In 1998 a survey was published on the extent to which software engineering papers validate the claims made in those papers. The survey looked at publications in 1985, 1990 and 1995. This current paper updates that survey with data from 2000 and 2005. The basic conclusion is that the situation is improving. One earlier complaint that access to data repositories was difficult is becoming less prevalent and the percentage of papers including validation is increasing.
Marvin V. Zelkowitz