Unleash Your Inner Healthcare Data

A free twice-monthly guide to understanding, presenting and using healthcare data, information and evidence.

Katherine S. Rowell @ Associates

Katherine S. Rowell @ Associates

CHECK IT OUT

Workshops & Seminars 

JOIN OUR LIST

Join Our Mailing List


 

View my profile on LinkedIn

In This Issue

March 18, 2011                               Volume 2 Issue 6
 

Greetings!


I know that pie charts are popular...but like my mother always told me, popularity isn't everything. 

 

In today's newsletter I explain the problem with pie charts and do my best to convince you that bar charts have a whole lot more going for them.

 

I also just finished reading Charles Seife's new book Proofiness--The Dark Arts of Mathematical Deception.  I think you'll love learning what the title means as much as I did!

 

Here is cheering in spring!

 

All the best,

Kathy Signature 

Katherine S. Rowell

Principal | Founder

 

P.S. The permanent link for this newsletter is here.

Bye, Bye American Pie (Charts)


I admit it - I didn't embrace the Hospital Consumer Assessment of Healthcare Providers and Systems (HCAHPS-pronounced "H-Caps") survey when it was first introduced.  I thought it was too simplistic to address the extraordinarily complex issues in our healthcare system.  Okay, truth be told -- I was a "complexity snob" and (hold the phone and mark the calendar) I was dead wrong.


As the first national, standardized, publicly reported survey of patients' perspectives of hospital care, HCAHPS data is an important and valuable tool in institution's efforts to improve the delivery of care.  It is also an important resource for patients who wish to understand how well an institution is performing on important measures such as communicating with patients - - which scientific studies and Dr. Phil remind us time and again are essential to improving everything in life.


But here is the problem -- even institutions with outstanding scores on physician and nurse communication don't communicate these survey results very well to the public i.e., the patients and families they serve.  Rather, all too frequently I find that institutions show this data on their public facing websites using pie charts.  And although it would be a bit over the top to say that pie charts are evil, it would be true to say that they are a bad choice for displaying data. 


The standard argument in favor of pie charts is that they show "part-to-whole" relationships optimally, but in fact a bar chart can do the job a whole lot better. 

 

Take a look at the following pie charts of patient satisfaction scores -- notice how you have to orient yourself to the colors and the labels and you have to look back and forth to make any sort of comparisons between the measures.  It is not easy to see at a glance where the institution is doing a good job and where they could improve.   

 

Additionally, the differences in the size of the parts of the pie charts are hard to compare -- you have to hold information in your mind and then do a bit of mental and visual gymnastics to try and discern variances and patterns.  

 

  

                                Pie Charts of Patient Satisfaction


And, did you notice another problem with the pie charts?  Look closely.  The labels got switched somewhere in the reporting of the information -- red, blue and green mean something different on each of the measures.  And because of the way the labels are formatted the pie chart for the measure "Communication with Physicians" is smaller than all of the other charts. 


As Edward Tufte wrote in his classic book The Visual Display of Quantitative Information, "the only worse design than a pie chart is several of them." 


Now consider the following display of the information in a horizontal bar graph. 

  

 

Pie Chart Redesigned to Bar Graph


By listing the categories of responses on the Y axis just once the graph becomes less cluttered and by repeating the same scale on the X axis for each measure the viewer is easily oriented to the information - it is clear and consistent across all of the measures. 

 

And now, at a glance, the viewer can see the ratio of responses for each measure and they can also see how in the overarching category of communication the institution is performing.  It is clear, simple and effortless.


So remember the sage words of my friend, colleague and data visualization expert Stephen Few to "Save the Pies for Desert." 

 

Oh yeah, there is a visual for that. 

Picture of Pie Eaten and Yet to be Eaten
Source: Edward Tufte August 9, 2009

 Reading Room

"Proofiness--the art of using bogus mathematical arguments to prove something that you know in your heart is true -- even when it's not." Charles Seife

 

Proofiness: The Dark Arts of Mathematical Deception

by Charles Seife 

  

 

Charles Seife's latest book Proofiness is a good and entertaining way to revisit important concepts about how to weigh supposed "proof" of complex issues. But I have to admit-- I think the thing I may like the best about Seife's new book is the title.


Although he never says so directly, the book's title alludes to "truthiness" - the Word of the Year in 2005, according to the American Dialect Society, which defined it as "the quality of preferring concepts or facts one wishes to be true, rather than concepts or facts known to be true."

 

ProofinessThe term was popularized by Stephen Colbert in the first episode of "The Colbert Report." The numerical cousin of truthiness is proofiness: "the art of using bogus mathematical arguments to prove something that you know in your heart is true -- even when it's not."


Seife, a veteran science writer who teaches journalism at New York University, examines the many ways that people fudge with numbers, sometimes just to sell more moisturizer but also to ruin our economy, rig our elections, convict the innocent and undercount the needy. Many of his stories would be funny if they weren't so infuriating.


But most important are the underlying mathematical issues that these stories educate the reader about including:


-- Falsifying numbers


-- Comparing apples with oranges


-- Cherry-picking data


-- Apple polishing (giving technically correct, but deliberately misleading, numbers.)


-- Potemkin numbers (phony statistics based on wrong or nonexistent calculations.)


-- Disestimation (giving too much meaning to a measurement, and not qualifying it enough.)


Although many of these concepts are familiar to me, the book was a terrific read and an even better civics lesson.  

 



We teach and advise organizations how to correctly capture and manage their healthcare data and how to communicate it clearly and compellingly.

For more details on our educational and advisory services:

Call us: 617-797-6471 | Email us: info@ksrowell.com 

Visit Our Website: www.ksrowell.com

Privacy Statement: Your privacy is important to us that is why we never share your information with anyone else ever-as in never-no way-no how.

Copyright © 2011, Katherine S. Rowell & Associates, LLC. All rights reserved.