Comments on: Measure Data Quality http://xldennis.wordpress.com/2008/01/28/measure-data-quality/ XL-Dennis on VSTO & .NET & Excel Tue, 10 Nov 2009 17:33:20 +0000 http://wordpress.com/ hourly 1 By: Jon Peltier http://xldennis.wordpress.com/2008/01/28/measure-data-quality/#comment-10632 Jon Peltier Tue, 29 Jan 2008 11:50:40 +0000 http://xldennis.wordpress.com/?p=254#comment-10632 A table is also good. I started with a table to make my charts, three columns: {EM, ST, SV}, populated with Valid or Blank. This made it easier to get all 8 (i.e., 2^3) sets of conditions in the output, where you had only seven in the pie. A table is also good. I started with a table to make my charts, three columns: {EM, ST, SV}, populated with Valid or Blank. This made it easier to get all 8 (i.e., 2^3) sets of conditions in the output, where you had only seven in the pie.

]]> By: Dennis Wallentin http://xldennis.wordpress.com/2008/01/28/measure-data-quality/#comment-10631 Dennis Wallentin Tue, 29 Jan 2008 09:48:04 +0000 http://xldennis.wordpress.com/?p=254#comment-10631 Hi Jon and Dermot, Thanks for Your comments which are highly appreciated. I'm in favour of tables. Mainly due to my shortcoming to create relevant charts models as well as due to my semi-color blindness. I will update the chart accordingly. Kind regards, Dennis Hi Jon and Dermot,

Thanks for Your comments which are highly appreciated.

I’m in favour of tables.
Mainly due to my shortcoming to create relevant charts models as well as due to my semi-color blindness.

I will update the chart accordingly.

Kind regards,
Dennis

]]> By: dermot http://xldennis.wordpress.com/2008/01/28/measure-data-quality/#comment-10630 dermot Mon, 28 Jan 2008 21:17:21 +0000 http://xldennis.wordpress.com/?p=254#comment-10630 Dennis, I find it hard to read the pie slices without text labels (the eye has to keep jumping between the pie and legend), also it's not clear which slices are "good" and which are "bad". There's no way to take the information in at a glance, as busy executives need to do. In this case, I think a Venn diagram would do the best job because it would clearly show the different combinations. Alternatively, I would suggest a (sideways) bar chart with the bars arranged in order from best(nothing missing) to worst (all missing). Dennis, I find it hard to read the pie slices without text labels (the eye has to keep jumping between the pie and legend), also it’s not clear which slices are “good” and which are “bad”. There’s no way to take the information in at a glance, as busy executives need to do.

In this case, I think a Venn diagram would do the best job because it would clearly show the different combinations.

Alternatively, I would suggest a (sideways) bar chart with the bars arranged in order from best(nothing missing) to worst (all missing).

]]> By: Jon Peltier http://xldennis.wordpress.com/2008/01/28/measure-data-quality/#comment-10629 Jon Peltier Mon, 28 Jan 2008 20:29:22 +0000 http://xldennis.wordpress.com/?p=254#comment-10629 Let's not perpetuate the presentation paradigm that favors 3D exploded pie charts. Either a bar chart (http://www.freefilehosting.net/show/3b5j6) or a dot plot (http://www.freefilehosting.net/show/3b5j4) would have been a better choice. Let’s not perpetuate the presentation paradigm that favors 3D exploded pie charts. Either a bar chart (http://www.freefilehosting.net/show/3b5j6) or a dot plot (http://www.freefilehosting.net/show/3b5j4) would have been a better choice.

]]>