4 important Fundamental Principles of Test Construction:
The fundamental principles of test construction are such as (a) Validity, (b) Reliability (c) Standardisation (d) their evaluation.
Tests should have validity, that is, they should actually measure what they purport to measure. A perfectly valid test would prospective employees in exactly the same relationship to one another as they would stand after trial on the job.
The validity of a test is determined by the relationship between the test and some criterion of efficiency on the job. The coefficient of correlation has become the most widely employed index of validity. It is a Statistical index expressing the degree of relationship between the test results and some criterion of efficiency on the job.
There are two procedures by which the validity of a test is determined, each supplementary to the other, in the first instances, the test may be administered to employees of known ability already on the job.
Those employees who are known to be most efficient score highest on the test, while the least efficient employees score lowest, the test has validity.
The second procedure in checking on the validity of a test, and one that should be employed to supplement the foregoing consists in follow-up studies of the performance of these employees who have been selected through it. No test can be said to be truly successful until both these procedures have been employed.
By the reliability of a test is meant the consistency with which it serves as a measuring instrument.
If a test is reliable a person taking at two different times should make substantially the same score time. Under ideal conditions a test can never be any more than a sample of the ability being measured. No test is of value in personnel work unless it has a high degree of reliability. This is usually determined by one of three methods.
1. by giving the test to the same group at two separate times and correlating the resultant series of test scores.
2. by giving two or more different (but equivalent) forms of the same test and correlating the resultant test scores, or
3. but the so-called split-half or odd even, method.
When reliability is determined by the latter method, the test is given but once, but the items are divided and scores on one-half of the items are correlated with scores on the other half.
The process of Standardisation includes:
1. The scaling of test items in term of difficulty, and
2. The establishment of norms.
More important as an element in the Standardisation of personnel tests is the scaling of test items in terms of difficulty. To be of functional value in a test, each item must be of such difficulty as to be missed by a part of the examines but not by all.
That is, no item is of discriminatory value which is answered either correctly or incorrectly by everyone. Secondly, since items of different degrees of difficulty will discriminate between persons of different degrees of ability or achievement, it is essential that the items be well distributed along the difficulty scale.
However, there is difference of opinion amongst the psychologists in this matter of scaling of test items.
The norms can be established upon the basis of performance on the test over a period of time. Perfectly adequate norms for a given examination would embrace the entire population eligible for that examination.
For practical purposes, however, norms established upon the basis of a representative sample are sufficient. What constitutes adequacy, therefore, would vary for examinations for different positions in which varying qualifications are involved.
The evaluation of test results, involving as it does all the problems of scoring and weighting of items and the assignment of relative weights to tests used in a battery, it surrounded with highly technical considerations.