STAGES IN TEST CONSTRUCTION


I.    Planning the Test

 

A.    Determining the Objectives

B.    Preparing the Table of Specifications

C.    Selecting the Appropriate Item Format

D.    Writing the Test Items

E.    Editing the Test Items

 

II.   Trying Out the Test

 

A.    Administering the First Tryout – then Item Analysis

B.    Administering the Second Try-out then Item Analysis

C.    Preparing the Final Form of the Test

 

III.  Establishing Test Validity

 

IV.  Establishing the Test Reliability

 

V.    Interpreting the Test Score

 

MAJOR CONSIDERATIONS IN TEST CONSTRUCTION

 

The following are the major considerations in test construction:


Type of Test

       Our usual idea of testing is an in-class test that is administered by the teacher. However, there are many variations on this theme: group tests, individual tests, written tests, oral tests, speed tests, power tests, pretests and post tests. Each of these has different characteristics that must be considered when the tests are planned.

 

       If it is a take-home test rather than an in-class test, how do you make sure that students work independently, have equal access to sources and resources, or spend a sufficient but not enormous amount of time on the task? If it is a pretest, should it

exactly match the past test so that a gain score can be computed, or should the pretest contain items that are diagnostic of prerequisite skills and knowledge?

If it is an achievement test, should partial credit be awarded, should there be penalties for guessing, or should points be deducted for grammar and spelling errors?

 

 

       Obviously, the test plan must include a wide array of issues. Anticipating these potential problems allows the test constructor to develop positions or policies that are consistent with his or her testing philosophy. These can then be communicated to students, administrators, parents and other who may be affected by the testing program. Make a list of the objectives, the subject matter taught, and activities undertaken. These are contained in the daily lesson plans of the teacher and in the references or textbook used. Such tests are usually very indirect methods that only approximate real-world applications. The constraints in classroom testing are often due to time and the development level of the students.

 

 

Test Length

 

       A major decision in the test planning is how many items should be included on the test. There should be enough to cover the content adequately but the length of the class period or the attention span or fatigue limits of the students usually restrict the test length. Decisions about test length are usually based on practical constraints more than on theoretical considerations.

 

       Most teachers want test scores to be determined by how much the student understands rather than by how quickly he or she answers the questions. Thus, teachers prefer power tests, where at least 90 percent of the students have time to attempt 90 percent of the test items. Just how many items will fit into a given test occasion is something that is learned through experience with similar groups of students.

 

 

Item Formats

 

       Determining what kind of items to include on the test is major decision. Should they be objectively scored formats such as multiple choice or matching type? Should they cause the students to organize their own thoughts through short answer or essay formats?

 

These are important questions that can be answered only by the teacher in terms of the local context, his or her students, his or her classroom, and the specific purpose of the test. Once the planning decision are made, the item writing begins. This tank is often the most feared by the beginning test constructors. However, the procedure are more common sense than formal rules.

 

POINTS TO BE CONSIDERED IN PREPARING A TEST

 

1.          Are the instructional objectives clearly defined?

2.          What knowledge, skills and attitudes do you want to measure?

3.          Did you prepare a table of specifications?

4.          Did you formulate well defined and clear test items?

5.          Did you employ correct English in writing the items?

6.          Did you avoid giving clues to the correct answer?

7.          Did you test the important ideas rather than the trivial?

8.          Did you adapt the test’s difficulty to your students ability?

9.          Did you avoid using textbook jargons?

10.      Did you cast the items in positive form?

11.      Did you prepare a scoring key?

12.      Does each item have a single correct answer?

13.      Did you review your items?

 

 

GENERAL PRINCIPLES IN CONSTRUCTING DIFFERENT TYPES OF TESTS

 

             The test items should be selected very carefully. Only important facts should be included.

             The test should have extensive sampling of items.

             The test items should be carefully expressed in simple, clear, definite, and meaningful sentences.

             There should be only one possible correct response for each test item.

             Each item should be independent. Leading clues to other items should be avoided.

             Lifting sentences from books should not be done to encourage thinking and understanding.

             The first person personal pronouns I and we should not be used.

             Various types of test items should be made to avoid monotony.

             Majority of the test items should be of moderate difficulty. Few difficult and few easy items should be included.

             The test items should be arranged in an ascending order of difficulty. Easy items should be at the beginning to encourage the examinee to pursue the test and the most difficult items should be at the end.

             Clear, concise, and complete directions should precede all types of test. Sample test items may be provided for expected responses.

             Items which can be answered by previous experience alone without knowledge of the subject matter should not be included.

             Catchy words should not be used in the test items.

             Test items must be based upon the objectives of the course and upon the course content.

             The test should measure the degree of achievement or determine the difficulties of the learners.

             The test should emphasize ability to apply and use facts as well as knowledge of facts.

             The test should be of such length that it can be completed within the time allotted by all or nearly all of the pupils. The teacher should perform the test herself to determine its approximate time allotment.

             Rules governing good language expression, grammar, spelling, punctuation, and capitalization should be observed in all items.

             Information on how scoring will be done should be provided.

             Scoring keys in correcting and scoring tests should be provided.


POINTERS TO BE OBSERVED IN CONSTRUCTING AND SCORING THE DIFFERENT TYPES OF TESTS

 

A.    RECALL TYPES

 

1.    Simple recall type

 

a.          This type consists of questions calling for a single word or expression as an answer.

 

b.          Items usually begin with who, where, when and what.

 

c.           Score is the number of correct answers.

