Two images.  The first is from the Speaker of the House (via).  The second is from the Federal Reserve Bank of Atlanta (link).  A great demonstration of how seemingly minor changes (the data series you choose; the visualization choices you make) come to bear on the conclusions you draw.

The first plots absolute job losses (in 1000s of jobs), for the current recession, the 1991 recession and the 2001 recession.

The second plots the unemployment rate, comparing the current recession to an average of 1991, 2001, 1973-75 and 1982-83.

Which years do you pick?  How do you measure unemployment?  Do you compare specific recessions or aggregate them?  Seemingly minor design choices with a serious bearing on your conclusions about the comparative severity of this recession.