The Lemon Fool

Currently reading the 1956 popular science book How to Lie with Statistics by Darrell Huff.

Seems pretty good so far, and it has a varied selection of applicable and interesting real-world examples.

plaguedbyfoibles wrote:Currently reading the 1956 popular science book How to Lie with Statistics by Darrell Huff.

Seems pretty good so far, and it has a varied selection of applicable and interesting real-world examples.

I haven't read it but I''m sure it has as much relevance today as it did then, if not more so. The burgeoning Data Science industry is all about telling stories it seems. Too much data can be as dangerous as not enough.

moorfield wrote:I haven't read it but I''m sure it has as much relevance today as it did then, if not more so. The burgeoning Data Science industry is all about telling stories it seems. Too much data can be as dangerous as not enough.

Information overload

I'm a human being. I am not a database. My memory isn't stable or reliable. It's human. I cannot digest 50 emails per day and recall the contents in 6 months time. When I had twenty sheep and two cows I was able to make a note of that in my book. Today I need to know

1. Where my twenty sheep were born
2. Who were their Mum and Dad
3. Where my twenty sheep were kept
4. When they were vaccinated
5. Where they were slaughtered
6. How far they had to travel to the slaughterhouse

Now I own a network computer and have one employee just filling in all the above details

But I still have 20 sheep

AiY(D)

Is that enough to fall to sleep to if you count them

AsleepInYorkshire wrote:
But I still have 20 sheep

Er, no you don't, judging by #5 above. You need a new employee.

pje16 wrote:Is that enough to fall to sleep to if you count them

I've counted them and including VAT I've got 23

AiY(D)

So VAT is still 15% in Yorkshire then

pje16 wrote:So VAT is still 15% in Yorkshire then

Nope A Yorkshireman never tells you where his last sheep is

AiY(D)

moorfield wrote:

plaguedbyfoibles wrote:Currently reading the 1956 popular science book How to Lie with Statistics by Darrell Huff.

Seems pretty good so far, and it has a varied selection of applicable and interesting real-world examples.

I haven't read it but I''m sure it has as much relevance today as it did then, if not more so. The burgeoning Data Science industry is all about telling stories it seems. Too much data can be as dangerous as not enough.

Yeah I've read that the recommended modern follow-up to this is The Tiger That Isn't: Seeing Through a World of Numbers, penned by a former chairman of the UK Statistics Authority (Andrew Dilnot), and the creator of the BBC Radio 4 / World Service programme More or Less (Michael Blastland).