Estimation of written answers.

187

profoundly ignorant of the data on which all such generalization ought to rest.

written

I will suppose that you have framed your eight or ten The estiquestions in view of the actual knowledge, both of the mation of ordinary scholar, and of the best who want an oppor- answers. tunity of distinguishing themselves; it then becomes necessary to estimate the answers. On the whole, the ordinary arithmetical test is the fairest and the least liable to error. You determine on a maximum, say 100, to represent the highest attainable excellence. You then assign a due proportion of marks to each question according to its difficulty. It is a good plan to distribute about go in this way, reserving the last ten for style, neatness and finish, and general skill of arrangement. In distributing your 90 marks among, say ten questions, you will give perhaps 12 to one, and 6 to another, according to the amount of knowledge and intelligence required to produce a perfect answer. But I would not tell the scholars which questions carried most marks. It is not good that they should be speculating and enquiring what are the relative values of different answers in your mind. It is enough to tell them to select those questions which they can answer best; and you will judge, if one fastens on the purely memory work, while another chooses to give the best of his time to those questions which require some thought and originality to answer them, how such answers ought to be estimated.

As you read each answer in turn, you should set down How to read an the proportion of the maximum number assigned to that examiparticular question which the answer deserves. It is nation essential that this should be done with each question, paper, and that there should be no room left for caprice or hasty impression by attempting to mark the value of the paper as a whole. Nevertheless, before passing on to

another paper, and while your recollections are perfectly fresh, it is well to add up the result and see if the total appears to represent fairly the general merit of the paper considered as a whole. For it may be that the scholar though evidently writing from a full mind, has mismanaged his time, has given needlessly elaborate answers, say to four questions, for which he has the maximum marks, and yet has a smaller total than an inferior scholar who has attempted eight questions, and has scored a fair number for each. This should be set right at once by the addition of a few marks for general ability. It is not safe, or really equitable, to leave the total of each scholar's marks to be added up afterwards.

In mathematics it is not difficult for a student, by doing all the exercises right, both in method and result, to obtain the full number of marks. But in other subjects the maximum will rarely, or ever be attained, as it will represent in the examiner's mind the highest conceivable standard of excellence, and it is very unlikely that this will be attained in every one of a number of questions in History or Literature. So in most subjects I should regard as a good paper that which obtained three-quarters, and as a fair or passable paper that which received half of the marks.

Great care should be taken to keep your own judgment equitably balanced while you are reading. So before marking any, it is well to read over several papers, choosing, if you have any sort of clue, one or two likely to be good, and one or two likely to be indifferent, and so fix the standard of what it is reasonable to expect. With this standard in your mind it will be fair to begin marking the answers one by one. If you are examining for any prize or competition it is needful to give the papers a second reading, comparing not only paper with paper,

Negative marks.

189

but answer with answer. For ordinary pass examinations

this is not necessary.

marks.

It is sometimes asked whether negative marks should Negative ever be given, or marks deducted for ignorance. That depends on the kind of ignorance. Mere absence of knowledge ought not to be counted as a fault, otherwise than as depriving the pupil of the marks which would have been due to knowledge. It ought not, I think, to be punished by the subtraction of marks to which other knowledge would entitle him. But the sort of pretentious ignorance which makes blunders and mistakes them. for knowledge, which indulges in grand, sonorous and vague statements carefully constructed to conceal the lack of true information ought to be punished as a fault. A bad and inflated style, false spelling, the use of words which are not understood, may not unreasonably be visited with the forfeiture of marks to which the mere memory work would be entitled. But you must make allowance for a few very innocent blunders, such as will be inevitable among young people who are being put to this sort of test without much previous practice. When a scholar tells you that "we derive a good deal of our early knowledge of English History from an ancient chronicler named Adam Bede," that "Buckingham was at first a friend of Dryden, but that he afterwards became one of his contemporaries," or that "Sir Wm. Temple was a statesman in the time of Charles II. who had a hand in the Triple Alliance, and who in later life acquired some odium by writing Essays and Reviews," you may set it Venial and down as mere bewilderment, which does not mean ig-punishable norance, which would be corrected by a moment's thought, and should therefore not be counted as a fault. On the other hand, a blunder such as that of the man who, in commenting on the passage in Milton referring to "our

blunders.

The

66

sage and serious poet Spenser" as a better moralist than Scotus or Aquinas," said that these worthies were "two licentious poets of the period;" or that of the student who said that "John Locke was a poet who was knighted by queen Elizabeth," or that of him who wrote that "the Americans were so grateful for the services of George Washington that they made him a peer," ought to be reckoned as a fault to be punished, because in each case it is a mere guess, put out rather dishonestly with the chance of its being right or with the deliberate intention of practising on the possible ignorance or carelessness of the examiner.

Even in class work, the course of oral questioning may sometimes be advantageously interrupted, by requiring the answer to be given by all the students immediately in writing instead of word of mouth. If you want to know whether all the class knows a French verb, or a number of dates, or a group of names, this is an expeditious and very thorough method. And here, when you have examined the note books by the plan of mutual correction or otherwise, the result may well be tabulated in a numerical form. But in ordinary oral questioning of a class and estimating its result, I do not think it is quite possible to adopt the arithmetical mode of measurement with perfect exactness, and therefore I would not use it at all, but employ other symbols such as Excellent, Good, Fair, Moderate, which are better fitted to describe general impressions.

And yet now the most important thing remains to be morality said. This whole problem of examinations and the right of examinations. way of conducting them and preparing for them touches very nearly the morality of the school life. Look well to the influence which the examinations you use are having

The morality of Examinations.

191

on the ideal of work and duty which your scholar is forming. Ask yourself often if that which will enable. him to do best in examination is also that which is best for him to learn. Watch how the prospect of the examination tells upon his methods of study, his sense of honour, his love of truth. Determine that whatever happens, you will not pay too heavy a price for success in examinations. Discountenance resolutely all tricks, all special study of past papers, and of the idiosyncracies of examiners, and all speculations as to what it will and what it will not "pay" to learn. It is because sufficient regard is not paid to these considerations, that many thoughtful persons now are fain to denounce examinations altogether, as the bane of all true learning, and as utterly antagonistic to the highest aims of a teacher. There ought however to be no such antagonism. In their proper place, examinations have done great service to education, and are capable of doing yet more. they can only do this on one condition. sure that for us, and for our pupils, success in examinations shall not be regarded as an end, but as a means towards the higher end of real culture, self-knowledge and thoughtfulness. And let us keep in mind for them and for ourselves the old sound maxim: "Take care of everything but the examination, and let the examination take care of itself."

But

Let us make