Talk:Religion in the United Kingdom

This article is of interest to the following WikiProjects:

WikiProject United Kingdom (Rated C-class, Top-importance) v t e This article is within the scope of WikiProject United Kingdom, a collaborative effort to improve the coverage of the United Kingdom on Wikipedia. If you would like to participate, please visit the project page, where you can join the discussion and see a list of open tasks.United KingdomWikipedia:WikiProject United KingdomTemplate:WikiProject United KingdomUnited Kingdom articles  C  This article has been rated as C-Class on the project's quality scale.  Top  This article has been rated as Top-importance on the project's importance scale. WikiProject Religion (Rated C-class, Low-importance) v t e This article is within the scope of WikiProject Religion, a project to improve Wikipedia's articles on Religion-related subjects. Please participate by editing the article, and help us assess and improve articles to good and 1.0 standards, or visit the wikiproject page for more details.ReligionWikipedia:WikiProject ReligionTemplate:WikiProject ReligionReligion articles  C  This article has been rated as C-Class on the project's quality scale.  Low  This article has been rated as Low-importance on the project's importance scale. WikiProject Christianity / Anglicanism / Latter Day Saints (Rated C-class, Low-importance) v t e This article is within the scope of WikiProject Christianity, a collaborative effort to improve the coverage of Christianity on Wikipedia. If you would like to participate, please visit the project page, where you can join the discussion and see a list of open tasks.ChristianityWikipedia:WikiProject ChristianityTemplate:WikiProject ChristianityChristianity articles  C  This article has been rated as C-Class on the project's quality scale.  Low  This article has been rated as Low-importance on the project's importance scale. This article is supported by WikiProject Anglicanism (marked as Low-importance). This article is supported by WikiProject Latter Day Saint movement (marked as Low-importance).

May 2018

Anonymous IP 79.27.8.245, or I should say FrankCesco26 (same style and same geographic location), your removal of this pie chart (reproduced here on the right) is not justifiable. The study in question is not a random market research poll, is published by two Catholic universities (St Mary, London and the Institut Catholique de Paris), and is based on the European Social Survey. But, more important to our issue, it represents a different section of the population (the youth) and therefore would be similar to a table presenting the % of religious believers among different age groups, similar to this.--Wddan (talk) 14:54, 1 May 2018 (UTC)

NB: FrankCesco26 confirmed to be the IP and justified the removal by saying that "we can't add a chart for every survey", not being consistent with what done later (see below).--Wddan (talk) 15:06, 6 May 2018 (UTC)

6 May

FrankCesco26: You recently added a pie chart representing the breakdown of Christian denominations according to an Ipsos MORI (again, a market research company) reading of the 2001 census (thus the most outdated census), while on 1 May you removed this pie chart (represented on the right) claiming that "we can't add a chart for every survey", and you didn't answer to the request for clarifications above. Now, what is the criterion by which you consider a pie chart which represents a different section of the population and based on a recent survey to be less worthy of representation than a pie chart which represents a nearly two-decades years old census and represents only England and Wales, which are not the entire UK?

I would like to know the opinion of other users, including Iryna Harpy, Nillurcheier, JimRenge, Grsd and others who patrol this type of articles.--Wddan (talk) 14:12, 6 May 2018 (UTC)

The same goes for this pie chart in religion in England, which, besides, is itself misplaced since it represents England and Wales, two different countries of the UK. Why should we represent the 2001 census?--Wddan (talk) 14:19, 6 May 2018 (UTC)

The sample changes the weight that should be given to a determined source. The total sample for UK in the aggregated 2014 and 2016 data is 282 for 2014 and 278 for 2016, then only 560; it is too small to have given a such high visibility, even in the top of the section. It has to be considered a minor survey, that complete the article. Instead, the Ipsos survey, with sample of 5.500, is needed to show the consistency of the christian denominations in England and Wales (in Scotland and Northern Ireland the census also asks for the christian denominations, as you can check in the constituent countries' religion articles), since no survey does it with the census approaches and a consistent sample. The reliability of the data can be proven if you compare it with the 2011 census data, the numbers are right, considering the year gap. Also; as you can see, I didn't add the survey in the Survey list but rather in the "Religious affiliations" section because it gives important data about the christian denominations that fit that section, so I didn't contradict my own words. FrankCesco26 (talk) 20:09, 6 May 2018 (UTC)

