Upgrade to Pro
— share decks privately, control downloads, hide ads and more …
Speaker Deck
Features
Speaker Deck
PRO
Sign in
Sign up for free
Search
Search
Numbers Need Context!
Search
clearpreso
September 29, 2011
1
99
Numbers Need Context!
clearpreso
September 29, 2011
Tweet
Share
More Decks by clearpreso
See All by clearpreso
Startup Gathering Presentation
clearpreso
0
42
US Visa Options for Business
clearpreso
1
41
TaskMessenger
clearpreso
0
42
Research at UCD Dublin
clearpreso
0
50
Presentations Matter: Simple tips for awesome presentations
clearpreso
3
250
FullHealth Medical
clearpreso
0
53
Presentations Matter
clearpreso
1
130
Don't Blame PowerPoint: It's just a vehicle
clearpreso
2
180
Presentations Matter!
clearpreso
1
95
Featured
See All Featured
実際に使うSQLの書き方 徹底解説 / pgcon21j-tutorial
soudai
121
39k
Git: the NoSQL Database
bkeepers
PRO
422
63k
Creating an realtime collaboration tool: Agile Flush - .NET Oxford
marcduiker
14
1.5k
Unsuck your backbone
ammeep
663
57k
Agile that works and the tools we love
rasmusluckow
325
20k
Ruby is Unlike a Banana
tanoku
96
10k
GitHub's CSS Performance
jonrohan
1025
450k
The Invisible Side of Design
smashingmag
294
49k
GraphQLの誤解/rethinking-graphql
sonatard
50
9.2k
Designing with Data
zakiwarfel
96
4.8k
Become a Pro
speakerdeck
PRO
11
4.5k
Fashionably flexible responsive web design (full day workshop)
malarkey
398
65k
Transcript
Context Numbers need
Although not everyone might agree, Statistics are…
Although not everyone might agree, Statistics are… Interesting
Although not everyone might agree, Statistics are… Interesting Educational
Although not everyone might agree, Statistics are… Interesting Educational Important
But it can often be hard to understand them without
CONTEXT
Our media abounds with context-less numbers
Let’s take a look at a few…
“Bacon increases risk of colorectal cancer by 20%”
“Bacon increases risk of colorectal cancer by 20%” “1 extra
unit of alcohol a day increases a woman’s risk of breast cancer by 12%”
“Bacon increases risk of colorectal cancer by 20%” “1 extra
unit of alcohol a day increases a woman’s risk of breast cancer by 12%” “Just 2 units of alcohol a day reduces risk of heart disease by 17%”
But what do the percentages really mean in real life?
And is there a more relevant way we could be
describing this information
“Bacon increases risk of Colorectal cancer by 20%”
But just how bad is that?
Lets assume that ordinarily 5 out of 100 people develop
Colorectal cancer.
Therefore if all 100 ate 3 extra rashers every day...
The number could rise to six
Therefore if all 100 ate 3 extra rashers every day...
The number could rise to six
So therefore…
So therefore… Bacon increases risk of Colorectal cancer by 20%
So therefore… Bacon increases risk of Colorectal cancer by 20%
Is the exact same as saying
So therefore… Bacon increases risk of Colorectal cancer by 20%
Is the exact same as saying About 1 extra case per 100 people
Next up…
“One extra unit of alcohol a day increases a woman's
risk of breast cancer by 12%”
But just how bad is that?
Let’s assume about 10 out of 100 women have breast
cancer in a lifetime
If all 100 drink an extra unit a day, that
number rises to 11.
If all 100 drink an extra unit a day, that
number rises to 11.
So therefore…
So therefore… One extra unit of alcohol a day increases
a woman's risk of breast cancer by 12%”
So therefore… One extra unit of alcohol a day increases
a woman's risk of breast cancer by 12%” Is the exact same as saying
So therefore… One extra unit of alcohol a day increases
a woman's risk of breast cancer by 12%” Is the exact same as saying About 1 extra case per 100 people
And Finally…
“Just two units a day reduces risk of heart disease
by 17%”
But just how GOOD is that?
About 32 out of 100 women have coronary heart disease
in a lifetime.
If all 100 were to drink 2 units of alcohol
every day… that number would fall to 27
If all 100 were to drink 2 units of alcohol
every day… that number would fall to 27
So therefore…
So therefore… Just two units a day reduces risk of
heart disease by 17%
So therefore… Just two units a day reduces risk of
heart disease by 17% Is the exact same as saying
So therefore… Just two units a day reduces risk of
heart disease by 17% Is the exact same as saying About 5 fewer cases per 100 people
So we have a little problem
Depending on the story
20% can equal 1 person in 100
20% can equal 1 person in 100 12% can equal
1 person in 100
20% can equal 1 person in 100 12% can equal
1 person in 100 17% can equal 5 people in 100
So…
Percentage figures without CONTEXT are meaningless and can often misleading
When talking about health matters it isn’t enough to deal
with percentages
Think about number of lives affected
And…
We must hold our media to account and ensure that
journalists do not go for the FLASHY, SHINY and TRENDY (not to mention misleading) sounding story rather than the more relevant and helpful one...
Because…
Those… “Who might rely on the news for guidance, deserve
something more relevant” (M. Blastland)
You can find me at: www.clearpreso.com www.twitter.com/clearpreso