Talk:Religion in Eritrea

Religion Pie Chart

Given that there's a huge variance in estimates of the population of Eritrea, it shouldn't come as a surprise that sources vary widely about the religious mix of that population. The text does a good job of laying out that diversity of opinions, listing a range of sources and their estimates (from 63% to 47% Christian). Given the diversity of data points, having a very prominent chart at the start of the article showing just one data point doesn't present a balanced view of the sources. Pew is a respected organization, but with the other sources all close to 50/50 Christian/Muslim, having the chart on the top of the page just show the Pew data point is WP:UNDUE. I recommend we either (a) remove the chart entirely, or (b) instead include a chart showing estimates from multiple organizations. BubbaJoe123456 (talk) 22:42, 28 December 2021 (UTC)

Pew source is the most credible source, the USDPS cites it in their own report. USDPS don’t conduct any estimates. To present a range estimate using a second hand source with no data points is making own research. You also cannot mix several sources and combine them to come up with an estimate. There aren’t many other credible available estimates of the religious composition. Leechjoel9 (talk) 08:47, 29 December 2021 (UTC)

Why do you regard Pew as "the most credible source"? It's a good organization, but they're an outlier here, and it appears that they've just taken their composition estimate from a 2002 survey - they didn't create or develop it themselves. The more recent version of that survey, conducted in 2010, came in at 60% Christian/Muslim. On top of that, there are multiple credible sources that put the religious demographics around 50/50, including the US gov't, the most current publication of which no longer cites Pew, but rather states that the population is "split in half," with 49% Christian and 49% Muslim. The article text is balanced, but having that very prominently placed pie chart just showing a single data source isn't. Do you have any suggestions on how to incorporate the differing opinions about Eritrea's religious mix into the graphic? BubbaJoe123456 (talk) 23:13, 29 December 2021 (UTC)

Pew research has the most comprehensive data of the religious composition of the world, of course their data and studies are considered among the most credible. The data of the pie chart is based on the results of the “2010 Pew Research Center demographic study” of more than 230 countries and territories. The study relied on more than 2,500 censuses, surveys, and population registers. The have adjusted some data since then and even have estimates for many years ahead. It is also consistent with The 2011 edition of the Encyclopedia of Global Religion report, a report that also further analyzed the regional distribution within Eritrea, so the pew research data is not an outlier. There aren’t multiple other sources as you claim. The USDPS source does not provide any data in this field. Removing the pie chart just for the sake of it is not correct. Explanation that some sources put estimates in the 49/49 is already present in the article in text format. None of the other sources that presumably put estimates in the 49/49 has any available data for a pie chart. Different data are used and good for different purposes, some are used in table format, text and others in the form of pie chart.Leechjoel9 (talk) 10:47, 30 December 2021 (UTC)

The data of the pie chart is based on the results of the “2010 Pew Research Center demographic study” of more than 230 countries and territories. If you had looked at the underlying Pew report, in the section where they describe their sources, you'd see it says, on page 12, "Estimates based on 2002 Demographic and Health Survey, adjusted to account for underrepresented religious groups." So, Pew's just using another entity's survey for their Eritrea number, they didn't develop it themselves. As I noted above, the UNICEF survey they use has been superceded by a 2010 survey which asked the same questions, and got answers of 60.8% and 39% for Christian and Muslim. So, there's good reason to think that Pew number, based on a 19-year-old survey, is at least slightly overstated. The Encyclopedia of Global Religion appears to source the same 2002 survey as Pew. The most current estimate from the US government is 49%/49%, although they don't provide a methodology. Bottom line, we've got Pew at 63/37, based on a 19-year-old survey, a somewhat more recent survey at 61/39, the US gov't at 49/49, and ACS-Italia and Aid to the Church in Need with a very slight Muslim majority. I suggest replacing the pie chart with the below chart, showing the 2002 and 2010 survey results, along with the US gov't estimate, as a representative of the entities that estimate a roughly even mix.

