DOUBLE IMPRESSION MARKING OF ENGLISH LANGUAGE ESSAY AND SUMMARY QUESTIONS
Robert E. Wood, Barry F. A. Quinn · Educational Review · 1976
Impression marking of English Language essay and summary questions by pairs of examiners is shown, as expected, to be more reliable than single marking. Since the reduction in spread of marks is accompanied by a greater reduction in the errors, overall discrimination is improved. Given the limited statistical information available, it is concluded that pairing of examiners can as well be done by random or quasi‐random means as by attempts at calculated matching. Impression marking will give different results to analytic marking, with its emphasis on technical points. Whether this is desirable or not is a matter for debate. The indications are, however, that the practical consequences of any differences will not be great. By switching from analytic to impression marking #opeven without introducing multiple marking at the same time#cp a candidate's result would probable be no more affected than if he or she were to be marked by one examiner rather than another.