081005-1347 EST
Some other references stimulated by this thread:
This reference above
http://www.unitedstatesaction.com/emp_and_faraday_cages.htm
led me to
http://www.arrl.org/tis/info/pdf/88615.pdf
In general you shield electric fields from something by enclosing the something in a conductive box. At low frequencies this does not have much effect on magnetic fields. Electric fields are capacitively coupled.
You shield from magnetic fields by a magnetic material box.
Twisted wire cables are not by themselves shields but provide a means to cancel the induced voltage from a distributed magnetic field. Telephone cables and CT-5 cables use twisted pairs for this purpose.
Radiated energy is shielded by electrostatic and/or magnetic shields.
Another type of noise problem is from ground (common) paths that have an unwanted current and this current produces a voltage difference along the common path. Shielding has no effect on this type of noise.
See my web site for a discussion on noise and grounding at
http://www.beta-a2.com/noise_grnd.html
Following is an old reference with a small section on shielding:
"High-Frequency Measurements", August Hund, McGraw-Hill, 1933, p 43. I never had this book for a class, but it is interesting to look back at knowledge that had existed by this moderately early date. The superheterodyne was only a few years old at this time, and commercial radio broadcasting had only begun at the beginning of the 1920s. It is also interesting to note the very great advances from 1930 to 1940. Note that radar existed on 7 December 1941 and worked, but failure of communication because of many factors resulted in a major disaster.
From the investigation after 7 December:
http://www.ibiblio.org/pha/pha/congress/part_3.html#138
Scroll down a bit to the radar discussion.
The problem of noise in an information channel includes many factors over and above the electrical circuits.
.