David A. Walker




This paper examines Smith’s (2003) proposed standard error of a proportion index associated with the idea of reliability as sufficiency of information. A detailed table indexing all of the standard error values affiliated with assessments that range from 5 to 100 items, where students scored as low as 50% correct and 50% incorrect to as high as 95% correct and 5% incorrect, calculated in increments of 1 percentage point, is presented, along with distributional qualities. Examples using this measure for classroom teachers and higher education instructors of assessment are provided. Accessed 19,072 times on https://pareonline.net from June 05, 2005 to December 31, 2019. For downloads from January 1, 2020 forward, please click on the PlumX Metrics link to the right.