The following table show the results from a famous social experiment in the U.S. on recidivism for criminals who were convicted and later released from prison (Rossi, Berk, and Lenihan, AJSVol 88 No 2, September 1982).Recidivism reflects whether or not they were rearrested after being released.The treatment is the study was whether or not the subject received a stipend after they were released from prison (the control group did not get the stipend).A confounding factor was whether the subject found work after being released.The table below shows the relationship between the treatment and finding work.

Did the Subject Work Treatment NO Yes ROW Total Control 163 223 386 Stipend 304 245 549 Column Total 467 468 935