Chart showing multiple estimates of the religious makeup of the Eritrean population

BubbaJoe123456 (talk) 18:08, 30 December 2021 (UTC)

They composed the study in 2010 when it was published, and in 2015 stated the same. They didn’t rely on a 2002 study, to claim that is against WP:NPOV, you can not just make up assumptions. Pew got estimates many years ahead. The Encyclopedia of Global Religion is a different study and it bases partly their data (not completely) on an official census from 2002, which is consistent with the pew research center. These three studies are three different studies and not the same study. The official census is the only official census to date. There is no available data for your charts, only second hand cited sources with no data points or real estimates. The available data sources is pew pointing to 63%C/36%M. Most religion articles have one pie chart for a good overview of the religion and the Pew one is sufficient. The various figures are present in plain text which is good enough. Also again a reminder, you cannot combine two or several studies to come up with your own range estimate e.i 47-67%Christians, since the studies are different and based on different methodologies. That is why the line was removed but also the duplicate line which you reverted.Leechjoel9 (talk) 13:46, 31 December 2021 (UTC)

As I stated and cited above, Pew's study itself states that their "Composition" estimate for Eritrea is "based on 2002 Demographic and Health Survey, adjusted to account for underrepresented religious groups." They didn’t rely on a 2002 study, to claim that is against WP:NPOV, you can not just make up assumptions. I didn't make up anything, I cited my statement with support from the source. You have no grounds to accuse me of dishonesty. Withdraw your attack immediately (by striking through the claim, not deleting it), and apologize. BubbaJoe123456 (talk) 19:31, 3 January 2022 (UTC)

The main point was that it says “Based on”, that you are correct of. For clarification what I meant was it would be incorrect to assume that pew and the 2002 study was completely the same, even though it is based on it. Pew put their number at 62.9 % while the census of 2002 (according to Encyclopedia of Global Religion 2011) was 64% for Christians, 36% Muslims so it differs a bit. The third of your chart displays as showing majority Muslim which the US source do not claim. The aid for church don’t show methodology or year and isn’t a study and point to same numbers as the unavailable ASC-Italia link. The sources that can be considered credible and available are Pew 2020, survey of 2010, and to some extent the US source of 2021. Using no pie chart would not favour any of the sources but adding the two other as well might be better. Leechjoel9 (talk) 03:15, 4 January 2022 (UTC)

Pew put their number at 62.9 % while the census of 2002 (according to Encyclopedia of Global Religion 2011) was 64% for Christians, 36% Muslims so it differs a bit. Why do you refer to the Encyclopedia of Global Religion, rather than the actual survey (not census)? The 2002 survey states 57.7% of respondents said they were Orthodox Christian, 4.6% Catholic, and 0.7% Protestant, or 63.0%. Each of those percentages is rounded to the nearest 0.1%, however. If you use the actual # of responses, you get 5048 Orthodox, 400 Catholic, and 60 Protestant, out of 8756 total, or 62.9%, exactly in line with the figure that Pew used. There's no evidence that Pew didn't just take the results directly from that 2002 survey. The third of your chart displays as showing majority Muslim which the US source do not claim. It does not. It shows 49% Christian, 49% Muslim (as stated in the USCIRF source), with the remainder as "Other," as is clearly show in the labeling. Given that you have no objection to showing the 2002 survey and the 2010 survey as sources, and consider the 2021 US source credible as well, the chart shown above clearly meets your concerns, so I've inserted it into the article. BubbaJoe123456 (talk) 21:36, 4 January 2022 (UTC)

That’s not what I wrote.I did not aprove of removal of Pew chart. I specifically wrote I do not object keeping the Pew chart, alongside, 2010 survey and perhaps the US chart or remove them all along. Leechjoel9 (talk) 08:51, 5 January 2022 (UTC)

That’s not what I wrote.I did not aprove of removal of Pew chart. I specifically wrote I do not object keeping the Pew chart, alongside, 2010 survey and perhaps the US chart or remove them all along. That's exactly what I did. I kept the Pew data (sourced to the actual underlying source, the 2002 survey), and put it alongside the 2010 survey and the US estimate, in a single chart, rather than having three huge pie charts. Can you clarify what your objection is here? BubbaJoe123456 (talk) 14:07, 5 January 2022 (UTC)

No that’s not what you did, you removed Pew chart and their estimates, and replaced it with another source. Pew estimates are from 2020. The three estimate are presented directly from their sources. Having three pie charts are not an issue. Leechjoel9 (talk) 14:31, 5 January 2022 (UTC)

The Pew religion share estimate for 2020 is EXACTLY the same as their 2010 estimate (as are their 2030, 2040, and 2050 estimate). It's also EXACTLY the same as the 2002 Demographic and Health Survey results (62.9% Christian, 36.6% Muslim). They're clearly just using that survey, and assuming the mix of religions stays constant. When they explicitly state the source of their estimate, it makes more sense to cite that underlying source. That said, I changed the column label to "Pew Research (Based on 2002 Survey) to make sure Pew's name is referenced in the chart.BubbaJoe123456 (talk)

