Covid 19 reporting app - public webinar

[BuzzShitbagBobbly]

You may be aware of this reporting app. It's designed to "help slow the spread of COVID-19" by getting users to self report daily, based on a number of relevant factors (age, sex, location, symptoms etc).

HOWEVER, almost straight after releasing the first version of the app which simply asked for users' sex, the developers decided to add a pointless and potentially harmful "gender identity" question with no scientific justification or purpose whatsoever.

Recording someone as a woman [or a man] when they are the opposite sex skews facts and data, because we already know it affects each sex differently. Adding the potential for innocent OR bad faith actors to add error into what should be science-based research is an appalling way to proceed.

If any of the spokeswomen/experts from the various women's groups (Dr Nic? Dr FondofBeetles?) are free, the developers are holding a webinar this Thursday at 2pm BST

I'm sure they will be delighted to explain how knowing who has special womanly feelz will positively impact and add much-needed clarity to their research into this live and brutally fatal pandemic.

zoom.us/webinar/register/6915873908274/WN_X0KrlhfVTE-q7v539VKz8w

'we just want to be nice and include eveybody'.
We’ve already changed it once for one set of fuckwits, who shouted very loudly and we’re not going to change it back now, just because some women asked.

Theoretically, self-identified gender could be relevant. As in, it’s potentially possible that people who identify as female have more caring responsibilities or people who identify as male are less likely to visit the doctor. Absolutely no idea if this is the case but it’s theoretically possible if we accept that some female/male stereotypes relate to socialisation.
I think the big worry would be if they didn’t record biological sex. I can’t see the harm in recording both.

OK but, Covid Team, what do you normally do if people argue with the validity of your most fundamental data points after you’ve already collected analysed them? Do you go back and retrospectively change them so they don’t get ‘upset’?

Or do you say, sorry, we can’t change anything at all because that would compromise the validity of our findings?
Do you follow that up with saying that it’s essential to be as accurate as you can be with your results and to be as simple and straightforward in your questioning when collecting data from the public so you get both high levels of participation and that participants leave with an understanding of why they been asked the questions they have?

(So to cover that ‘why’ part off properly you actually should have provided an optional info bit on ‘gender and sex categories’ in the information sheet, before seeking consent to take part and before asking the survey questions. This info should have explained why gender identity isn’t relevant to this data set so isn’t going to be asked about and should have asked people not to complete the form at all if they don’t want to enter their biological sex information, which (you should set out) is an unchanging category regardless of subsequent gender identity feelings and of any change of legal sex status (GRC).

Then you should have explained that this is important because for a participant to give a gender category answer to a sex question would be obscuring the dataset for their own results but for everyone else as well, which could ultimately lead to misleading conclusions which could be dangerous in a time of global pandemic.

You could explain that the validity and ultimately the downstream effectiveness and use of any self-reported dataset relies on participants and researchers acting responsibly, having trust and solidarity with each other and with other participants. Which means that participants should only complete this data if they can respect that point which is integral to the aims of this research.

Helplines and further information if you need help and support in this area are .... and ....

They may as well ask for our favourite colour for all the use it is.

Oh I agree it's useless. But it's also harmless, unless it's a replacement for (rather than an addition to) biological sex. There are many only so many battles I can fight, and I'd rather save my energy on this one.

It's not harmless though.

It's inviting a major corruption opportunity of the data. One that is entirely avoidable.
In a life and death situation. With life and death implications.

Imagine if AGE had two questions.
Actual age.
And age you wish you were/age you believe you are mentally.

And the company invites everyone to complete age truthfully, and also encourages the collection of a false age too. Implying that factual age is probably not that important a data point.

And then, your first analysis released to the press says "21 year olds at higher risk than previously thought".

What does that mean? Will your audience instantly know that age no longer necessarily means age?

EVERY analysis will then have to be probed further. That's if people even realise that age no longer means what it always has because it's become acceptable to offer "age you want to be" in place of "age".

Real 21 year olds? Or imaginary ones? Are you sure your database hasn't accidentally pooled data from both fields together until it's irretrievably muddled? Does everyone at the company know that "21 year old" can mean two things: definitely 21, and also definitely not 21? Is everyone aware of the existence of the entirely false data and able to ascertain, given a variable such as "21 year old" NOT to make the mistake of treating that at face value, ever?

It's preposterous and dangerous to encourage ANY collection of false data in a pandemic, especially data that relates to a material biological characteristic, especially when we already know that biological characteristic is profoundly affecting outcomes already, and especially when the last ten years of UK administrative history reveal that EVERY organisation that has collected real AND false data 'for inclusivity' has gone on to conflate the two beyond repair.

Including THE NHS which can no longer tell you whether any given patient is ACTUALLY male or female, only that their 'current gender' says that they apparently identify as one or the other.

This exercise ALWAYS FAILS.
Telling yourself that you can and will keep the two types of data separate and never confuse them is futile.

