Premium
Examination Grades: how many should there be?
Author(s) -
Cresswell M. J.
Publication year - 1986
Publication title -
british educational research journal
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.171
H-Index - 89
eISSN - 1469-3518
pISSN - 0141-1926
DOI - 10.1080/0141192860120104
Subject(s) - grading (engineering) , grading scale , mathematics education , psychology , relation (database) , scale (ratio) , computer science , data mining , medicine , civil engineering , physics , surgery , quantum mechanics , engineering
There is no generally accepted rationale for deciding the number of grades which should be used to report examination results. Two schools of thought on this matter have been identified in the literature. One view is that the number of grades should reflect the reliability of the underlying mark scale. The other view focusses upon the loss of information incurred when the mark scale is reduced to a number of fairly coarse categories. The first of these views usually implies the adoption of a relatively small number of grades; the second view implies the use of a considerably larger number of grades. In this paper, the various factors which determine the relative merits of these two schools of thought are considered in relation to the different functions which examinations fulfil. The intention is to enable better informed choices to be made concerning the number of grades to be used in the grading scheme of any particidar examination.