Test Scoring

Holding an examination is only half of evaluation. Once the papers are written, somebody has to read them, give marks, and keep a record of those marks so they can be used later. Together these jobs are called test score and record keeping. Everything begins with the first job: checking the answer sheets and deciding what each answer is worth.

Picture a class 10 teacher with fifty answer sheets from the same question paper on the desk. Each sheet should tell how much one student has learnt. If two answers of the same quality get 8 marks in one sheet and 5 in another, the marks are no longer measuring learning. They are measuring the mood of the person marking. Scoring therefore has to be accurate and uniform. Uniform means that an answer of a given quality earns the same marks in whichever sheet it appears.

A written examination can contain two kinds of question, objective and subjective, and the two are marked in very different ways.

An objective question has one fixed answer. Ask 'What is the capital of Nepal?' and the correct answer is the same whoever writes it. Such answers are easy to mark, so their marking carries fewer weaknesses than the marking of subjective answers. Fewer is not none, though. When there are many sheets and the marker hurries, mistakes creep into adding up the marks. If a sheet has 32 marks on one page and 14 on another and the total is written as 44 instead of 46, the student loses 2 marks they had fairly earned.

A subjective question asks the student to analyse and explain. Ask 'Why is discipline needed in school?' and one student writes five short points, another writes a long paragraph full of examples, and a third adds a story from their own life. No two answers look alike, because each depends on the student's writing style, choice of words, way of expressing ideas and originality. Marking such answers is not simple, and when the marks are not uniform, some students are treated unfairly.

One safeguard is to appoint a moderator, a second examiner who looks over marked sheets so that carelessness, prejudice or favouritism does not decide a student's marks. Even this does not solve every problem. If subjective answers are to be marked accurately and uniformly, the person marking must also follow certain practices of their own.

Five practices help a marker give subjective answers accurate and uniform marks:

• (a) Prepare the answer outline in advance.
• (b) Use a suitable marking method.
• (c) Decide in advance about unrelated elements.
• (d) Mark question by question.
• (e) Mark without looking at the name.

The answer key is written at the same time as the question paper, not after the sheets arrive. When the marker has already settled what a good answer must contain, answers written in very different ways can still be judged correctly against the same standard.

The key also decides beforehand how many marks go to each quality of an answer:

• how well the content is organised
• knowledge of the content
• presentation
• originality
• correct language, when the subject is a language

Suppose a 10 mark question is split as 4 marks for knowledge of the content, 2 for organisation, 2 for presentation and 2 for originality. Now all fifty sheets are measured with the same ruler, and that brings uniformity to the marking.

For some questions the marker can mark point by point. In a 4 mark question such as 'Write four features of formal education', each correct point can earn 1 mark. The marker can then see exactly which points were given and which were missed.

Some answers, such as essays, cannot be broken into points. These can be sorted instead. The marker reads them all and places them into groups of good answers and poor answers, then gives marks that suit each group. Because every answer is compared with the others before any marks are written, careless marking becomes unlikely and the marking stays impartial.

Answers often carry things that have nothing to do with the question itself: punctuation, poor handwriting, crossings out and clumsy sentences. These can pull a mark down even when the content is good. A history answer may have every fact right, yet if the writing is hard to read and full of crossings out, a marker can end up giving fewer marks without even noticing why.

So the marker decides before opening the first sheet how much, if at all, such things will count. When that decision is made in advance, one rule applies to every student and the marking is impartial.

A student usually writes some answers well and others poorly. After reading an excellent first answer, a marker starts to think 'this one is bright' and feels like giving extra marks to the weaker answers that follow. A poor first answer does the opposite. This pull of one impression over the rest is called the halo effect, and a fair marker must not let it decide the marks.

The remedy is question-wise marking. With fifty sheets, the marker checks question 1 in every sheet, then question 2 in every sheet, then question 3, and so on. Because one question is marked for everyone in a single sitting, every student is measured with the same standard, and one answer has little chance of colouring the marks of another.

The same fairness applies to a student's behaviour outside the paper. A student who is noisy in class or late with homework still earns full marks for a good answer. Only what is written on the sheet is marked.

When the marker does not look at the student's name or symbol number, the marking has the best chance of being impartial.

In an internal examination set by the school, knowing the name does help the teacher. It shows the level each student in the class has reached, which makes it easier to give each one feedback. But a name can also bring in a wrong influence, such as a favourite student or one the teacher is annoyed with. In an external examination the marker does not know the students at all, so this problem does not arise.

Whoever is marking, each answer is read to the end before any marks are given. This makes the marking more reliable.

Once the sheets are marked, the results have to be recorded. Writing down which student achieved how much in which subject is called record keeping.

For a long time the traditional system recorded achievement only as marks and percentages. It has now been replaced by the letter grading system, in which a student's learning achievement is placed in a band and each band is named by a letter.

Under the old system, two students scoring 84 and 85 percent were ranked apart by a single mark, and hard-working students were often left upset over a difference that small. Under letter grading both receive an A, and that kind of distress has come to an end.

Under the present letter grading system, marks are turned into grades and grade points as follows:

Each band in the table is 10 percent wide, and the grade point falls by 0.4 from one band to the next. To find a student's grade:

• Take the student's marks as a percentage, say 67 percent.
• Find the band it falls in. 67 percent lies in the row '60 and above but below 70'.
• Read the grade, its meaning and its grade point from that row. Here the grade is B, meaning Good, with a grade point of 2.8.

Letter grading has replaced the traditional system up to school level, and it has brought several gains for students and parents:

• Students get the chance to choose subjects that match their ability.
• The situation of failing the examination has been removed.
• A difference of a mark or two no longer costs a student a good position, as it did under the marks system.
• Parents can see which group their children fall into.
• No student has to lose heart because of being declared failed.
• Parents can use the grades to advise their children on which subjects to study next.
• An environment is created in which students can study the subjects they are interested in.

Letter grading is not free of weaknesses. Even so, compared with the traditional marks system, it is regarded as the more scientific way of recording achievement.