standard error n
also SE
(in statistics and testing) a statistic used for determining the degree to
which the estimate of a population parameter is likely to differ from
the computed sample statistic. The standard error of a statistic provides an
indication of how accurate an estimate it is of the population parameter.
One commonly used standard error is the standard error of the mean,
which indicates how close the mean of the observed sample is to the mean
of the entire population.
standard error of measurement n
also SEM
an estimate of the range of scores wherein a test taker’s true score lies.
The standard error of measurement decreases as the reliability of a test
increases which is shown by the following formula:
SEM=SD
where SD=the standard deviation of test scores
r=the reliability estimate of a test
For example, a test taker obtained a score of 85 on an ESL reading test that
has a standard deviation of 12 and a reliability coefficient of .91. The test’s
SEM is estimated as follows:
SEM= 12 = 12 =(12)(0.3) =3.6
As a test taker’s true scores are expected to distribute normally if this per-
son took the same test an infinite number of times, this person’s true score
would be expected to lie within +or -one SEM of this person’s observed
score 68% of the time (i.e. between 88.6 and 81.4) and within +or -two
SEM of this person’s observed score 95% of the time (i.e. between 92.2 and
77.8) (see the normal curve).
standard error of the mean n
see standard error
standardization^1 n standardize v
the process of making some aspect of language usage^1 conform to a
standard variety. This may take place in connection with the writing
system or the spelling system of a particular language and is usually imple-
mented by a government authority. For example, a standardized system has
been introduced in Malaysia and Indonesia, which provides a common
standard for the spelling of Malay and Indonesian, which are both varieties
of the same language.
0.091 - rstandard error1 - 0.19