Statistics Without Context

by Neal Levene on Wednesday, August 5, 2009 · 3 comments

in Best of Simple Complexity, Statistics

I saw a great slide deck by Edward Kavanaugh titled Statistics Without Context on SlideShare. I have embedded the whole presentation at the bottom of this post. [Update: Ed Kavanaugh is entered in a presentation contest. If you like this post, please click here to vote for his presentation. There is a green thumbs up icon that allows you to vote. ]

First, what is context?

The context is the circumstances in which an event occurs.

If you are asked, “How many apples did you buy?”

“20% more than yesterday!” might be the answer from an annoying person.

20% is just a statistic. What you bought yesterday is the context. You need to know that 10 apples were bought yesterday to know that 12 apples were bought today.

Edward provides three examples of claims:

Claims: 1. Bacon increases risk of colorectal cancer by 20%; 2. One extra unit of alcohol a day increases a woman's risk of breast cancer by 12%, 3. Just 2 units a day reduces risk of heart disease by 17%

He then puts these statistics in context:

  1. 5 out of 100 people contract colorectal cancer.


  2. baconcancer1 Statistics Without Context


  3. If incidence rose by 20%, that would be 1 more person for a total of 6 (5 x 120% = 6)


  4. baconcancer2 Statistics Without Context


  5. So, “Bacon increases risk of colorectal cancer by 20%” is the same as saying it causes 1 more case in 100.


  6. baconcancer3 Statistics Without Context


Edward then goes through the same analysis for the other two claims. In the end:

statistical context

Unless you add some context to the percentage, there is no meaning.

You can see the whole presentation below.

Social Media Links:
  • RSS
  • email
  • Twitter
  • Facebook
  • del.icio.us
  • StumbleUpon
  • Digg
  • Google Bookmarks
  • Technorati
  • Sphinn
  • SphereIt
  • Mixx
  • LinkedIn
  • Reddit
  • Live
  • Netvibes
  • Yahoo! Bookmarks
  • Yahoo! Buzz
  • PDF
  • Print

We welcome your feedback. Please leave us a comment below. If you haven't already, there is no time like the present to subscribe to the RSS feed.

Category and Tags

This post filed in the following categories:

  • Best of Simple Complexity - The best posts of Simple Complexity as judged by the post authors.
  • Statistics - Statistics is a mathematical science pertaining to the collection, analysis, interpretation or explanation, and presentation of data.

About the Author

This post was written by Neal Levene, CEO of InnovaTech, Inc., who blogs about data and business issues here at Simple Complexity and about a variety of other topics at NealLevene.com. Find Neal on LinkedIn or follow him on Twitter. Neal is available to speak to your organization on a variety of topics. You may also use Simple Complexity's Contact Form.

Comments

{ 1 trackback }

Data visualisation « Liberty Resources
Tuesday, February 16, 2010 at 12:47 pm

{ 2 comments… read them below or add one }

1 Ed Kavanagh Wednesday, August 5, 2009 at 3:09 pm

Hey Neal,

I am a biiiig fan of the blog so its quite an honour to be included on it!
I really feel passionately about promoting understanding of statistics and I am quite tired of the constant abuse of numbers by the media in general.

For any slideshare users please click through and give it a vote for the presentation comeptition! (shameless plug over)

Ed

Reply

2 Bella Hurrell Thursday, September 3, 2009 at 9:43 am

If you liked this, then it’s likely you will appreciate the slideshow that helped to inspire Ed Kavanagh above. You can find it here http://news.bbc.co.uk/1/hi/magazine/7937382.stm
Written by Michael Blastland, the graphics were produced by BBC news website design team. It’s Michael’s mission to help us all make sense of numbers and his regular column on the BBC news website is a great read.

Reply

Leave a Comment

You can use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>

Previous post:

Next post:

WordPress Admin