That’s your point of view, the estimates clearly state that they are estimates of the year 2020. You cannot remove data, edit these and make your own interpretations of them.Leechjoel9 (talk) 16:22, 5 January 2022 (UTC)

Are the two of you agreed on this much at least, that the Pew figures are predictions for 2020, 2030 etc that they published in 2010? You seem to differ on whether those predictions used the exact same proportions that the 2002 survey found or whether Pew's methodology was more complicated than that, and I notice that they do say for Eritrea "Religion of incoming migrants based on religious composition of the immigrant's origin country.", but do we need to dig so deep? What we need to tell the reader is that this isn't a 2020 survey, nor is it a later estimate of what the situation was in 2020, though the ordinary reader would naturally assume it was one or the other if we simply labelled it "2020". Instead we should find a way to tell the reader that it's an old forecast for 2020, published ten years earlier. NebY (talk) 18:40, 6 January 2022 (UTC)

pov pushing by groznia

There is nothing wrong in show two 2020 graphs on religion in eritrea your pov pushing choosing one and ignoring the other take tgis to the talk page because your the only one who is pov pushing. 2A02:C7C:30D9:C400:689E:EC6E:4731:A70E (talk) 12:36, 4 October 2022 (UTC)

It's you who is POV pushing across multiple pages. Selectively using figures that inflate the Muslim figures even if they are old estimates [1]. The various estimates regarding religion in Eritrea are already elaborated in the demographics section. There's no need to repeat them again in the lead.Groznia (talk) 03:23, 5 October 2022 (UTC)

I used cia world factbook as a source you chose arda a unknown company thats why I reverted that then I chose pew 2020 estimate best. Funny you accuse me of inflating muslim numbers where your edir history shows you have done the same multiple times even here. Nope your purposely choosing a graph that suits you there is norhing wrong with having two in the lead. — Preceding unsigned comment added by 2A02:C7C:30D9:C400:931B:9E73:4771:A174 (talk) 11:36, 5 October 2022 (UTC)

Protection and one-revert restriction

I have protected this article for three days due to the persistent edit-warring. Furthermore, I am also placing a discretionary sanctions notice that no editor should revert more than once within a 24-hour period, otherwise they may be blocked. Ritchie333 ^{(talk) (cont)} 14:14, 5 October 2022 (UTC)

Groznia whats the exact issue both sources give a 2020 estimate

Why are choosing one over the other why not keep both? 2A02:C7C:30D9:C400:6745:82EE:9699:2839 (talk) 14:18, 13 October 2022 (UTC)

As explained to you several times before, the varied statistics are already elaborated in the Faiths and denominations section. I won't be replying to anymore discussion threads that you will create unless you have something constructive to say.Groznia (talk) 06:49, 15 October 2022 (UTC)

All was added was a graph and the arda statistic having two graphs are fine either remove the graph or put its dishonest to place only one when the statistic differ so much. — Preceding unsigned comment added by 2A02:C7C:30D9:C400:5566:F657:1E84:5285 (talk) 12:02, 15 October 2022 (UTC)

Since the dispute between the two of you is in my revision, I would like to explain it. The latest 2020 ARDA estimate (source) I had added to the lead is from a reputable source and was never previously mentioned in the Faiths and denominations or in another section in the article. Even if it was mentioned in that section, adding it into the lead would be better per WP:NPOV as numerous sources give substantially different estimates and choosing one over the other may amount to WP:Cherrypicking as the lead should provide a well-balanced perspective on the topic. Unionsa1 (talk) 20:03, 15 October 2022 (UTC)

consensus for the arda estimate and two graphs added

If no one has a objection then it will be added tomorrow by 12pm. 2A02:C7C:30D9:C400:161D:BF9C:CA41:EE31 (talk) 18:28, 13 October 2022 (UTC)

About your recent addition - the reference you have provided in the new graph (<ref name=":6" />) is undefined. Do you think you could provide the citation you used? Aidan9382 _(talk) 14:50, 14 October 2022 (UTC)

Nevermind, a bot handled it. Aidan9382 _(talk) 15:05, 14 October 2022 (UTC)