a full poll, I see no evidence of them weighting for smaller sample or generating what margin of error would now be. Its a question not a statement I am making, where is the evidence to support them being treated as proper polls, all be it small ones. Soosider3 (talk) 11:44, 7 December 2023 (UTC)

It getting on 2 weeks since I first posted my queries about this poll, in particular issue of link to data tables and use of subsets.

I am minded to remove them until we can establish there usability Soosider3 (talk) 10:16, 10 December 2023 (UTC)

This from 'theseislands' article is an amendment from the original publication, it offers an explanation as to how the numbers were arrived at, in essence they ask the same question twice at different parts of the interview then combine those numbers. Once eraly on and again later in process, there is no publication of all the questions between those referred to. This is unacceptable, as I am sure you realise that questioning order can and does have an impact on returns, to then merge these 2 separate and different responses to masquerade as a single questioning is highly dubious and to present as one response is misleading. I intend to delete Focaldata information until those who might support there inclusion can answer those queries. Soosider3 (talk) 13:42, 14 December 2023 (UTC)

I don't accept this at all, they are BPC members, not you. I'm going to revert these changes.

Your incessant determination to remove any poll which doesn't support your political agenda is abusive in the extreme, and needs to stop. RERTwiki (talk) 16:44, 14 December 2023 (UTC)

The assertion that about my motives do you no credit. Look at the data posted and try to square it with what was published by the client. The client stated in their article that they had asked the question twice as q6 and again as q24, effectively creating 2 subsets and then amalgamating them, at best highly questionable activity. I first raised my doubts over 2 weeks ago, but it has taken my action to garner a response. At present the pollster has not yet published the data on their website ( contrary to box rules) By all means explain if my rationale is faulty, I have always understood we don't publish subsets adding 2 together does not make it okay. I would remind you that not only the wording of questions but the order they are asked can impact results Soosider3 (talk) 17:09, 14 December 2023 (UTC)

The procedure specified in the InTheseIslands article is fine. They asked everyone both questions, and in order to avoid the bias implicit in, say, asking yes/no first and remain/leave second, they split the population into two and asked each in different orders. This seems eminently sensible to me, and obviously they have thought about the problem. They are the professionals here.

One can make a reasonable argument that visibility of both yes/no and remain/leave framing is what people would see in an actual campaign. For certain there will be a campaign for "Yes!" and the unionists would have to be blind and stupid not to campaign on "Remain". That is, asking both questions plausibly creates a more realistic response.

We are not quality control for the polling industry. This poll is good enough for What Scotland Thinks, and is produced by a BPC member. We should have a strong bias in favour of inclusion, for reasons previously cited. The poll should stay in the list.

Yes, the poll favours NO, but by about the same margin that MORI always favours YES and much less than the outlandish margins FindOutNow has favoured YES in some polls.

I don't appear to have the same conception of credit as you. As my grandmother used to say, least said soonest mended. RERTwiki (talk) 10:21, 15 December 2023 (UTC)

Your grandmother sounds a very wise person.

I would respond by pointing out that it is not the pollster offering these explanations but rather the client, that the pollster has not issue these tables (as of yet) and there is no explanation of methodology from the pollster. We should always be cautious about sources of information and how much weight to give them and would suggest that a campaign group maybe somewhat less reliable than the actual pollsters ( who are missing from this discussion)

Please note that it is well established in polling circles that when the question is asked in the polling questionnaire can impact its result, just as the questions asked leading to the question can also impact the result, we have no data as to what Questions 11 to 24 might have been.

As I say there is some highly questionable stuff happening here, as well as not polling 16/17 years old or weighing for likelihood to vote, we have (to the best of my knowledge) the Q being asked twice at different stages of questionnaire.

This is unusual practice and should be noted as such, perhaps we should include a separate line for each of the small polls with the appropriate information in the notes field, as if not subsets they are small polls

Notable that yet again we have a mention of the "methodology bias" we had some discussion on here earlier in the year around Findoutnow poll, I had did a wee bit of research on that and think there maybe a link between what pollsters weigh for in previous voting (ie Indyref, Brexit, Last GE and last Holyrood) and some of the variation. Perhaps we should have 4 wee columns to highlight this with an x if weighed to that measure Soosider3 (talk) 12:52, 15 December 2023 (UTC)

I don’t think 4 more columns is a good idea.

I’ve seen polls before in other contexts that randomise question orders for the reasons you mention. This is the same. I think we just have to take the poll at face value.

The links allow really keen readers to form their own opinions about polls.

If you want to add a brief note in the usual place highlighting what they have done I think that’s entirely reasonable. RERTwiki (talk) 10:13, 18 December 2023 (UTC)

Fully in agreement with RERTwiki here. Lets avoid any confirmation bias. AlloDoon (talk) 10:50, 20 December 2023 (UTC)

Oh behave yourself, lets report the data as it is presented in data tables and not as per client. Soosider3 (talk) 12:15, 20 December 2023 (UTC)

Aye, perhaps a note is the way to go at least we seem to have some consensus on that, I am aware of the asking the questions in a random order process, however that is not what happened here, where the same Q was asked at 2 different stages, in effect producing 2 small polls (at best) perhaps we should report them as reported by data tables ie as 2 polls rather than as client presented them, that is usually our default position - to go to the tables as most reliable source of data, information clearly presented in a manner that is obvious to the reader, rather than how the client would prefer it. Otherwise we are actually in danger of manipulating the data, check out the data tables pollster has presented them as seperate that is how we should record them

I see that Focaldata have at last published the tables on their website and it only took a complaint to BPC to get it done, perhaps not a ringing endorsement of them as political pollsters. Still no statement from them as to methodology used. Soosider3 (talk) 12:13, 20 December 2023 (UTC)

IPSOS poll removal

Happy New Year!!

I noticed when updating the charts that two IPSOS polls from back in '22 have been removed from the main table.

It doesn't make any difference to the charts, as they were already marked as non-standard questions and have never been included. I noticed they were missing when I didn't have to mark them for removal.

There is a comment in the edit history to the effect that they didn't ask about a future referendum, which seems quite unlikely, but might be grounds for excluding them.

I'm generally very negative on memory-holing interesting data, and would rather this was back as it was.

What do people think? RERTwiki (talk) 11:53, 17 January 2024 (UTC)

and a happy new year to you.

That's an odd one as notes field had already referred to it. Given length of time polls had been present I would assume our more regular editors were happy enough, therefore have reverted them. Soosider3 (talk) 12:18, 18 January 2024 (UTC)