BCA 2020 Demographics Report

[JoshW]

Based on what, 4 people on ukcaving not wanting to give it?

On my knowledge of statistics & the British Canoeing results from a membership more than 5 times that of BCA. The unknown errors will be much larger than the samples.

Wasn't aware of the British Canoeing survey, but from a quick google (feel free to correct me if I'm wrong) it looks to be a survey run separately to the membership renewals, something you were advocating as being a more effective way of obtaining data about 3 comments ago.

 

[mikem]

Yep, there is a lot more info in the document than just those demographics - but way more people have still said they don't want to answer those particular questions, which makes the results inconclusive (& the format makes it far harder to identify an individual than including it in membership request)

 

[nearlywhite]

Yep, there is a lot more info in the document than just those demographics - but way more people have still said they don't want to answer those particular questions, which makes the results inconclusive (& the format makes it far harder to identify an individual than including it in membership request)

I'm confused here - 51.1% answered both identifiers, which is more than 13.4% that gave no data at all?

EDIT: Ah I think you're on about the other report never mind.

 

[FionaH]

Cracking news everyone, the old white men have sorted it all out. "I see no problem" & "why bother collecting data?". Great job, we can all go home.

Christmas wishlist perhaps?
https://www.waterstones.com/book/how-to-be-an-antiracist/ibram-x-kendi/9781847925992

I suppose you don't see the irony of making a racist (and agist/ sexist) comment to support anti-racism?

Are you being systemically oppressed by cavetroll's comment, Pete?

I doubt it.

No, clearly I am not. Neither would it be systemically oppressive if you replaced the word "white" for the word "black" in the comment however I'm sure that nobody would argue that it was racist if you did.

Making a derogatory comment based on someone's skin colour is racist. This hardly benefits a discussion about inclusivity.

Reverse racism doesn't exist. Racism against white people does not exist. Racial prejudice does, yes. Racism does not. Racism operates on both the individual and the institutional (systemic) level.

As to your point about it not benefiting a discussion about inclusivity: okay.

 

[mikem]

Almost all of whom were...

(His whole point was it was inappropriate)

 

[2xw]

On my knowledge of statistics

Would you oblige me by using it to demonstrate what you mean? (to avoid the accusation of an argument from authority)

I'm not entirely sure how the survey you've quoted from canoeing is comparable to the single/few variables that Josh suggested.

Can you elaborate on which statistical methodology you would use to analyse a BCA dataset of approx ~2500 samples, and why you think the errors would be much higher than the samples? What is it you mean by error in this instance? How would you quantify error in this case?

What's the minimum sample size you deem to be effective, and why, and why does this preclude what is essentially effort-free data collection?

 

[mikem]

The survey asks those precise questions & the number of replies being anything other than white are far less than those not wishing to say - doesn't matter what criteria you want to apply (& they have a dataset of almost 3000 respondents, which is comparable to BCA membership). You can't accurately quantify the error as you don't have any further info.

 

[JoshW]

The survey asks those precise questions & the number of replies being anything other than white are far less than those not wishing to say - doesn't matter what criteria you want to apply (& they have a dataset of almost 3000 respondents, which is comparable to BCA membership). You can't accurately quantify the error as you don't have any further info.

at the very least it's clarified that there's only 10% of unknowns, whereas at the moment we're at 100% of unknowns because we haven't asked the questions?

it's very easy to find out how many people will say 'prefer not to say' and that's to ask the question. some information is better than absolute guesswork, and the actions taken will be adjusted based on the confidence in the outcome i.e. the % of prefer not to says.

 

[2xw]

And your hypothesis (conjecture, speculation) is that this will be the same with the BCA data but you are simultaneously unwilling to see that hypothesis tested? A sort of Schr?dinger's opinion: "I'm definitely right and we shouldn't bother checking".


And I assume general work arounds for this, and methods for quantifying that uncertainty, you are opposed to as well?

 

[mikem]

I think cavers being ornery buggers won't make things easier for you. I don't have any problem with you trying to collect it (apart from there being no way to do it at moment).

 

[JoshW]

I don't have any problem with you trying to collect it (apart from there being no way to do it at moment).

Then I?m not entirely sure what you were hoping to achieve by undermining the results before the bca has even decided to request the information let alone received it or analysed it..

 

[mikem]

Because I don't believe it will tell you anything new, due to the dataset being too small

 

[JoshW]

Because I don't believe it will tell you anything new, due to the dataset being too small

Yes but what were you hoping to achieve? 

 

[PeteHall]

Because I don't believe it will tell you anything new, due to the dataset being too small

Yes but what were you hoping to achieve? 

