Several reasons why the “N” of a survey (number of respondents to a question) is important

I paid attention to the Small Law Department Compensation Survey conducted in the summer of 2008 by ACC and Empsight, www.accempsight.com, at page 9 of the Review Sample. The table on that page lists widely varying numbers of companies who gave data for seven benchmark metrics. For example, the number of attorneys is an “N” of 314 while it is an “N” of only 118 for “Other Law Department Professionals” (not attorneys, paralegals, or administrative staff). I assume that means 196 respondents did not have any Other Law Department Professionals, but it is also possible that some respondents did not fill in that question.

My deeper point is that the average number of attorneys is 2.3 whereas the average number of other law department professionals is 2.0, but from very different numbers of respondents. I don’t think you can reliably say anything about the putative ratio of 2.3 to 2.0. Likewise, the number of companies listed as giving data about paralegals (the “N” for paralegals) is 182, midway between the attorney N and the other professionals N.

Again, since 314 companies had at least one lawyer, did 132 departments have no paralegals? If so, the average of the entire group would be lower than the average for the 182 that reported one or more paralegals. The average given is 1.0, so I assume they divided the total number of reported paralegals by 314.

I hope you did not try to savor this post before your first cup of coffee!