Message Board

Looking for comments

MikeC — Fri, 16 May 2008 03:33:19 GMT

I am really interested in hearing any comments on the attached report.

Some background, this report is generated in MS Access and the purpose is to show the breakdown for the sources of wiring related defects identified in a certain time span. The bottom third of the report shows a trend of total defects as well as a trend of defects by each source. These trend charts have a target displayed as well as the icon used by the company to indicate the performance (red X for target missed by >20%, yellow triangle for missed by <=20%, green circle for on target)

Any comments and/or suggestions for improvement would be greatly appreciated.

Forum: Graph design

Bullet graph name

xan — Wed, 14 May 2008 17:35:48 GMT

Does the name "bullet graph" have an etymology that might aid localization? I always assumed it was related to the projectile sense of the word, but it's hard to see an obvious relation.

I explained it to a translator as showing the path of a bullet that was fired into a cylinder of thick gel. Anything better? Or more to the point, are there any existing translations of "bullet graph"?

Forum: Graph design

Classic example

grasshopper — Fri, 09 May 2008 15:09:17 GMT

This one's nothing earth-shattering, but I think a classic example
showing that even a simple-looking graph can be deceptive,
based on how the axes are scaled.

In this graph, it looks like there is a huge increase in "social networking"
job trends (not sure exactly what that is, btw) - but if pay careful attention
to the y-axis, you'll notice that it's still less than .1% (note that's 1/10th
of 1% ... not a value of .10 or 10% ... which is yet another subtle
deception of this graph).

From the following 'indeed.com' webpage:
http://www.indeed.com/jobtrends

(see attached jobgraph.png)

Forum: Examples of bad graphs and dashboards

Readers with low vision

mbarnum — Mon, 28 Apr 2008 17:22:43 GMT

Chapter 8 - Table Design gives all sorts of useful techniques to make table data more easily readable or scannable. One technique--fill color in alternating rows--is recommended in wide tables to aid scanning across long rows of data.

I have a colleague with muscle-imbalanced eyes who says that fill color isn't just helpful in wide tables, but in all tables four rows or larger. She says users with low vision will struggle to read the table if the fill color isn't there (among other things like large text, proper alignment of numbers, etc.).

I would think that it would depend on the table. I wouldn't think a two column table with 10 rows and sufficient white space would need a fill color, but add another column or two, and yeah--you'd probably need a fill color.

Two questions: 1) Do you agree with my colleague? and 2) How did low-vision readers factor into the research you did for SMtN?

Thanks in advance for your answer...this book has been tremendously enlightening for me already.

-Matt

Forum: The book "Show Me the Numbers"

Reverse Y axis

spilsburt — Mon, 28 Apr 2008 14:07:06 GMT

Hi,

I am creating a simple line chart and would welcome some advice on whether to use a reverse scale on the Y axis.

The chart I am producing needs to show that a reduction in value is the desired 'direction of travel'.

To do this in the past I have put an image of an arrow pointing down and the word "improving" next to it.

However, someone has suggested reversing the scale so that the largest number is at the bottom of the axis and zero is at the top of the axis.

I think they both work but believe that an increasing value should always go up regardless of whether or not it is the desired direction.

I welcome any alternative views.

Thanks

Forum: Graph design

Get Inside Your Data (?)

grasshopper — Fri, 25 Apr 2008 19:01:05 GMT

If pie charts aren't bad enough...
If 3d pie charts aren't bad enough...
What about 3d pie charts with floating slices!?!

Apparently this is an image from the virtual/walk-through "Second Life"
game thingey. The person who posted to the blog theorizes...

"At the moment, its just cool to clamber all over, but I have an inkling that this is an effective way to understand statistics. I can imagine walking through interlinked data and seeing relationships between figures represented spatially ..."

http://virtual.librariesinteract.info/page/2/

Hmm -- maybe this *is* the way the new generation will be learning
things? I know we're losing a bit here because we're just looking at
a static snapshot, but I can't imagine that this would add to the
understanding of the data, better than (say) a good-old bar chart.

