An ever more important aspect of quality assurance in testing is the understanding of measurement uncertainties and the calculation of uncertainty budgets across the entire test system. This involves the assessment of uncertainties (note these are different to measurement errors) for each part of the measurement system and combining these to create an overall system uncertainty statement, eg "Test levels are within 2.5% with a confidence level of 95%"

Free Download - Uncertainty calculation spreadsheet

Some simple definitions:
Accuracy - qualitative term only that requires interval and uncertainty to be known to be meaningful.
Confidence level - the statement of how sure we are of the stated result, eg "with 95% confidence"
Confidence interval (or just interval) - margin within which the "true value" being measured can be said to lie, with a given level of confidence.
Error - is the difference between the measured value and the "true value".
Error correction - when possible we should always correct for known errors in the system, ie if you have an accelerometer calibration sheet that shows the error at a certain frequency then this can be corrected.
Tolerance - The allowable range of values that can be tolerated - requires error and uncertainty to be known.
Uncertainty - is the quantification of the doubt about the measurement result.

There are a number of excellent references available:

The National Physical Laboratory also has a beginners guide to measurement uncertainty at http://www.npl.co.uk/npl/publications/good_practice/uncert/

The UKAS website has some interesting pointers at http://www.ukas.com/information_centre/technical/technical_uncertain.asp

Those people who enjoy a challenge may like to visit the NIST website and view their thoughts http://www.physics.nist.gov/cuu/Uncertainty/

Last one, a detailed and lengthy pdf file with good info if you have time to spare http://www.a2la2.net/guidance/est_mu_testing.pdf