 

 

2.    Completion type

 

a.          Only important words or phrases should be omitted to avoid confusion.

 

b.          Blanks should be of equal lengths.

 

c.           The blank, as much as possible, is placed near or at the end of the sentence.

 

d.          Articles a,an, and the should not be provided before the omitted word or phrase to avoid clues for answers.

 

3.    Enumeration type

 

a.          The exact number of expected answers should be stated

 

b.          Blanks should be of equal lengths

 

c.           Score is the number of correct answers.

 

 

4.    Identification type

 

a.          The items should make an examinee think of a word, number, or group of words that would complete the statement or answer the problem.

 

b.          Score is the number of correct answers.

 

B.    RECOGNITION TYPES

 

1.    True-false or alternate-response type

 

 

 

a.          Declarative sentences should be used.

 

b.          The number of “true and“false items should be more or less equal.

 

c.           The truth or falsity of the sentence should not be too evident.

 

d.          Negative statements should be avoided.

 

e.          The “modified true-false” is more preferable than the “plain true-false”

 

f.            In arranging the items, avoid the regular recurrence of “true” and “false” statements.

 

g.          Avoid using specific determiners like; all, always, never, None, nothing, most, often, some, etc. and avoid weak statements as may, sometimes, as a rule in general etc.

 

h.          Minimize the use of qualitative terms like: few, great, many, more, etc.

 

i.            Avoid leading clues to answers in all items.

 

j.            Score is the number of correct answers in “modified true-false and right answers minus wrong answers in “Plain true-false”.

 

 

2.    Yes-No type

 

a.          The items should be in interrogative sentences

b.          The same rules as in “true-false” are applied.

 

3.    Multiple-response type

 

a.          There should be three to five choices. The number of choices used in first item should be the same number of choices in all the items of this type of test.

 

b.          The choices should be numbered or lettered so that only the number or letter can be written on the bank provided.

 

c.           If the choices are figures, they should be arranged in ascending order.

 

d.          Avoid the use of “a” or “an” as the last word prior to the listing of the responses.

 

e.          Random occurrence of responses should be employed

 

f.            The choices should be related in some way or should be at the end of the statements.

g.          The choices should be related in some way or should belong to the same class.

 

h.          Avoid the use of “none of these” as one of the choices.

 

i.            Score is the number of correct answers.

 

4.    Best Answer type

 

a.          There should be three to five choices all of which are right but vary in their degree of merit, importance or desirability

 

b.          The other rules for multiple response items are applied here.

 

c.           Score is the number of correct answers.

 

 

5.    Matching type

 

 

             There should be two columns. Under “A” are the stimuli which should be longer and more descriptive than the responses under column “B”. The response may be a word, a phrase a number or a formula.

 

             The stimuli under column “A” should be numbered and the responses under column “B” should be lettered. Answers will be indicated by letters only on lines provided in column “A”

 

             The number of pairs usually should not exceed twenty items. Less than ten introduces chance elements. Twenty pairs may be used but more than twenty is decidedly wasteful of time.

 

             The number of responses in column “B” should be two or more than the number of items in Column “A” to avoid guessing.

 

             Only one correct matching for each item should be possible.

 

             Matching sets should neither be too long nor too short.

 

             All items should be on the same page to avoid turning of pages in the process of matching pairs.

 

             Score is the number of correct answers.


C.    ESSAY TYPE EXAMINATIONS

 

Common types of essay questions. (The types are related to purposes of which the essay examination are to be used.)

 

             Comparison of two things

 

             Explanation of the use or meaning of a statement or passage

 

             Analysis

 

             Decision for or against

 

             Discussion

 

How to construct essay examinations.

 

             Determine the objectives or essentials for each question to be evaluated

 

             Phrase question in simple, clear and concise language.

 

             Suit the length of the question to the time available for answering the essay examination. The teacher should try to answer the test herself.

             Scoring:

 

a.          Have model answer in advance

b.          Indicate the number of points for each question

c.           Score a point for each essential

 

ADVANTAGES AND DISADVANTAGES OF THE OBJECTIVE TYPE OF TESTS

 

Advantages

 

             The objective test is free from personal bias in scoring

 

             It is easy to score. With a scoring key, the test can be corrected by different individuals without affecting the accuracy of the grades given.

 

             It has high validity because it is comprehensive with wide sampling of essentials.

 

             It is less time-consuming since many items can be answered in a given time.

 

             It is fair to students since the slow writers can accomplish the test as fast as the fast writers.

 

 

Disadvantages

 

 

             It is difficult to construct and requires more time to prepare.

             It does not afford the students the opportunity in training for self- and thought organization

 

             It cannot be used to test ability in theme writing or journalistic writing.

 

ADVANTAGES AND DISADVANTAGES OF THE ESSAY TYPE OF TESTS

 

Advantages

 

             The essay examination can be used in practically all subjects of the school curriculum.

 

             It trains students for thought organization and self-expression

 

             It affords students opportunities to express their originality and independence of thinking

 

             Only the essay test can be used in some subjects like composition writing which cannot be tested by the objective type test.

 

             Essay examination measures higher mental abilities like comparison, interpretation, criticism, defense of opinion and decision.

 

             The essay test is easily prepared.

 

             It is expensive

 

 

Disadvantages

 

             The limited sampling of items makes the test unreliable measure of achievements or abilities

 

             Questions usually are not well prepared

 

             Scoring is highly subjective due to the influence of the correctors personal judgement

 

             Grading of the essay test is inaccurate measure of pupils achievements due to subjectivity of scoring

 

 

BACK