Apologies for not tagging you there @alexandria-ah.bsky.social but to be fair yesterday you did proudly tell me you didn't have any social media so...

Just saw this thread while searching the hashtag and I had the same feeling this year of not feeling the desire (or need?) to do the live updates. May be because I no longer have a role, but also just feels like evidence that the relationship to social media has transformed over the past year

I should say a massive well done to everyone involved in the conference, particularly @jen-studies-death.bsky.social and Alex who put so much into making it happen. SocRel really is an annual highlight and it just doesn't happen without folk like them working thanklessly throughout the year

Ultimately from a data/sampling/methodological perspective it's their reputation at stake, and they were happy for it to go to print as written.

Worth adding, however, that we wouldn't be allowed to publish inaccurate analysis of the data - YouGov have reviewed everything in the report and have signed it off from a data accuracy perspective. If we bastardised the data they would publish a retraction and distance themselves

As is standard, the methodology and data tables on published data are available on request from YouGov (I don't have the direct link to hand atm), and we are exploring making more available in the future. All the relevant questions and sample sizes/info are also included in the report.

Sorry have to jump but as I say, a fuller FAQ coming soon.

Ultimately we have to acknowledge that, short of a full census (which does not ask about religious practice but that's a longer story) all surveys are going to have their limitations. Even British Social Attitudes! YouGov are, however, *incredibly* good at this form of polling.

[..] Denominational counts are measured VERY differently to a YouGov survey, and depend ultimately on a wide range of volunteers. I should say the data - and not just because of the findings - has been warmly received by senior figures within denoms because it gives deeper insights

- Having said that, the trend differences between ours and the denom counts are both curious and fascinating (though in some cases not as far out as would appear once margin of error is taken into account). The methodological differences between the approaches though are not insignificant [...]

- I would caution heavily against extrapolating specific numbers from percentages, particularly from smaller sample sizes, partly due to the larger margin of error at this scale. We avoid giving denominational, age etc *figures* for this reason

- A social desirability/"Shy Tory" dismissal needs to address convincingly firstly why it wasn't in effect in 2018 and secondly why these respondents look, across the whole breadth of the survey, like committed churchgoers. It's a coherent data story

Appreciate the blog, a Bsky thread probably not the best place to respond in full but we will have an FAQ piece coming out very soon addressing a lot of your queries. A couple of quick bullets though:

The BES data is available for download - www.britishelectionstudy.com/data/ - but all requires additional analysis, so this graph is based on The Economist's analysis of the data.

Then today in the Guardian, looking for further research into what is drawing young men into the church - but looking with a sense of optimism www.theguardian.com/commentisfre...