Home > Statistics Every Writer Should Know > The Stats Board > Discusssion
basal and ceiling scores
Message posted by on September 18, 1999 at 12:00 AM (ET)
I am evaluating a language test for a class. The test is not standardized. Is it still possible to have basal and ceiling scores? If so, how do I determine what they are? Thanks alot.
READERS RESPOND:
(In chronological order. Most recent at the bottom.)
Re: basal and ceiling scores
Message posted by Bill on September 21, 1999 at 12:00 AM (ET)
The ceiling effect (or basal) means that the range of difficulty of the test items is limited. Thus the test does not measure the entire range possible on the dimension being measured, e.g., a language test. If a student correctly answers 95 out of 100 language test items then he can only improve his language test score by 5 items or about 5%. If another student answers 50 items correctly on the same test then he can improve his performance by 50 items or 100%. Thus the ceiling effect places an artificial restriction on the distribution of gain scores across levels of initial ability. (Borg and Gall, 1983) Yes, it is possible to have a basement or ceiling effect on a non-standardized test. I have never seen, however, an objective measure of this effect. I guess it comes down to knowing what you are trying to measure, the limitations of your measurement instrument (including the ceiling effect), and a common sense interpretation of the data.
Borg and Gall. Educational Research. Longman Press. New York, 1983.
Your $5 contribution helps cover part the $500 annual cost of keeping this site online.
