The Census Review Act

[Miestră Schivă, UrN-GC]

There are many in the Christian religions who would consider a protestant as NOT a Christian and many who would consider a Catholic as NOT a Christian. I'm unsure if the same thing applies to Suni and Shia in Islam or not.

Only the "takfiris" (ISIS types) would say that Shi'a Muslims were infidels, rather than "believers who've got it a bit wrong". That said, other branches of Islam are often considered beyond the pale, eg. the Ahmadiyya in Pakistan.

Right now, in my other country of the United States of America, government officials are killing data collection programs of all kinds because they don't see any immediate use for them. It is foolish behavior.

Well, they are establishing OTHER types of data collection. Like, registers of autistic people, or trans people. For later elimination. Which is precisely why France doesn't collect that material - a list of Jewish citizens was of great interest to the occupation/collaborationist authorities in WW2.

[Baron Alexandreu Davinescu]

Right now, in my other country of the United States of America, government officials are killing data collection programs of all kinds because they don't see any immediate use for them. It is foolish behavior.

Well, they are establishing OTHER types of data collection. Like, registers of autistic people, or trans people. For later elimination. Which is precisely why France doesn't collect that material - a list of Jewish citizens was of great interest to the occupation/collaborationist authorities in WW2.

Yes, although I don't think they've actually proceeded to collecting that information, thankfully.  I think they would have a hard time actually doing so, although we'll see.

Also, there's a very big difference between a list of names and an anonymized survey!  I think it would be obviously insane to have a list of Jewish Talossans or something.

[Breneir Tzaracomprada]

If we are able to gather useful and neat information without the breakdowns for any of the faiths and it is marginally easier to compile the data then I'd recommend we just remove the denominations. Or we simply add the options for the other faiths too. The current framework allows only Christians and Muslims to express a more specific subset. I don't think this is fair to Talossans who are not Christian or Muslim (however few they may be).

[Baron Alexandreu Davinescu]

If we change the breakdown of the religion data, it would be somewhat less useful and somewhat less easy to compile, since it wouldn't be less easily compared to earlier versions.  It's not a big deal, but we shouldn't make the change without a good reason.  Hypothetical offense probably isn't a good reason.

I'm not Christian, by the way: I say this as a lover of data.

[Breneir Tzaracomprada]

If we change the breakdown of the religion data, it would be somewhat less useful and somewhat less easy to compile, since it wouldn't be less easily compared to earlier versions.  It's not a big deal, but we shouldn't make the change without a good reason.  Hypothetical offense probably isn't a good reason.

I'm not Christian, by the way: I say this as a lover of data.

That's fair, ultimately it's not that big of a deal, I agree. It was something that came to mind when reviewing the questions again. I don't think a change like this would have a big impact on data quality though.

[Sir Ian Plätschisch]

Every method of bucketing is going to have some conceivable issue (for example, I'm Protestant, but I'm much closer to Catholicism than I am to some other denominations considered Protestant)

Consistency is probably best barring a very compelling reason to change.

[Breneir Tzaracomprada]

Every method of bucketing is going to have some conceivable issue (for example, I'm Protestant, but I'm much closer to Catholicism than I am to some other denominations considered Protestant)

Consistency is probably best barring a very compelling reason to change.

But you have more self-descriptive options, as a Christian, to describe your faith than a Jewish, Buddhist, or non-religious respondent despite the fact that those faiths also have denominations or sectarian subdivisions too. That is the compelling reason in my opinion. We have no practical need for the additional options for Christians and Muslims beyond general interest and the data are also not tied to any existing or currently planned programs that I know of. What is questionable, then, to me is that we would have more interest in a breakdown for Christian and Muslim respondents than any other faith. If that is not the case then there seems little issue with making the change.

I am less concerned with consistency between censuses because this is not a huge and disruptive change. I am more concerned with consistency as the options are presented to respondents as a matter of fairness.

[King Txec]

I'm curious why this wasn't an issue during the last census when you were the Seneschal but it seems like an issue now. I'm not casting any aspersions, but it does seem as if you might be looking for an issue when there really isn't much of one. Frankly, if this is really a big deal, we should just remove the religion question entirely if we really are making those of other faiths uncomfortable.

However, any data analyst will tell you that data must be consistent for it to mean anything, so changing the questions between censuses will give wrong conclusions. Of course, did we actually do anything at all with the data from the last census? The only effect I think it had in practice was maintenance of the citizenship rolls.

-Txec R

[Breneir Tzaracomprada]

I'm curious why this wasn't an issue during the last census when you were the Seneschal but it seems like an issue now. I'm not casting any aspersions, but it does seem as if you might be looking for an issue when there really isn't much of one. Frankly, if this is really a big deal, we should just remove the religion question entirely if we really are making those of other faiths uncomfortable.

However, any data analyst will tell you that data must be consistent for it to mean anything, so changing the questions between censuses will give wrong conclusions. Of course, did we actually do anything at all with the data from the last census? The only effect I think it had in practice was maintenance of the citizenship rolls.

-Txec R

It honestly didn't come to mind when I was seneschal probably because I was focused on matters related to working with you on the National Skills Survey if I recall correctly which was a great deal of fun. Unfortunately, that didn't end up developing into its intended purpose. I am not sure why we would rather remove the question entirely than just make it standardized in presentation. That is kind of a weird and overly dramatic response especially if it is not such a big issue, King. I am a former data analyst (five years in the Seniors and Disabilities Unit of the Department of Health) for the State of Alaska...and still don't see such a simple change as negatively impacting data quality.

[Sir Lüc]

No questions should be mandatory

This is already the case as per Lexh.C.1.2.2.1.

[Baron Alexandreu Davinescu]

Maybe let's not worry about people getting hypothetically offended on this one, so we can keep the data consistent.

[Breneir Tzaracomprada]

Maybe let's not worry about people getting hypothetically offended on this one, so we can keep the data consistent.

Sure, Baron. We won't worry about people getting offended on this one. :)

[Baron Alexandreu Davinescu]

Hypothetically offended, yes. If someone is legitimately offended, that will be different, and we can cross that bridge when we come to it.

[Breneir Tzaracomprada]

Hypothetically offended, yes. If someone is legitimately offended, that will be different, and we can cross that bridge when we come to it.

Sure, Baron.

[Sir Lüc]

Added questions on age and civil status (as on the previous census) and on education level (seems to make sense for a census).