Profile picture
Sergio Graziosi @GraziosiSergio
, 25 tweets, 9 min read Read on Twitter
Yesterday @bbcnickrobinson made a disturbing mistake, while interviewing Jean-Claude Trichet about #Brexit and financial services for the Today Program (@BBCR4Today)
The mistake boils down to broadcasting demonstrably false information, I assume @BBCNews will want to rectify.
The details.
The offending claim was made shortly after 1:15:03 (running time). Full audio is here:…
@bbcnickrobinson " is a year to go until Britain is due to leave the EU and there is no evidence that Britain is changing her mind.."
Nick's claim is unquestionably false, but concurrently aligned with what the current government would like us to believe.
Thus, the growing suspicions that part of the BBC is becoming a mere mouthpiece for the government appear disturbingly founded. I expect a retraction.
The evidence:
Nick's claim is false, because the best evidence available points to the fact that Britain's population is slowly but unquestionably changing its mind.
The best source of such evidence comes from You-Gov Trackers:…
This evidence is the best we have because the shift in opinion is slow but steady, You-Gov's repeated polling at different times allows to detect trends while reducing the impact of sampling errors.

➡️The trend is clear!

(source:… - graph is mine).
Multiple polling dates allow to compare mean values across given periods. In this graph I'm showing the average difference ("to leave was right" minus "it was wrong"), pooled in groups of 3 months, with error bars around the 95% Confidence Intervals.
Once again, the trend is *impossible to miss*, and the pooling allows to estimate how confident we can be about it.
Result: the best evidence available clearly shows that Britain is changing her mind. Saying the opposite is false, saying we don't have evidence is also false.
Saying such falsehoods during a recorded interview and then broadcasting it within one of the most prestigious (and supposedly trustworthy) public news programs is the kind of error that REQUIRES a *visible* correction.
This is what I am asking the @BBC to provide.
It goes without saying that the public service should never broadcast false information and depict it as true.
I want to believe this was a genuine mistake, and therefore assume that the BBC will be happy to officially rectify as soon as possible...
PS I will provide a technical explanation of the figures included shortly (in a sub-thread) along with access to the spreadsheet I've used to produce them ('cause I'm a geek, and proud of it 😎).
In the main thread above, I make a few claims and show some digested data. I think it's worth expanding on both. Reasons: transparency and accountability.
My underlying reasoning might be wrong, in which case I would like to be corrected!
The main claim I want to clarify is that "YouGov's data is the best evidence we have".
I say this because it's the only data I know of which consists of frequently repeated polls, all using the same question and methods. This comes with significant advantages...
First and foremost, YouGov's data allows to spot trends. The methodological uniformity justifies comparing like with like. Even if we don't trust (we shouldn't) the absolute numbers, we can and should trust the trends they expose. The data does show the direction of change.
Secondarily, the fact that single polls are repeated *frequently* allows to make a rough (approximate) estimation of how precise they are. Other short-term factors probably contribute to the measured variance, thus, what we calculate is an overestimation of our uncertainty.
We end up having a measure of the Maximum level of uncertainty, which is nice, since even when we look at the confidence intervals (second pic), the trend is clearly visible.
[Note: it's also interesting that the confidence intervals get broader along the way!]
Moreover, since we know there is a visible trend, the my calculations mean that we are underestimating the variance even more (some of it is due to the trend, and I did not try to account for its contribution).
I think that this makes my analysis *more robust*.
That's because I am deliberately overestimating uncertainty. Thus, we reach the details of the second pic. You can access the original spreadsheet here:…
It includes the data, data source and a link explaining the maths I've used.
In my opinion, these considerations make the pooling of sequential polls justifiable.
To remain as rigorous as possible, I've used weighted means (based on Sample Size of individual polls), which does make the Maths more exotic (hence the need to make my analysis public).
Overall, considering that I've biased my analysis so that it makes trend-spotting harder, I think the evidence is very clear.
@bbcnickrobinson's claim is false and needs to be corrected.
The strange thing is that he could have said "[...] there is no evidence that the UK government is changing its mind". This would have been broadly correct, unfortunately: the government is trying hard to pretend that opinions aren't changing!
That's why I had to do this:
It is unacceptable for the BBC to help the government in hiding the truth. It is exactly the opposite of its main role in British society.
@BBCNews exists to inform the public, not to spread politically-motivated misinformation!
We cannot stop anyone from applying some subtle spin, but we can and we must point out all the cases where it stops being spin and becomes a broadcast of (demonstrable) falsehoods.

It worked in the past, so let's see how it will go this time!

After 17 hours, still no response from @bbcnickrobinson.
Nick, I know you are a busy man, and that I am being annoying, but this remains important: I would like to hear your thoughts. Thanks.
Hello @bbcnickrobinson. Apologies for being a pain. Could you please address my points in the thread above?

I still believe that @BBC journalists prefer to correct themselves when they happen to make a mistake.

Why this prolonged silence?
this thread started on Thursday.
It is now Sunday.
No reply from either @bbcnickrobinson, @BBCr4today or @BBCNews :-(
Missing some Tweet in this thread?
You can try to force a refresh.

Like this thread? Get email updates or save it to PDF!

Subscribe to Sergio Graziosi
Profile picture

Get real-time email alerts when new unrolls are available from this author!

This content may be removed anytime!

Twitter may remove this content at anytime, convert it as a PDF, save and print for later use!

Try unrolling a thread yourself!

how to unroll video

1) Follow Thread Reader App on Twitter so you can easily mention us!

2) Go to a Twitter thread (series of Tweets by the same owner) and mention us with a keyword "unroll" @threadreaderapp unroll

You can practice here first or read more on our help page!

Did Thread Reader help you today?

Support us! We are indie developers!

This site is made by just three indie developers on a laptop doing marketing, support and development! Read more about the story.

Become a Premium Member and get exclusive features!

Premium member ($3.00/month or $30.00/year)

Too expensive? Make a small donation by buying us coffee ($5) or help with server cost ($10)

Donate via Paypal Become our Patreon

Thank you for your support!