Data Collection

[TartanCurtain]

I want to feed back on this data collection but am struggling to articulate the issues.

I cannot understand that these were the options (you can only choose one) -

Man (including trans man)

Woman (inluding trans woman)

Transgender

Non Binary

Other/Prefer not to say

I think I find the first 2 options annoying as they present a point of view that I don't feel comfortable with.

But I don't understand why you could be trans and choose one of the top 2 or the third.

How does this help the people collecting the data? What would it tell them? Not much that could be relied upon I don't think?

They’ve gone wrong by conflating two protected characteristics from the Equality Act into one question. Sex and Gender Reassignment are separate protected characteristics.

To make their data comparable with other data sets they should ask one question about Sex (m/f). Everyone has that protected characteristic. Then a separate question asking whether the individual has the protected characteristic of Gender Reassignment. Not everyone does. If the individual believes they do there should be the option of saying what it is.

Making data comparable is important. There’s no point in saying you have 50% of users in some category if the category isn’t clearly defined and matches what everyone else is measuring.

I think I find the first 2 options annoying as they present a point of view that I don't feel comfortable with.
They are not just annoying but will collect non-useful data.

If you want to give them feedback, point them in the direction of the Recommendations in the Sullivan Review: https://www.gov.uk/government/publications/independent-review-of-data-statistics-and-research-on-sex-and-gender/review-of-data-statistics-and-research-on-sex-and-gender-executive-summary#recommendations

OP was there a sex category question? Or was this question the only one asked in relation sex?

There was an acronym used in computing, GIGO. It stands for garbage in, garbage out. It would seem to apply here. I can't think of any useful data that would be collected here. It's a big ol' virtue signal though.

That does seem like a very poorly thought out question. As you say the option there of choosing ‘transgender’ becomes confusing when there are three other options that are transgender identities.

perhaps they should have made that option ‘other transgender identity’. Either way, even if it was coded with clearee option, unless it has a sex category question, it could be meaningless data being collected.

I don't know if it casts any light on the thinking behind this, but, when the Sullivan Review came out, I was told the following by a trans friend who is fairly well up on the ideology:

Sullivan is transphobic so her research is worthless.

Sullivan is wrong to propose a ban on collecting data about acquired gender (NB she didn't).

Sullivan is wrong to propose that sex data, if collected, should always include birth sex, because trans people prefer to lie rather than reveal their birth sex.

The correct information to collect, in relation to sex, is birth sex, unless the responder has an acquired gender, in which case they should respond with their acquired gender. It should be a simple M or F and not broken down further, unless the research is about trans issues. So the first two options in the OP are 'correct'.

When I asked what to do if birth sex was relevant to the objective of the research - say, to work out demand for mammography, the answer was: some transwomen have breasts. Therefore, (women + transwomen) acts as a good-enough proxy for (people who need mammography). Ideological purity is more important than precision

This is an issue. As seen by the census.

What any data collection agency can only hope for is that the anomalies that those false answers cause is insignificant. Ie that there are so few that it won’t make a difference. Obviously, a full population audit such as the census will have significant issues as we saw.

In the past, I was taught that at times a data check could be done if contact details were available. ie, you could call a random selection to check their answers but also get deeper insight. That could be done if it was important data, with an eye to find out how many people had replied ideologically vs accurately. Ie. if a % was identified as being ideologically answered, that % is applied as a weighting to the results.

But that is an expensive exercise and unlikely to happen.

Spot on @Helleofabore. It's also really important to have accurate sex data in a healthcare setting. So many symptoms and disease risks differ between men and women. The range of what is 'normal' in test results is also determined by sex, not gender.
There's an excellent Women's Rights Network report that spells out the GIGO (credit to @lcakethereforeIam ) mess of NHS patient records are now in.
https://www.womensrights.network/nhs-data-report

This was the only one.

There were questions re sexuality and ethnicity separately but no others on the sex/gender topics

This is my feeling.

