Education is concerned with behavioural change,moral,cognitive and skills development. To assess these developments and progresses teachers must test the pupils fairly and effectively with the four major characteristics of all types of tests in mind ,thus comprehensive scope of a test ,its discriminatory ,validity and reliability.
What is a test comprehensive scope? It is that the test should cover the subject area fairly thoroughly.Unfairly treated examination paper consists of very few searching questions on the small part of the syllabus.This type of testing is very unfair to those children who are very acumed on particular part that they will spell disaster when examined on the other part.
Such lack of balance when setting a test makes it comparatively invalid and useless.
The hypothesis is that probably the examiner is tired,lazy or ignorant such that he/she resorts to set one further question worth 20 marks to make it total of 100 marks on the expense of some concepts in the sections of the syllabus .For instance in the grade seven mathematics examination papers ,fractions,money ,distance or time is not test instead concentrating on addition , subtraction,multiplication ,division without including some of the mathematical concepts need to be tested. It should be clear that comprehensive tests are not risked pre-examination question spotting.
Discriminatory test is that which differentiate between best and worst performers according to the scores obtained and spread from top to bottom. However the test items should be relatively and fairly simple to avoid heavy concentration of scores at the bottom or so easy that enables everyone to easily score unnecessary high marks. A test should cover evenly selected questions that are testing the potentials of all learners and be in the ability range and ensures good distribution of marks that satisfies streaming of learners according to their abilities. An interquartile range of 25 or more out of 100 as a normal distribution graph is good.
A discriminatory test is not that one which is a mixture of easy ,too difficulty and beyond capacity questions to the ability level of the pupils.Such test lost its reliability ,so test should be 50% possible to children,if 5% of pupils get it right it is too difficulty to be useful and above all if all scored 95% it is too easy to be useful.A badly worded or trick test is not useful.
A test should be valid. It should accurately measure a particular ability and have a clear purpose.Mixing the objectives of the test oftenly invalidates the test, for example the teacher test two pupils in composition writing.One pupil Susan got 18/20 and John got 10/20 okay.John's work is of poor handwriting ,but good grammar and spelling ,so the teacher deducts to marks at every paragraph . Susan's work is neat well presented in good handwriting and awarded extra two marks for neatness. The result therefore is that John will get 15/20 and Susan will get 18/20. In this case is the test valid? The teacher instead marked handwriting as shown above which invalidates the test. Hence the teacher should set and mark the test according to the purpose and avoid attempting to test more different skills at the same time.
The realibility of a test can be set to the same group,marked by different markers from different backgrounds at the different times and scores are roughly the same. This is a consistent test. It reliability is because of the objectives and independent from personal feelings,attitudes ,hatred,love or favouritism of markers.
Reliable test maintains its consistency in the scores ,even if repeated later ,but it may of no value if it is not valid. Reliability is not helpful if we are not clear of what we are measuring. However test reliability is very low if its is not faithful in its validity.Oral tests and mental aptitude tests some are both doubtful in its validity and reliability.
Tests which are comprehensive,discriminatory ,valid and reliable are useful if adopted by teachers in the teaching and learning process.
No comments:
Post a Comment