Two alternative approaches to educational testing that must be thoroughly understood are norm-referenced testing and criterion-referenced testing. Although there are similarities between these two approaches to testing, there are also fundamental differences between norm and criterion referenced testing.There have been disputations about the relative virtues of norm and criterion-referenced measurements for a long time. However, a fundamental fact is recognised by most of concerned people that norm-referenced and criterion-referenced testing are complementary approaches. Check for Educational Evaluations in US at UT Evaluators
(i) Criterion-Referenced Evaluation:
When the evaluation is concerned with the performance of the individual in terms of what he can do or the behaviour he can demonstrate, is termed as criterion- referenced evaluation. In this evaluation there is a reference to a criterion.But there is no reference to the performance of other individuals in the group. In it we refer an individual’s performance to a predetermined criterion which is well defined.
(i) Raman got 93 marks in a test of Mathematics.
(ii) A typist types 60 words per minute.
(iii) Amit’s score in a reading test is 70.
A simple working definition of Criterion-Referenced Evaluation :
A criterion-referenced test is used to ascertain an individual’s status with respect to a defined achievement domain.In the above examples there is no reference to the performance of other members of the group. Thus criterion-referenced evaluation determines an individual’s status with reference to well defined criterion behaviour.
It is an attempt to interpret test results in terms of clearly defined learning outcomes which serve as referents (criteria). Success of criterion-reference test lies in the delineation of all defined levels of achievement which are usually specified in terms of behaviourally stated instructional objectives.
The purpose of criterion-referenced evaluation/test is to assess the objectives. It is the objective based test. The objectives are assessed, in terms of behavioural changes among the students.Such type of test assesses the ability of the learner in relation to the criterion behaviour. Glasar (1963) first used this term, ‘Criterion-reference test’ to describe the learner’s achievement on a performance continuum.Hively and Millman (1974) suggested a new term, ‘domain-referenced test’ and to them the word ‘domain’ has a wider connotation. A criterion referenced test can measure one or more assessment domain.
(ii) Norm Referenced Evaluation:
Norm-referenced evaluation is the traditional class-based assignment of numerals to the attribute being measured. It means that the measurement act relates to some norm, group or a typical performance.It is an attempt to interpret the test results in terms of the performance of a certain group. This group is a norm group because it serves as a referent of norm for making judgements.
Test scores are neither interpreted in terms of an individual (self-referenced) nor in terms of a standard of performance or a pre-determined acceptable level of achievement called the criterion behaviour (criterion-referenced). The measurement is made in terms of a class or any other norm group.Almost all our classroom tests, public examinations and standardised tests are norm-referenced as they are interpreted in terms of a particular class and judgements are formed with reference to the class. Educational Evaluations in US visit here
(i) Raman stood first in Mathematics test in his class.
(ii) The typist who types 60 words per minute stands above 90 percent of the typists who appeared the interview.
(iii) Amit surpasses 65% of students of his class in reading test.
A simple working definition Norm Referenced Evaluation :
A norm-referenced test is used to ascertain an individual’s status with respect to the performance of other individuals on that test.In the above examples, the person’s performance is compared to others of their group and the relative standing position of the person in his/her group is mentioned. We compare an individual’s performance with similar information about the performance of others.
That is why selection decisions always depend on norm- referenced judgements. A major requirement of norm-referenced judgements is that individuals being measured and individuals forming the group or norm, are alike. In norm-referenced tests very easy and very difficult items are discarded and items of medium difficulty are preferred because our aim is to study relative achievement.