Note: Images may be inserted into your messages by uploading a file attachment (see "Manage Attachments"). Even though it doesn't appear when previewed, the image will appear at the end of your message once it is posted.
Register Latest Topics

  Author   Comment  

Posts: 167
Reply with quote  #1 

These sparkline graphs are also a novel invention.  I especially like the p 141 version where the background shows 'limits'.  My preference for these would be to apply shewhart chart methodology (statistical limits instead of spec limits), so that one can readily see if there is system sourced variation beyond normal.  System variation tracking is/can be an early indicator of out of spec production.  To that end, it would also be helpful to subtly highlight each point that is outside of the limits.  It's a little bit difficult to see that points are outside of the gray box when they are very close to the edge.  We then might miss the fact that we need to do something about the system that's being tracked.

Bill Droogendyk

Posts: 802
Reply with quote  #2 

I find that Tufte's sparklines are often an ideal match for dashboards, especially as an alternative to less informative trend arrows. Background fills can be used to expose points where thresholds of various types have been exceeded, including measures of statistical variance from the norm, such as standard deviations or particular percentiles.

Whether values that fall outside these thresholds ought to be highlighted in some way to make them stand out even more depends on how important slight transgressions are to the viewer. In some cases, if the transgression is slight, it doesn't matter if they are missed by the viewer, but in others, even a hair's width of exception might be critical.

Stephen Few
Previous Topic | Next Topic

Quick Navigation: