402 SECTIONAL TRANSACTIONS.— L. 



offered to him by employment overseas in the development of the vast field of 

 undeveloped resources that is needed to feed the fast-developing manufacturing 

 resources of the home country. 



Demonstration of The Work of the Leeds Schools Music and Drama 

 League. Address by Sir Henry Hadow, C.B.E. 



Tuesday, September 6. 



Discussion on School Examinations. Dr. P. B. Ballard, Dr. J. M. 

 Crofts, Mr. B. C. Wallis, Mr. J. H. Arnold. 



Dr. P. B. Ballard. — The Art of Examining Children. 



Examining is measuring, and the very essence of measurement is accuracy. Hence 

 modern efforts at the reform of examinations aim at making the testing objective. 

 An attempt is made to eliminate so far as may be the element of chance. Luck 

 enters largely into the current examination system which tests knowledge by very 

 limited samples, and uses formal English composition almost exclusively as the 

 mode of response. As a result the testing has a low degree of ' reliability,' in the 

 technical sense of that word. 



An examination may be diagnostic, or evaluative, or competitive. The purpose 

 of the examination determines its content and its difficulty. It is rare that an examina- 

 tion can adequately serve more than one of the purposes referred to above. What- 

 ever its purpose, however, it should satisfy certain criteria. It should, so far as possible, 

 distinguish between promise and performance ; it should be independent of the 

 personality and mood of examiner ; it should be reliable in the sense that a similar 

 examination given the same candidates on, say, the following day would yield virtually 

 the same results; and it should cover a reasonably wide field of knowledge and capacity. 

 The modern tendency is towards a large number of brief questions easy to mark, 

 instead of a small number of questions difficult, if not impossible, to mark. 



Dr. J. M. Crofts. — The First and Second School Certificate Examinations 

 and Standardisation of Results. 



An account is given of the procedure followed by most examining bodies in an 

 endeavour to standardise their results ; the procedure affects appointment of ex- 

 aminers, setting papers, and, above all, marking scripts. It is shown that it is com- 

 paratively easy for examiners to place a batch of candidates in order of merit, but 

 that there is no certainty of opinion as to where the pass line should be drawn. The 

 chief examiners in a subject review the standard of marking adopted by each assistant 

 examiner, and apply such corrections to his scale of marking as they deem necessary. 

 There is no guarantee, however, that the chief examiners have adopted the right 

 standard ; they are always examiners of much experience, and doubtless they are 

 fairly successful hi obtaining a satisfactory standard, but they have no absolute unit 

 of measurement to go by, they can give only their personal opinion, and testing of 

 this personal opinion shows a very considerable variation among experienced 

 examiners. It is suggested that where the number of candidates is large, it is fair to 

 assume that approximately the same percentage should pass from year to year. 

 Such an assumption cannot be proved, but it is supported by the constant difference 

 in performance between boys and girls over a series of years as regards standard 

 of work sent in, and in the dispersion of the frequencies of distribution. 



For the Higher School Certificate Examination, on which many scholarships are 

 awarded, other considerations enter, as the value of marks in literary subjects has 

 to be balanced against those in scientific or other subjects, and different types of 

 distribution curves are obtained in different groups of subjects. It is suggested that 

 the marks as sent in by the examiners should be modified so that they all conform 

 to the same type of curve, and so equate the chances for scholarships of candidates 

 in different groups. 



