Search This Blog

Showing posts with label statistics. Show all posts
Showing posts with label statistics. Show all posts

Monday, November 22, 2010

Justices' Opinions

I have never seen such misleading graphs as those accompanying the article Justices' Opinions Often Long on Words but Short on Guidelines (New York Times, November 18, 2010: A1, A22).

The graph on number of cases taken seems to shrink to zilch in 1090 because the graph does not have a zero point; the bottom edge is at some unspecified point between 50 and 100 cases. This is unacceptable.

The second graph on the number of words has the opposite problem. It seems as though reports in the 1950s were very brief and that they increased exponentially in the 1970's and have since leveled off. Again this graph has no zero point; the bottom edge is somewhere about 1900 words. If the graph had begun at zero words, the change would not have appeared so dramatic.

Readers really do expect better of the Times' graphic artists. Their watchword should be: always include the zero.


Sent to New York Times [& to the Public Editor]