The 2018 survey should be represented since it is about the smaller young population, and has been published by important academic institutions. The 2001 survey is, instead, outdated and useless, being no more than bulky garbage for this article which needs far more than some graphics to be improved (it would need rewriting through good sources, in good language... Wikipedia should be more than an amass of statistics). Also consider improving your English.--Wddan (talk) 23:03, 6 May 2018 (UTC)

First of all, you didn't understand that the survey is not from 2001, but uses the same criteria of the latest census at the time and it's actually from the end of 2008, and it improves the article giving a reliable data on the christian denominations, I see nothing wrong with it. If we wanted to show the religious affiliation of the younger population we could have used the census data, that is the only reliable source for it. As for your source, I see it as a minor survey based on a very small sample which is a further survey in the cluster. Also, thank you for the advice but my English is good. FrankCesco26 (talk) 05:44, 7 May 2018 (UTC)

I dislike an inflation of pie charts. As a prominent visual clue they should be used for overall population info only 1 maybe 2 per article. Data of regional, age or other subgroups can be mentioned of course - if database and authors make it over the quality hurdle, but nor illustrated on the same level as main data. And don't comment my English - I am not a native speaker. --Nillurcheier (talk) 07:37, 7 May 2018 (UTC)

I agree with the criterion to not to add more than two pie charts to an article, and possibly have just one pie chart based on the best survey or on the census when available, to avoid pie chart overcrowding. However, I am in favour of representing surveys if the source is an important one (such as an academic institution or European Social Surveys, as in the case of the 2018 survey) and if the survey represents a different section of the population (such as the youth), especially if the article is large enough to accommodate the maximum of two pie charts. I am not in favour of representing surveys when they are from random market research companies and think tanks. So, I think the 2008 Ipsos MORI survey to be an unnecessary addition to the article, also because it is dated.--Wddan (talk) 18:07, 7 May 2018 (UTC)

The Ipsos survey is essential to the article and it hasn't to be removed. It provides clear and reliable data that for England and Wales that was missing in the article, using the same methodology as the census, so in line with the major source of the article and based on a 10 times bigger sample than the 2014-2016 ESS one. Ipsos also is one of the most well-regarded research agencies, as you said for the Ipsos Global Trends survey then and not a random one as you are trying to appoint now. FrankCesco26 (talk) 18:47, 7 May 2018 (UTC)

Ah, and I forgot to say that the 2008 Ipsos MORI survey is not about the UK but about England and Wales, which hare half of the UK. The UK includes England, Wales, Scotland and Northern Ireland. This makes the 2008 survey even more unwarranted in this article.

About Ipsos, I changed my idea compared to what I thought some months ago when I added the Ipsos Global Trends to the article about religion in France and other ones (from which I now think it should be removed). Now I recognise it to be a low-quality source. Ipsos is a market research company, not comparable to an institutional source (either academic, European or governmental).--Wddan (talk) 19:17, 7 May 2018 (UTC)

Ipsos Global Trends are very low quality surveys, but this survey from Ipsos MORI is very good for our case, where just England and Wales don't have detailed data about the christian denominations and this survey gives an accurate profile of them. We can't say that a source is better or worser just by saying which publishing agency is better. Please just leave as it is. FrankCesco26 (talk) 19:54, 7 May 2018 (UTC)

The publishing agency is indeed fundamental, and institutional ones (either academic, European or governmental) should be favoured over market research agencies and think tanks.--Wddan (talk) 19:57, 7 May 2018 (UTC)

Well, the only thing that make a survey less reliable or more reliable is the methodology (chiefly question wording, sample size and weighting). The 2009 Ipsos survey has the same wording of the 2001 (and 2011) census, and thus the data is compatible with it; the sample size is robust and permits reliable estimates that are corrected with a successive weighting using census data, since the survey collected data on age, location and ethnicity. The reliablity of the source is clear. FrankCesco26 (talk) 15:14, 9 May 2018 (UTC)

I disagree with FrankCesco that the reference for the youth survey is not good. It is actually a reliable source. St. Mary's University seems to be a respected institution that is more than 150 years old. I think seeing what the young thinks of religion is important information. I think it should be restored. But some even better information would be the Brits' religious persuasions or lack thereof per age segment. Thinker78 (talk) 03:58, 8 May 2018 (UTC)

The thing I was questioning was the very small sample, not the source. It's too small to give accurate estimates even with the better weighting. There is census data that can refer to the same age group and as well for other ones that can be used to make a detailed table of religious affiliation by age group. That woould be way more reliable than the survey. FrankCesco26 (talk) 15:17, 9 May 2018 (UTC)

Well, I'm guessing that St. Mary's, being a respected institution, might have some criteria in place for statistics that it publishes. If you are accepting St. Mary's as a reliable source, then you accept the information that source contains as reliable. Otherwise you are saying that the source is not reliable. Thinker78 (talk) 23:29, 9 May 2018 (UTC)

While it's true that the St. Mary's University source seems to be a respected source; but there is no need to be highlighted by special pie chart since the article already has a pie chart from the offical census; and there are other reliable sources as well in that section; so no need to select one source that does not even present the whole population (18-29) and create another pie chart for it.--desmay (talk) 19:10, 8 May 2018 (UTC)

I think that the ESS is a good source, and its sample of 1000 is ok if it's used in its entirety, but the sample is too small to provide accurate estimates of the religious affiliation for age group. Few different answers can significantly change the final result, and even if the better weights are used, there is a too wide margin of error. FrankCesco26 (talk) 12:45, 10 May 2018 (UTC)

I completely agree with you, the article doesn't need to have highlighed a survey based on a such small sample and not referring to the whole population. It would be better to use the latest census data and dividing it for age groups to have a similar data but with the most reliable source.FrankCesco26 (talk) 15:14, 9 May 2018 (UTC)

I agree that if there is census data that divides the information in age groups that should be used. But if it is not added my position is that the youth graphic should be restored because even though it doesn't represent all age groups, it provides important information regarding what the younger generation thinks about religion. Thinker78 (talk) 23:33, 9 May 2018 (UTC)

In the Office for National Statistics' website can be found detailed tables about the religious affiliation by age, gender, ethnicity and citizenship, so we can use that. I've just found the data we're searching here, it refers only to England and Wales, and is from the 2011 Census. It says that the data for Scotland and Northern Ireland can be found in the respective statistical offices, I'll search for them. https://www.ons.gov.uk/aboutus/transparencyandgovernance/freedomofinformationfoi/ethnicityandreligionbyage

EDIT: This is the link for Northern Ireland's data: www.ninis2.nisra.gov.uk/Download/Census%202011_Winzip/2011/DC2116NI%20(a).ZIP , the table we need is in NI \ DC2116NI.xls

EDIT2: I've found the Scottish data using the Census Data Explorer, "Standard Outputs", "Religion by Sex by Age", "Scotland" and I've exported this table: https://ufile.io/8iw9z Now we can merge the tables to obtain the data for the whole UK. FrankCesco26 (talk) 12:45, 10 May 2018 (UTC)

Is there the need to specify the sample and the questions of a survey in the text body of the article?

Based on WP:NOTSTATSBOOK and WP:NOTMANUAL, I think that it is excessive and unencyclopedic to specify, in the main text body of the article, the consistency of the sample of any survey and the questions which were asked to the sampled people (see this case). I am not saying that these are not important things, but the main text body should show just the outcomes of the survey, and the methodology should be explained (if the explanation has some relevance) in the reference or in a footnote.--Wddan (talk) 16:08, 15 May 2018 (UTC)

I think that a clear explanation of the survey methodology must be given in the text right under the data, and must not be hidden in references of footnotes since it gives crucial information about the wording of the question, that is one of the main factors that affect the differences between the censuses and the other surveys. The sample size also is very important, since surveys with a more robust sample are usually more reliable than surveys with a smaller sample in which a few different answers can drastically affect the final data, even with the best weights. I would also add information on what kind of weight has been used, but I think that is too excessive. FrankCesco26 (talk) 17:53, 15 May 2018 (UTC)

Hindu Council UK link

The Hindu Council UK link should probably be to http://www.hinducounciluk.org/ now. 94.30.84.71 (talk) 17:22, 31 March 2019 (UTC)