WorcesterThen: 1890s

 

 

How Not to Do Statistical Analysis

 

In the 1890s, an official of Worcester Public Schools, in the annual report of the school department, presented statistics on the numbers of pupils by grade year, then offered some interpretive narrative.  We’ll bypass his name and the exact year of the report in the interest of leaving the deceased to rest in peace.

 

The data:

                 Grade 1    2,235         Grade 7      879

                 Grade 2    1,802         Grade 8      807

                 Grade 3    1,569         Grade 9      721

                 Grade 4    1,648         High School:

                 Grade 5    1,420           Classical  522

                 Grade 6    1,358           English     579

 

 

His first observation was that there was a steady decline by grade, except for Grade 4.  He speculated on the probable cause of the exception, but it is not important here. (Nor, for that matter, is anything else in his report, aside from a little harmless humor.)   

 

He then got down to work, seeking to draw meaningful conclusions from the pattern observed. Here, in the form of an image from a photograph, is his assessment of the enrollment data.  The curvature reflects the difficulties in getting pages to lie flat when taking pictures on a table at the Worcester Public Library.

 

 

 

Uh oh. That’s a pretty nasty mortality rate for young children.  Could there have been an alternative interpretation, something along the lines of a minor surge in the numbers of young families causing the rise in the numbers of pupils in the earliest years of school?  I wonder what he thought when he saw a corresponding increase in second graders the following year. It’s probably just as well that few parents ever read these reports.