CRITERION-REFERENCED TEST (CRT) AND NORM-REFERENCED TEST (NRT)

1. SIGNIFICANCE
Revolutionary changes have taken place in the field of teaching id learning processes in the last four decades. Individual instruction emphasized. So designing of the tests should be such as to reflect e changes in what the individual has learned. Again as an individual ogresses through a programme of instruction educators want to know various junctures, precisely what the student knows and whether he ows enough to proceed further. CRT and NRT are used to measure 5 attainment of the learner. The fundamental distinction between IT and CRT is based on the manner in which one interprets the results tile learner's test performance. In the case of NRT one interprets runners test performance according to the performance of 'others', case of CRT; one interprets learner's performance in relation to well 5ned class of knowledge, skills, attitudes and the like. The following sensation provides details of CRT and NRT.

2. CRITERION REFERENCED TEST

Glaser (1963) first used the term CRT to highlight the need for tests can describe the position of a learner on a performance continuum, er than the learner's rank within a group of learners. The word 'criterion' in the CRT refers to a domain of behaviors in criterion referencing one is interested in referencing of an Tines test performance to a well-defined domain of behavior ensuring an objective or skill.

I. Popham, W.J. and Husek, T.R. in their paper, "Implications of Criterion Referenced Measurement" published in Journal of Educational Measurement Vol VI (1969), define criterion-referenced measures as. 'Those which are used ascertain an individual status with respect to some criterion i.e. performance standard. It is because the individual is compared with some established criterion, rather than other individuals, that these measures are described as criterionreferenced. The meaning of an individual score is not dependent on comparison with other testees, We want to know what the individual can do, not how much he stands in comparison to others." 2. Ivon (1970) defines a Criterion-Referenced Test as "One consisting of items keyed to a set of behavior objectives. 3. Harris and Stewart (1971) define CRT as "One consisting of a sample of production tasks drawn from well-defined population of performance." 4. Glaser and Nitko, in their paper "Measurement in Learning and Instruction" in the book Educational Measurement (1971) (dited by R.L. Thorndik write, "A criterion-referenced test is one that is deliberately constructed to yield measurements that are directly interpretable in terms of special performance standards." 5. In the words of Gronlund , N.E. (1985), criterion-referenced test is "a. test designed to provide a measure of performance that is interpretable in terms of a clearly defined and delimited domain of learning tasks. 6. Sex Gilbert (1989) writes, "Criterion-referenced test relate a student's score on an achievement test to domain knowledge rather than to another student's score." 7. Hopkins, Stanley and Hopkins (1990) observe, "These tests have the purpose of identifying individuals who have or have not acquired certain basal

competencies. Different consequences for instruction such as remediation, may result." 8. Wiersma, William and Jurs, Stephen G. (1990) write "Criterion referenced tests describe the performance of the student in terms of actual skills or tasks that are included in the test. From above definitions it is concluded that Criterion Referenced Test is meant to measure the achievement of an examinee on a certain domain to find out his level of achievement in that domain. It has nothing to do with the achievement level of other examinees Characteristics of Criterion-Referenced Test 1. Its main objective is to measure student's achievement of curriculum based skills. 2. It is prepared for a particular grade or course level. 3. It has balanced representation of goals and objectives. 4. It is used to evaluate the curriculum plan instruction progress and, group student's interaction. 5. It can be administered before and after instruction. 6. It is generally reported in the form of I. II. III. IV. Minimum scores for partial and total mastery of main skill areas. Number of correct items. Percent of correct items. Derived score based on correct items and other factors.

Use of Criterion-Reference Testing

i. To discover the inadequacies in learner's learning and assist the weaker section of learners to reach the level of other students through a regular programme of remedial instruction.

ii. iii. iv.

To identify the master learners and non-master learners in a class. To find out the level of attainment of various objectives of instruction To find out the level at which a particular concept has been learnt. To better placement of concepts at different grade levels,

v.

To make instructional decisions of what to do with a learner in individually prescribed instruction programme.

Limitations of Criterion-referenced Testing Chase (1974) lists the following limitations: 1 Criterion-referenced test tells only whether a learner has reached proficiency in a task area but does not show how good or poor is the learner's level of ability. 2. Tasks included in the criterion-referenced test may be highly influenced by a given teacher's interests or biases, leading to general validity problem. 3. Criterion-referenced tests are important for only a small fraction of important educational achievements. On the contrary, promotion and assessment of various skills is a very important function of the school and it requires normreferenced testing.

