Premium
TEST LENGTH AND THE STANDARD ERROR OF MEASUREMENT
Author(s) -
GARDNER P. L.
Publication year - 1970
Publication title -
journal of educational measurement
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.917
H-Index - 47
eISSN - 1745-3984
pISSN - 0022-0655
DOI - 10.1111/j.1745-3984.1970.tb00728.x
Subject(s) - standard error , test (biology) , statistics , reliability (semiconductor) , mathematics , simple (philosophy) , observational error , paleontology , power (physics) , physics , philosophy , epistemology , quantum mechanics , biology
Lord (1959) has shown that the standard error of measurement of a test is, for all practical purposes, directly proportional to the square root of the number of items on the test. More specifically, Lord found empirically that the standard error of a test was equal to . if the reliability of the test was computed by the Kuder‐Richardson (KR) 20 formula. If the KR‐21 formula was used, the standard error was equal to .. The present paper sets out to show how these relationships may be derived from the defining formulas of reliability and standard error of measurement, if certain simple assumptions about values of test statistics are made.