It's like throwing a bucket of poisoned m&ms into a vat of good sweets, giving them a good shake and convincing yourself you'll be able to pick out the bad ones later.

The entire vat becomes unsafe.

You HAVE to preserve, INTACT, the meaning of female/male, girl/boy, and woman/man.
That means NEVER inviting any opportunity to collect data that says "women" that you know you've explicitly encouraged men to submit themselves as such.

It's been rank stupidity that has allowed data collection of 'female' data that knowingly includes males in the last couple of decades. It has irretrievably corrupted every important population level data analysis we have as a country.

But to continue to do this, eyes wide open, in the midst of a once in a century pandemic?

This is insanely dangerous right now.

Also, what are they doing about the millions of people who signed up before they put the gender question in?

Are they assuming that:
No trans people used the app
Or
That trans people used the app and recorded their sex
Or
Trans price used the app and recorded their identity
Or transpeople are not numerous enough to impact results anyway.

Yes barracker

Currently, a good number of people believe that TWAW/TMAM (or, at least, profess to believe that). It seems like they're making the best of a difficult situation by encouraging maximum participation.

I don't think anyone actually believes TWAW in the midst of a viral pandemic.

That kind of lie only survives as a luxurious artifact of an era where your life doesn't appear to depend upon whether you tell the truth about sex, or not.
It's just a virtual signalling lie that people are prepared to utter during a time when they are ok with pretending that all that matters is clothes and feelings. Because ignoring biology has consequences that you judge you can live with.

If scientists declare that a vaccine is safe for females and harmful for males, for example, I don't think the general public believe for one second that how a person identifies changes that risk.

We need unequivocal clarity right now.
The time for indulging bullshit about "women" meaning feelings not biology expired when whether you live or die is profoundly affected by whether you really ARE a woman or not.

Barracker, the option has already been added, the horse has bolted. Both sex and gender questions are present. It seems hypocritical to now complain about wasting time in a pandemic and then demand that the change be rolled back. The fact they're collecting both rather than changing the question demonstrates they acknowledge biological realities.

The NHS also tried to collect both.
They decided they would make sure they didn't conflate the two because it was critical to lives and health outcomes.
They produced risk analysis and hazard log documents to outline the risk of confusion.
However.
They failed.
The gender data subsequently supplanted the sex data entirely.
The sex data no longer exists.
It became impossible to undo.

There is no point at all declaring you understand biological realities if you are unable to keep unrealities from overwriting them.

The best way not to corrupt life and death data is not to invite people to lie when they provide data to you, and not to reassure yourself that you'll never mix up the women with the women when some of the women are women and some of the women are not women and both women and not women are held as data entitled women.

Even if I believed that this particular outfit could succeed in distinguishing sex from gender where the entire NHS database failed, despite their documented risk avoidance procedures and data information standards directives?
Even if they DID succeed, confidence in their data would be damaged.
When someone reports conclusions about male vs female data, knowledge that they explicitly invited and collected male data to be recorded as female and vice versa, means that the reliability of their conclusions will always be suspect.

Obscuring what can be learned from research data is at worst unsafe and at best makes the findings unreliable. It is often people’s whole job to ‘clean’ data etc to make sure this doesn’t happen. Yet these researchers did it on purpose.

It wastes individual participants’ time, because they think they’re contributing to something that can be reliably used. So that’s bad on an individual level but also bad for research participation because it makes people think it’s not worth taking part AND that the results of research in general can’t be trusted. That’s a very risky outlook for all of us and it’s really bad for researchers to contribute to it.

There is no point at all declaring you understand biological realities if you are unable to keep unrealities from overwriting them.

Which is precisely what just having a question about sex would likely achieve. If you have just a question of sex, it's likely that people who identify as a different gender from their sex will give their gender in that category. At worst, there will be two missed opportunities (during answering and during data processing) to separate sex from gender.

It wastes individual participants’ time, because they think they’re contributing to something that can be reliably used.

Wouldn't it be less likely to be reliable without separate questions?

Is it not significant though that some people might be taking hormone treatment which may, or may not, be significant?

So it appears that men are at higher risk than women but they don't know why. Is it due to hormones?

If so, would it not be important to know that s trans man might be taking testosterone or a trans woman taking oestrogen?

Basically, if the difference in outcome is due to hormones then isn't it valid to know that some people may not have the hormone profile of their birth sex?

Gronky I posted about that upthread. The researchers should have explained why they only need to ask one question, asking about biological sex.

The answer to that is likely to be covered in questions about medications.

There is no single hormone profile that exists as a template, especially in the female sex.
If men over 50 cared about emulating the female comparator for their age, they wouldn't take oestrogen at all. But of course, 65 year old men don't usually choose to transition into a post menopausal woman with low or undetectable oestrogen.