Reliability of Criterion-Referenced Tests Reliability can be increased by the following: 1. Paying proper attention to the construction of test items. 2. Paying proper attention to the test-setting. 3. Paying proper attention to the examinees. 4. Making the items similar. 5. Using appropriate wording.

6. Increasing the number of items. 7. Increasing the number of discriminating items. 8. Giving clear directions to the examinees. 9. Giving clear directions to the evaluations. 10. Motivating students; to perform better.

Interpretation of Criterion-Referenced Test 1. Interpretation should be based on what the items actually measure. 2. There should be sufficient items for each type of interpretation In case these are less than 10 items, they should be combined with other items through lengthening content items. 3. The test should contain easy as well difficult items otherwise^ Would be difficult to describe what low achievers could do.

Guidelines for Constructing, Administering and Evaluating Criterion-Referenced Tests Ronald Hambelton and Daniel Bignor (1978) have suggested the Following guidelines.

Objectives 1 The purposes of the test should be stated clearly and concisely 2. Objectives should be written in such a way that each item could be identified with one objective 3. There should be appropriate rationale for the inclusion of each Objective in the test. 4. The classroom teacher should be in a position to adapt the test to meet local situation.

5. The set of objectives of the test should be a true representative of content domain.

Test items 1. The test items should be free from any bias. 2. The set of test items should measure an objective for which it has been included in the test. 3. The test items should be valid indicators of the objectives for which they have been prepared. 4. The test items should be in appropriate form to measure the objectives for which they have been selected. 5. A heterogeneous sample should be used in the test items. 6. Item analysis should be used to detect flawed items. 7. The item review process should be described.

Administration 1. The test manual should specify the role and responsibilities of the examiner. 2. The test administrators should have adequate information relating to the purpose, time limits, answer sheets and scoring of test. 3. The directions of the test should be clear. 4. The test should be easy to score.

Test Layout 1. Test booklet should be attractively printed. 2. The layout of test booklet should be convenient for examiners.

Reliability (i) The test length should be sufficient enough to find out test score reliability. (ii) The sample of examinees used in finding out reliability should be adequate and representative. (iii) The reliability information should be provided in the test for each intended use of the test score. (iv) The reliability information provided in the test should be appropriate for the use of the score of the test.

Cut Off Scores 1. There should be a rationale for the selection of the method for determining cut-off scores. 2. There should be evidence for the validity of the chosen cut-off marks. Validity

1. The validity evidence should be adequate for the intended use of the test score. 2. The test manual should provide an appropriate discussion on the factor affecting the Validity of scores.

Criterion -rejerenced lest (CRT) & Norm-Referenced Test (NRT)

Finding reliability and validity Using test in the classroom

Selection of unit or topic

Final test form A and B

Delineation of domain and its description

Internal review of step

Specification of domain objective

Field try out of test form A and B

External review of steps

External review of step

Internal review of step

Framing test items for A and B

Internal review of step

Fig. 18.1. Flow chart showing various steps in development of criterion referenced tests (Source : Singh. Prilam. Cruerion-Rcferenced Testing: A Monograph NCERT. New Delhi, 1983

NORM-REFERENCED TEST (NRT) Gronlund (1976) states that non-referenced tests are "designed to. Students in order of achievement, from high to low, so that decisions j on relative achievements (e.g., selection, grouping, grading) can hade with greater confidence." Bormuth (1970) writes that a norm-referenced test is designed "to sure the growth in a student's attainment and to compare his level attainment with the-levels reached by other students and norm group." M.R. Gronlund defines Norm-Referenced Test as "a test designed provide. a measure of performance that is interpretable in terms of individuals, relative standing in some known group." bormulh (1&70) writes that aNRTis designed "to measure the growth student's attainment and to compare his level of attainment with levels reached by other students and norm group." This test is used primarily for comparing achievement of an examinee that of a large representative group of examinees at the same grade . The representative group is known as the 'Norm Group'. Norm up may be made up of examinees at the local level, district level, level or national level. Since the development of norm-referenced is expensive and time consuming. They are produced by commercial publishers. The following derived scores are generally reported these published tests.