We seem to be going in circles here 

In summary then:
There may be a proposal to the BCA to collect more data about membership, specifically "protected characteristics".
The BCA council may approve collection of this data on a voluntary basis.
Some people might be upset by this and some might even leave the BCA (probably not many).
Some people won't provide the data (probably quite a lot)
Some people will be happy to provide the data (probably quite a few)
The data may show something, or it may not.
The BCA may decide to try to do something about it, or it may not.
This may have a positive impact on British caving, it may have a negative impact on British caving, or it may have no impact at all.

On all the above, time will tell 

 

[Badlad]

[gmod] Thanks for the summary Pete. This thread seems to have gone off on a tangent. The subject of the OP is an age demographic report. Some forum users find the direction of travel has become quite unpleasant. I'm going to lock this thread for a short period. When it comes back up please think carefully how you post. Thank you[/gmod]

 

[mikem]

The only reason discussion went on so long was because I was responding to requests from others for clarification, but I didn't want to attempt the maths on my mobile, late at night, so looking at the British Canoeing data for race:
25% of questionaires didn't include any reply (they had 2241 responses from 2990 returns)
90% said white (of British or other nationalities)
7.5% preferred not to say
2.5% other races (combined, as individually none was more than 0.75%)

So, you could say the number was 2.7% (2.5/92.5) or somewhere between 2.5 & 10%, but these only reference 75% of those involved, so all you can definitely say is the number is somewhere between 1.9% & 32.5% of members.

For arguments sake, next year 25% more people say they are other races, so 70 total, but 10% more have answered the question, so 3289 - does that mean you've had a 25% increase or (70/3289, a reduction to 2.1%) a decrease of 16%? There's no way of knowing.

I would be very surprised if we didn't get similar results. The survey doesn't say how many didn't respond to other metrics, but sexuality was 81% hetero, 14% won't say, 5% other (biggest being 2%), so still 3 times as many obscure results & why these sorts of questions often don't provide you with realistic or even relatively comparable data.

https://britishcanoeing.org.uk/uploads/documents/Membership-Satisfaction-Survey-Report-2019-VF.pdf

 

[JoshW]

For arguments sake, next year 25% more people say they are other races, so 70 total, but 10% more have answered the question, so 3289 - does that mean you've had a 25% increase or (70/3289, a reduction to 2.1%) a decrease of 16%? There's no way of knowing.

2 years of data does not give you a trend, either one could be an outlier from the trend. My point is the more data you have over more years, the more you can start to make judgements on it. I think possibly with about 4/5 years of age splits, the BCA are approaching a point at which it can start to analyse a possible trend if there is an ageing population.

I would also hope that once the membership see that there isn't some big conspiracy over the way that the BCA would use the data (if my possible proposal goes ahead), that you'd see more and more people comfortable with giving a small portion of information about themselves, and the 'prefer not to say' would decrease over time, giving further accuracy to individual years results.

That's not to say that there is nothing that can be gained from 1 year of data, there are some conclusions that can be drawn, just that they would need to be heavily caveated.

 

[mrodoc]

The demographics report coincides with what I see out in the field ie far more female cavers (my club committee is about the 75/25 ratio) who on the whole are younger than the male cavers and on the male side an increasingly older population of active cavers. 

 

[ChrisJC]

My point is the more data you have over more years, the more you can start to make judgements on it. I think possibly with about 4/5 years of age splits, the BCA are approaching a point at which it can start to analyse a possible trend...

That does seem reasonable. I guess as you get more and more data, maybe the errors get smaller? Even if the stat is that there is a consistent proportion of people who prefer to avoid answering the question!

One other factor to consider though is this:
If you had run a sexuality survey in the 60's say, you would find almost no gay people. In the 2020's, you would find a much higher proportion. That is not because the actual numbers have changed, but society has changed and it is now uncontroversial to be gay. So care is require to be sure that the stats are an honest representation of the underlying truth.

Reverse racism doesn't exist. Racism against white people does not exist. Racial prejudice does, yes. Racism does not. Racism operates on both the individual and the institutional (systemic) level.

I will take that as a joke that isn't funny.

Chris.

 

[mikem]

Certainly cavers are staying active into later life nowadays - I wonder how much this is affected by improvements in kit, which has been shown to have a positive effect on the retention of e.g. female canoeists / kayakers in their sport?

There is also the possibility that higher percentages of men continue to pursue caving outside the club environment, whereas women are more likely to join a club (this has been noted in other outdoor activities).

(To clarify - mrodoc's committee ratio is male to female & is pretty much the same as overall ratio in that club)

& if there is a consistent proportion of people avoiding the question then it doesn't affect the comparison, the problem is it won't be consistent for quite some time, as people will do it to make a point & then not worry some time subsequently.