On estimates of test reliability.

Abstract

Critical discussions of the reliability problem have been prominent in educational literature for twenty years. While many questions involved have by now been settled, interest has recently been intensified by renewed criticism of the customary techniques. These attacks have been supplemented by numerous proposed substitutes, some of which are, in turn, receiving criticism. It now appears desirable to survey the questions arising regarding the split-half, Spearman-Brown, technique to determine whether that method must be abandoned.

Pouzdanost mjerenja u psihologiji

“…(3) (4) 5Spearmanova i Flanagan-Rulonova koncepcija pouzdanosti potakle su teorijsku raspravu o značenju pravog rezultata i pogreške mjerenja te procjeni i interpretaciji pouzdanosti kao koncepta (Cronbach, 1943(Cronbach, , 1947(Cronbach, , 1951Gulliksen, 1950;Guttman, 1945;Kelley, 1921;Kuder i Richardson, 1937;Loevinger, 1947;Lord, 1955aLord, , 1955bLord, , 1958Thurstone, 1931;Tryon, 1957). Tijekom desetljeća ova je rasprava rezultirala matematičkom formulacijom klasične teorije testova (Lord i Novick, 1968), još poznate pod nazivima klasična teorija mjerenja, odnosno teorija pravog rezultata.…”

Modelo De Sistema De Gestión De Calidad Para La Puesta en Marcha De Cursos No Presenciales: Instrumentos De Seguimiento Y Evaluación

“…Este método analiza la relación entre las diferentes partes de un mismo instrumento (Ledesma et al, 2002), y utiliza el coeficiente alfa de Cronbach para valorar la consistencia del cuestionario completo o consistencia global. (Cronbach, 1973;Llarena, 2008). El grado de fiabilidad exigido, reflejado en el valor de este coeficiente, varía en función del tipo de estudio.…”

Textual Data in Psychiatry

“…Adequate reliability is a prerequisite for validity, because measurement error introduces “noise” into statistical tests that can mask or distort a potential effect within the data (Anastasi, 1934; Cohen, 1990; Cronbach, 1943). Reliability of measurement ensures that the construct is precise, and validity then ensures that the construct is relevant to the nomological net of findings on that topic (Campbell and Fiske, 1959).…”