It won't tell anyone anything about reaching certain sex groups or whether they are reaching trans people. Because you can't separate any of it out into those individual groups reliably.

It just gives the impression of being 'into all this' but actually doesn't even serve trans people well. And makes people like me feel uncomfortable (I know us Hags don't matter though).

Its wrong for all the of the reasons others have already posted.

But also as part of the reason for collecting data is to compare the data to the census figures which determined the primary question should be about sex, ie biology.

And only as an option anyone's gender.

So as a statistical exercise it is worthless.

And I think there is even a paper about this on the EHRC web site about collecting data.

So not just a feeling it is wrong, it is wrong.

Q1 What is your sex? Male / Female
Q2 (optional) Do you have a gender identity? Here you can add in the whole alphabet options, male, female, non binary,

Similarly sexuality (assuming there is a valid point to collect it).

Are you opposite sex attracted ie heterosexual
Are you same sex attracted ie lesbian or gay
Are you bisexual
etc

Can you say who it is that has created this data collection.

I do wonder if this garbling of language is intentional. But given the choice between a conspiracy or a cock-up, it's usually a cock-up and that's gender woo in a nutshell. Utter bollocks!

Man (including trans man)
Woman (inluding trans woman)

My biggest issue is they have mixed the sexes here.

Man (including biological females)
Woman (including biological males)

If they were collecting data for marketing, e.g. to try and sell clothing, they might want the first two categories to know the kinds of clothes their customer-base want (masculine or feminine styles), and possibly also the 'non-binary' for gender neutral clothing, and the 'not saying' category, but the 'transgender' category is just rubbish.
Clutching at straws here.

You are struggling to articulate because it's a nonsense question. If a survey does not ask sensible question I answer prefer not to say (which should really be withheld). Apparently those who work in EDI hate it when surveys come back with all prefer not to say. If there is a place for free text anywhere I will say you can't be bothered to ask question about my sex so I'm not filling in your survey. It seems petty, but how else will they know! I will not comply with this nonsense. Men are not any type of woman.

It's a very badly designed question. Ignoring the ideological slant, the answers don't tell you much. Even if you wanted to group people based on gender identity, some trans women may answer 'woman including trans woman' and some may answer 'transgender'. You don't know how many of the 'transgender group are trans women. So you have no clear answers on how many people identify as women or men. Or how many are trans women or trans men or trans overall, if you want to know that. You also have no clear answers on how many people of each sex there are, as the 'women including trans women' includes both biological sexes. So no matter what your beliefs about sex and gender and what categories you are interested in, you get very limited information. It seems to be designed more around appeasing people filling it in than actually collecting any useful information. And it probably even fails at that, as 'transgender' doesn't fit in the categories promoted by LGBTQ organisations, and 'woman including trans women' will not make GC people happy. All in all, a fail.

I saw this drop down field at work this week and nearly lost my shit. Women and transwomen are totally different.

I wasn't sure which to complain to so I've parked it. It doesn't help that one of our IT team is trans so I'm trying to tread diplomatically.

We had a 'none stated' option so for the time being I'm logging everyone as that.

You are wise to tread carefully.
Can you frame any pushback in terms of the importance of data integrity and reference the Sullivan Review?
It would also help to quietly sound out colleagues to seek out any supporters. Organisations find it easier to dismiss an individual with concerns than it is two or three with similar concerns.

Idk why they've got 'transgender' as a standalone option while simultaneously telling trans people to include themselves in the sex they wish they are rather than birth sex.

So a trans person could be two of those things? But a regular person isn't really any of them? 😕

You might want to refer them to this guidance from the OSR https://osr.statisticsauthority.gov.uk/guidance/collecting-and-reporting-data-about-sex-and-gender-identity-in-official-statistics-a-guide-for-official-statistics-producers/ which is very strongly worded about collecting Sex/Gender data only with a clear idea of why you are collecting it, and ensuring that it is clear to users which box to tick.