NACE salary surveys' accuracy

[latent aaaack]

Are these things accurate?

About Salary Survey: Salary Survey is a quarterly report of starting salary offers to new college graduates in 70 disciplines at the bachelor's degree level. The survey compiles data from college and university career services offices nationwide. Salary Survey is issued in Winter, Spring, Summer, and Fall, with the Fall issue serving as the year-end report. (Salaries reported in this press release reflect offers to bachelor’s degree candidates.)

http://www.naceweb.org/press/display.asp?year=&prid=266

Wouldn't these surveys which are often quoted in the press in fact be completely un-scientific because they really on self-reported offers of those graduates that choose to answer the universities' survey (and who would tend to have done better) or do the universities' conduct an actually valid sampling of graduates?

In particular I wonder about the computer science starting salary figure since it's as high as that of an engineering degree which are much more difficult to get and are commonly thought to be among the highest paid professions.

 

[Whack01]

Demand more than difficulty and rarity determine the value of the degree. Computer science is in demand since there is a very large return on the investment. Also there is a huge gap in skill between tech savy and non technical people so trivial jobs become inordinately difficult.

For example, I spent 5 hours after I installed Linux trying to get my dvd playback to work compiling various .tar files & googling into every orifice of the internet. Yet in the end it turned out I needed to type one line at the bash prompt: yum install mplayer. The point being not that I'm retarded, but that if I had already known what to do I'd have done it in 2 seconds.

/ using Fedora 7 x86_64
// would have had it working in 10 minutes if the 64bit versions of the software had detected liboil properly when I used ./configure, or if I had just started by RTFM.
/// I have a BS in Computer Science