The question should be biological sex (with an explanation for those who willfully might choose to pretend they misunderstand), and medication / treatments.

Collecting data on males and registering them as females, or women, is a huge risk. Soliciting false data knowingly is worse.

Any risk/benefit exercise would knock this on the head instantly.

The researchers should have explained why they only need to ask one question, asking about biological sex.

I'm not going to say they had to do it the way they did but, if I were designing this survey, I'd probably opt for 2 simple questions, rather than a long explanation (which people are less likely to read). I understand that it sits better ideologically (no comment on the relative right/wrong/scientific accuracy of any ideologies in particular) but I think the way they did it will receive a greater number of more accurate responses. I've co-authored epidemiological surveys overseas and we're terribly pragmatic when it comes to cultural questions (getting small cross sections to give answers in isolation before interviewing them to identify pitfalls).

Collecting data on males and registering them as females, or women, is a huge risk. Soliciting false data knowingly is worse.

Has there been any indication that they're going to do this in this specific case? Bear in mind that the first edition they led with only asked the question of sex. They didn't drop it in subsequent editions, they merely added a clarifying question which, to me, is the simplest way (in terms of words and time taken answering) way to achieve greater accuracy, not less.

To be fair, I wonder how accurate or reliable the data collected is anyway. It's all self reported. I've been doing it for weeks but I could be putting any information into it. No one is checking that what I declare re symptoms is correct so how accurate is it,?

Bear in mind that the first edition they led with only asked the question of sex.

But Gronky they first asked for 'sex at birth' and then changed it to 'sex assigned at birth, making it less clear, then they added the gender question.

It's poorly constructed, no excuses for it.

they first asked for 'sex at birth' and then changed it to 'sex assigned at birth, making it less clear, then they added the gender question.

I'd say that's more meeting sensibilities than less clear. If they'd replaced sex with gender, I'd be the first to grab the scientifically-calibrated pitchfork.

No one is checking that what I declare re symptoms is correct so how accurate is it,?

I agree that it may not be terribly accurate in terms of giving robust numbers on infections but the data will be helpful in terms of looking for spikes in certain symptoms that exceed national averages as well as spikes which preceded confirmed infections. I've seen some very interesting and useful data come out of broad symptom reporting.

To be fair, the whole app is fairly pointless as it is being used by only a small, unrepresentative chunk of the population. They will get ok info on this chunk, but I don't see how they will be able to draw any broader population level conclusions.

if I were designing this survey

People who were interested in scientific facts probably drafted the very first version of this, with a view to getting it out quick, and improving it through user feedback and things they realised they missed themselves - pregnancy*, blood pressure, healthcare worker or not etc have all be added in for valid reasons.

But the gender ID question is simply to keep it in line with a US questionnaire (according to the update comment). We can all do ourselves an injury by desperately overreaching for some possible, logical scientific reason, but let's not give the emperor clothes he hasn't asked for. The lack of surrounding or supporting questions that could make this a valid q (hormones etc) show it to be nothing but a sop to those who would otherwise cause a nuisance about it.

And let's face it, its not like this is the first time this sort of takeover has happened - as the app says, they are following some previously set "standard" that nobody actually agreed to, it just sort of fell into usage.

I have just had a work call scheduled for tomorrow afternoon or I'd listen to the live webinar. I'm hoping they have a recording available afterwards.

(* I wonder what happens when the first pregnant man gets covid?)

Thank you Barracker for your clear analysis on this. Really helpful. Adding into something supposedly scientific a feature that has nothing to do with science will make us distrust the findings. And the NHS has made some significant mistakes on this, as many have said.

coronavirusexplained.ukri.org/en/article/cad0007/

This is a new COVID-19 explainer website by UKRI, nothing to do with the app we’ve been talking about. but I thought it was interesting to see because it carefully separates sex and gender. I don’t think it’s entirely successful in its definitions and could do with an edit- no reference to stereotypes and made up shit under ‘gender’ definition which apparently comprises men’, ‘women’ and ‘non-binary identities’. That doesn’t seem to work for gender critical people or for trans people.

It also doesn’t take account of where researchers (like the ones in the app we’ve been discussing above) deliberately obscure sex data collecting.

But still, the bar seems pretty fucking low these days and I am happy to see at least someone still writing in the view that sex is an actual biological reality.. even if it doesn’t quite hang together on the other hand because they don’t seem too clear about what ‘gender’ is.

‘ Sex and gender

Although sex and gender are often used interchangeably, they have separate meanings. Sex refers to the underlying biological profile of a person. It influences a range of bodily responses that are important in tackling infection or disease. Gender refers to the roles, behaviours, activities and attributes that any given society considers appropriate for men, women, and people with non-binary identities. Gender determines what is expected, valued or allowed in the behaviour of men and women.

Both sex and gender could have a role in explaining why men have worse outcomes of COVID-19 infection than women, but no conclusive studies have yet been carried out.‘