Test Scores and Reliability |
| Testing
specialists use two different numbers to measure the reliability
of a test. The first is called the reliability coefficient,
which ranges on a scale of 0-1 ("0" is the lowest possible value
and "1" is the highest). Most achievement tests have reliabilities
in the .7-.9 range.
A related concept is the standard error of measurement (SEM) that can be thought of as the amount of "fuzziness" inherent in number that we know is not perfect. "Technically, reliability
shows the extent to which test scores are free from errors of
measurement. No classroom test is perfectly reliable because
random errors operate to cause scores to vary or be inconsistent
from time to time and situation to situation. The goal is to
try to minimize these inevitable errors of measurement and thus
increase reliability." For more details, see a very good
description in: TEST
RELIABILITY by Lucy C. Jacobs, Ph.D. Also, check
out her HOW
TO WRITE BETTER TESTS: A Handbook for Improving Test Construction
Skills, which is very helpful - and you can print it out,
too. |
|
||||||