Just call me an old fogie, I guess! :)

Forum: Examples of bad graphs and dashboards

Dare to Compare

wd — Thur, 24 Apr 2008 13:56:21 GMT

Found this one in a communication seminar presentation. The quest is to compare current with essential for a list of 40 questions grouped by theme (as depicted by the basic hue). Answer values range from 1 through 7. How it begs for side by side bars!

Forum: Examples of bad graphs and dashboards

Need suggestion for mapping probability

metman — Wed, 09 Apr 2008 21:00:35 GMT

I'm looking forward to learning a lot here - it looks like a very supportive group of people! Suggestions appreciated!

My challenge is to show probability of an event at points across a map. Initially it is simply probability of a threat (say thunderstorms). Also want to show probability of a flying category (5 categories). I've got the message loud and clear that pie charts are bad, like in the attached sample. But I have not seen a good way of doing it. In the simple chart I want to avoid colours because we generally use and are familiar with "red is bad, yellow is heads-up, and green is good". In the multi-category sample attached, that colour scheme is intentional. Spatial arrangement is important - at a glance the user needs to see where the trouble spots are, or the chance of it.

Thanks!

Forum: Graph design

Motion charts provided by Google

jannepyykko — Wed, 26 Mar 2008 16:08:59 GMT

Hello,

I want to raise discussion about the new "motion chart" concept that Google just announced. The concept itself is familiar from GapMinder.org that Google bought last year. Now Google gives it for use for everybody.

Heres my first trial with the tool (its a static image, available here with motion):

I think the motion chart concept is great when exploring data. It helps understand phenomena that change over time. For example, I just wanted to play with the tool, put some macroeconomic data into it, and accidentally learned a new fact: since 1970 big European countries (UK, Germany, France, Italy, Spain) have adopted policies that make them more alike (bubbles moving near each other from starting positions).

What this really means is this: From now on we will see motion charts appearing more often on various web pages. Its time to discuss about the feature so that people can utilize it the best way!

Regards,
-Janne

Forum: Graph design

Dual-Scaled Axes in Graphs

bpierce — Tue, 25 Mar 2008 18:18:55 GMT

In his March 2008 Visual Business Intelligence article, Stephen Few wrote about the use of dual-scaled axes in graphs (for instance, a line graph that has a scale measured in dollars on its left side and a scale measured in units sold on its right side, with separate data corresponding to each). Stephen said that, originally, he accepted graphs with dual-scaled axes as a potentially useful solution. However, as time passed, he began to investigate dual-scaled axes more, and eventually came to the conclusion that they offer no unique benefits, but they do introduce several problems. In the article, Stephen said that he can think of no case where they are more effective than other methods, but he asked that people propose exceptions.

After reading the article, can you think of any cases that he did not address, where a graph with a dual-scaled axis would actually be the best solution?

Forum: Articles by Stephen Few

Redesigning an ExcelUser dashboard

grasshopper — Tue, 25 Mar 2008 12:50:44 GMT

I have to admit, although they violate a few of the Tuft-isms and Few-isms,
(in particular, the use of bold colors) Charley Kyd's ExcelUser dashboards
have the most visual appeal, imho, of any collection of dashboards I've seen:

http://www.exceluser.com/catalog/landdash1.htm

But I decided to see if I could reproduce one of them using a different software, and in doing so to also see if I could "improve" any of the graphics used in it. I chose the following dashboard:

http://www.exceluser.com/dash/xbw2.htm

In my version, I left the layout & colorscheme as-is, and only made
changes in the "graphics". Here are the things I changed:

I added html charttips (flyover text) to the two bar charts with 4 big bars,
so you can hover your mouse over them (in the html output) and see the
actual data values.

I changed the order of the columns in the tables of data, so that the
year is always on the left, and the month is always on the right
(so they are in the same order as the bar charts in the top section).

In the '%' increase/decrease values printed under the bar charts and
the food pictures, I made positive values green and negative values red.

I added an html drilldown on the "Mel's Diner" text, so you can click on it
and see Mel's Diner's webpage.

I changed the labels from "52-week" and "last 4 wks", to "Past year" and
"current month".