(a) Percentile Rank: A number telling the percentage of examinees a defined group scoring lower than the particular raw score. (b) Linear Standard Score: A number telling the location of the articular row score in relation to the mean and standard deviation of defined group. (c) Normalized Standard Score: A number telling the location of particular raw score in relation to a normal distribution defined in of a particular group. (d) Grade Equivalent Score: A number telling the grade placement which the particular raw score is the average for a defined group.

Chief Characteristics of a Norm-Referenced Test 1, its basic purpose is to measure student's achievement in curriculum sed skills 2. It classifies achievement as above average, average or below average for a given grade. 3. It is generally reported in the form of Percentile Rank, Linear Standard Score, Normalized Standard Score and Grade Equivalent Score.

Process for Developing Norm-Referenced Achievement Test According to L.M. Carey (1988), the process for developing Norm-Referenced Test may be divided into the following stages:1. Design Stage. It is done through i. ii. iii. iv. v. Curriculum analysis Selecting objectives to be measured Analyzing objectives for determining pre-requisite skills Developing, table of specifications for test Determining specifications for items 2. Development Stage This consists of i. Writing items according to specifications

ii. iii. iv. v. vi. vii.

Developing needed art work and illustrations Writing response directions and examples Writing administrative directions Reviewing items illustrations and directions. Developing test lay-out Developing simple test 3. Conducting Field Test. At this stage, test is tried out through. i. ii. iii. iv. v. vi. Selecting representative group. Administering test. Scoring Analyzing information. analyzing data and selecting items Developing final lest form.

4. Developing Test Norm. Norms of the lest are developed through i. ii. iii. iv. v. vi. Describing characteristics of population. Selecting representative norm group. Administering lest to norm group, Scoring. Converting raw scores to standard scores. Creating norm tables.

5. Writing Test Manual. The test manual is written through i. ii. iii. iv. v. Described the design process and skills measured Describing criteria used to select items Describing norm group selection procedure Describing norm group characteristics Describing lest characteristics reliability and standard error of measurement,

vi. vii. viii.

Describing standard administrative procedures Describing scoring procedures and derivation of standard scores Describing score interpretation procedures.

Reliability of Norm-Referenced Test Test length affects reliability. Other things being equal, the reliability of the test can be increased by increasing its length. Items of similar content also increase reliability. Items of moderate difficulty increase reliability over the items which are either too easy or too difficult. Increased range of performance of the examinees being tested tends to increase reliability. Validity of a Norm-Referenced Test i. ii. iii. iv. v. The validity of a norm-referenced test can be increased by Constructing items of proper difficulty level Increasing the test length Increasing the heterogeneity of the group which is being tested Administering the test under proper conditions

Suggestions for interpreting examinee's scores on Norm-Referenced Test Bescolt, G.A., Balow, J.H., Hogan, T.P. and Far, R.C., in their book Metropolitan Achievement Tests (1978) suggest the following points for score interpretation on norm-referenced tests:(a) Generally students should obtain the same scores which arc expected from them by a regular classroom teacher. If it does not happen and the students obtain too high or too low scores, the teacher should explore by instructional implications. (b) Generally any student should score within the same range in all the subjects. If there is too much gap in the score of one subject

in comparison with the other subject, the teacher should try to locate the reason for it in motivation, special interests and special difficulties, etc.
(c)

Small differences in sub-test scores may be due to chance error also for which the standard error of measurement should be used. (d) Information collected from various tests should be used to explain performance on other tests. Uses of Norm-referenced Testing N. Vasantha Ram Kumar and K.N. Lalithamal (1990), state that N.R.T. is useful (1) In aptitude testing for making differential prediction; (2) To get a reliable rank ordering of the pupils with respect to the achievement we are measuring; (3) To identify the pupils who have mastered the essentials of the course more than others; (4) To select the best of the applicants for a particular programme; (5) To find out how effective a programme is in comparison to other possible programmer. Drawbacks of Norm-referenced Testing According to the above mentioned authors some of the criticisms raised against these tests are: (1) Test items that are answered correctly by most of the pupils are not included in these tests because of their inadequate contribution to response variance. They will be' the items that deal with important concepts of course content. (2) There is lack of congruence between what the test measures and what is stressed in a local curriculum. (3) Norm-referencing promotes unhealthy competition and is injurious to selfconcepts of low scoring students.