The media and Covid-19: making a bad situation worse

Im noy going to defend the mail or any other legacy media outlet but all of those "fact checker" sites are nothing but political activists. You will find that any centrist or right leaning org is ranked poorly by them.
Quis custodiet ipsos custodes ... so reader beware .
 

Truxx

LE
Yes exactly. Im just going to put this out there that Barnard Castle is a town.

I went to bishops castle myself recently for work. Never saw any fortified buildings or site seeing tours, just a town.
But Barnard Castle does have a very big castle. But if he had been there he would have gone to Barnard Castle Castle.
 
But Barnard Castle does have a very big castle. But if he had been there he would have gone to Barnard Castle Castle.
99% of people will have heard Barnard Castle and assumed that he had gone to the tourist attraction, funnily enough the media did nothing to discourage that.
 
As many of you many know I'm a statistician in my day job. One of my biggest gripes during this crisis is the media deliberately misleading the public with data. Some of this is ignorance (e.g. not knowing that different countries only reporting hospital deaths etc). However, some of this is deliberate. Take the example below as illustrated by a fellow statistician. The 'analysis' was written by two people one of whom has a masters in economics, another of whom has a masters in data science. The mistake they make is so elementary (I would expect even my undergraduates not to make such a mistake), that my conclusion is that they are deliberately misleading readerss, with their analysis.
1590957244234.png
1590957244234.png
1590957265866.png
1590957244234.png1590957265866.png
 

Slime

LE
99% of people will have heard Barnard Castle and assumed that he had gone to the tourist attraction, funnily enough the media did nothing to discourage that.
I’m sure you are aware of some of the images of ‘Barnard Castle’ that the Beeb released as screen fillers while they discussed the visit.

Rather than do nothing to discourage it I’d suggest their still pics were to promote the idea.
 
As many of you many know I'm a statistician in my day job. One of my biggest gripes during this crisis is the media deliberately misleading the public with data. Some of this is ignorance (e.g. not knowing that different countries only reporting hospital deaths etc). However, some of this is deliberate. Take the example below as illustrated by a fellow statistician. The 'analysis' was written by two people one of whom has a masters in economics, another of whom has a masters in data science. The mistake they make is so elementary (I would expect even my undergraduates not to make such a mistake), that my conclusion is that they are deliberately misleading readerss, with their analysis. View attachment 478295View attachment 478295View attachment 478297View attachment 478295View attachment 478297
If I follow the twitter thing correctly, the FT claimed a 'significant' result with a causal probability (badly phrased) of less than 17%? Whoops.
 
If I follow the twitter thing correctly, the FT claimed a 'significant' result with a causal probability (badly phrased) of less than 17%? Whoops.
Exactly, he said that the relationship was significant when there was a 17% chance it wasn't.. As abit of context here, so at my university getting above 40% doesn't gurantee a student has passed the point, the student needs to demonstrate through their answers they had decent understanding of the material. If someone were to get 45% and draw conclusions like this, we would probably fail the student...
 
Exactly, he said that the relationship was significant when there was a 17% chance it wasn't.. As abit of context here, so at my university getting above 40% doesn't gurantee a student has passed the point, the student needs to demonstrate through their answers they had decent understanding of the material. If someone were to get 45% and draw conclusions like this, we would probably fail the student...
It's been a long time since I had to deal with p values (I can't remember the difference between them and confidence intervals) but I would have thought a p value of 0.16... meant that was the probability of a causal relationship? The magic number of p = 0.95 is still stuck in my head as the target value to be able to call anything a 'significant' result.

Unless I'm reading a lot into a typo.
 
It's been a long time since I had to deal with p values (I can't remember the difference between them and confidence intervals) but I would have thought a p value of 0.16... meant that was the probability of a causal relationship? The magic number of p = 0.95 is still stuck in my head as the target value.

Unless I'm reading a lot into a typo.
so a p value of 0.05 or below is considered significant. A p value of 0.05 means there is only a 5% chance that there is no relationship. The lower the p the better, usually in any scientific publication you need at least 0.05 to get your results deemed significant for publication.
 
so a p value of 0.05 or below is considered significant. A p value of 0.05 means there is only a 5% chance that there is no relationship. The lower the p the better, usually in any scientific publication you need at least 0.05 to get your results deemed significant for publication.
Cheers, knew it was a 95% probability for significance but got it the wrong way round.
 
so a p value of 0.05 or below is considered significant. A p value of 0.05 means there is only a 5% chance that there is no relationship. The lower the p the better, usually in any scientific publication you need at least 0.05 to get your results deemed significant for publication.
I just grew a beard, lost my social life, wife, kids, sex life, dignity and gained some 'amusing' Disney themed T-shirts and halitosis after reading this and partially understanding it...


Cheers, mate.

Now eagerly perusing Amazon for a veritable selection of stationery to cram into my top left shirt pocket.
 
I just grew a beard, lost my social life, wife, kids, sex life, dignity and gained some 'amusing' Disney themed T-shirts and halitosis after reading this and partially understanding it...


Cheers, mate.

Now eagerly perusing Amazon for a veritable selection of stationery to cram into my top left shirt pocket.
That's a very sarcastic reply for someone so obsessed with how people put things on plates ;)
 
I just grew a beard, lost my social life, wife, kids, sex life, dignity and gained some 'amusing' Disney themed T-shirts and halitosis after reading this and partially understanding it...


Cheers, mate.

Now eagerly perusing Amazon for a veritable selection of stationery to cram into my top left shirt pocket.
haha, just realized that there is also another fairly obvious point that should be clear to all. The sample size they have used in their 'analysis' is way too small. This means that even just one country having dodgy data can skew the results massively.
 
MEDIA: 2 million in the shielding group released from lockdown!

GOVERNMENT: Thats not what we said, they can sort of go out, but be careful!

DOCTORS: Don't go out! You're still in the top few percent of most vulnerable!
 
That's a very sarcastic reply for someone so obsessed with how people put things on plates ;)
Does that mean that @supermatelot was a Plate Layer's Mate during his time in the Andrew? Who knew!

Sent from my SM-G973F using Tapatalk
 
A wonderful piece of special pleading in today's Guardian:


Check out her CV at the bottom of the article - she has basically spent 30 years doing nothing...
Never mind her jobs:
"Interests: Playing the accordion with Oxford band the Mother Folkers."

Burn her!
 

TamH70

MIA
Never mind her jobs:
"Interests: Playing the accordion with Oxford band the Mother Folkers."

Burn her!
As long as she weighs the same as a duck.
 

New Posts

Latest Threads

Top