And in the "Stores" section, the Budget & Actual bar graphs just didn't
make any sense to me (not sure I understand budgeting -5% growth?)
So I changed that to show the actual-growth vs target-growth,
and used a bar-and-target chart (somewhat similar to a bullet graph)
rather than a paired-bar chart.

Here's my version (if you want to see the html charttips & drilldowns
then you'll have to click the link and view the version with the html
overlay):

http://robslink.com/SAS/democd31/restaurant.htm

Forum: Graph design

Re-designing IMF's "Mortgage Rate Reset" plot

grasshopper — Mon, 17 Mar 2008 12:21:55 GMT

The following graph caught my eye (both the look, and the content)...

   http://www.housingwire.com/wp-content/uploads/2007/10/chap1-8.gif

It's from the following IMF report:

   http://www.imf.org/external/pubs/ft/gfsr/2007/02/pdf/chap1.pdf

And, of course, I decided to try to improve upon it :)

Although there are arguments against stacked bar charts, I decided to
leave that as-is, and just try to make it the best stacked-bar possible.
Here are some changes I made:

   http://robslink.com/SAS/democd31/reset.htm

+ I labeled *every* year along the bottom, rather than just the odd ones.

+ I changed the tickmarks per year so they point down instead of up,
   so they are more visible dividers of the years.

+ I put tickmark values/labels along the left, rather than right.

+ I added a very light gray refline grid behind the bars.

+ And, I added html flyover-text so you can hover your mouse over
   the bar segments, and see the actual data values.

- I thought about changing the colors, but opted to leave them as-is,
   since the IMF's original document uses this "color scheme".

---

I'd be interested in hearing other suggestions/ideas! - maybe even a
total re-design as a different charttype(?) If you'd like to get a copy
of my code (which contains the data values as text) to use as a starting
place, you can grab it by clicking the text-link below the thumbnail on
my gallery page:

   http://robslink.com/SAS/democd31/aaaindex.htm

Forum: Graph design

Time series rank graphic

jmunoz — Tue, 04 Mar 2008 17:13:27 GMT

Hi All,

I've put together a dashboard ranking my company's top 25 salespeople. In the dashboard I'd like to include a microchart depicting the prior monthly rankings for each salesperson. I'm currently displaying the prior rankings relative to the salesperson's current month ranking. For example, if salesperson A is ranked 5 this month, and was 5 last month, 6 the month before that, and 3 the month before that, I want my graphic to show those changes in a way that the reader can not only see the change in rank, but also know the salesperson's actual prior rankings. Ideally, I'd like to show the last 12+ months rankings for each salesperson.

I've used microbarcharts and sparklines to visualize the rankings, but both methods provide graphs that aren't intuitive enough for explaining changes in rank. I've also tried win/lose charts, but they don't encode the relative rank of the salesperson, which is something I think is important.

I've attached a snippet of my dashboard showing the current rank, and the microbarcharts showing the prior 14 months rankings, the right most bar represents the current month's rankings.

Can anyone come up with something simple and compact (will fit in a standard width Excel cell) for this?

Thanks.

Forum: Visual data analysis techniques

cycle plots in Excel

luke — Wed, 20 Feb 2008 15:01:20 GMT

Does anyone have an Excel template for cycle plots? I was unable to get the one at processtrends.com to take new data. Thanks.

Forum: Graph design

The war on Coke - improved graphics

grasshopper — Mon, 18 Feb 2008 19:36:43 GMT

I found this map to be kind of interesting, but somewhat difficult to read:

Therefore I created my own version, with hopefully some improvements...

First I used more of a color 'gradient' - in their map, it's not intuitive
whether "bright red" or "dark red" is the highest, whereas in my map
I hope it's more intuitive.

Also, in their map the "Other" and "No Data" categories stick out very
prominently (with bright green, purple, etc). imho, those values are not
very important, and are not what the map is trying to show, and they
clutter an already overloaded map ... so I leave those areas white, and
don't show them in the legend.

And I use a white background, so it's not easily confused, or associated
with, the colors used in the map (like a green background might be).

Here's my new version - hopefully simplified, and more intuitive to read.
I'd be interested in hearing other suggestions! ...

Forum: Graph design