The selection tests aim at measuring such skills and abilities in a worker that are decided by job analysis to be essential for successful job performance. A test is an instrument designed to measure selected psychological factors.
Cronbach defines it as, “A test is a systematic procedure for comparing the behaviour of two or more persons.”
Learn about:- 1. Meaning of Selection Tests 2. Categories of Selection Tests 3. Validation 4. Reliability 5. Process 6. Advantages 7. Limitations 8. Precautions.
Selection Tests: Meaning, Types, Validation, Advantages, Limitations, Precautions and Process
Selection Tests – Meaning and Definitions
The selection tests aim at measuring such skills and abilities in a worker that are decided by job analysis to be essential for successful job performance. A test is an instrument designed to measure selected psychological factors.
ADVERTISEMENTS:
The basic assumption underlies the use of tests in personnel selection is that the individuals are different in their job-related abilities and skills and that these skills can be adequately and accurately measured for comparison. A great number of human abilities are complex and interrelated. Hence, these have to be understood in association with each other.
Selection tests are defined as under:
Cronbach defines it as, “A test is a systematic procedure for comparing the behaviour of two or more persons.”
Milton M. Blum opines, “A test is a sample of an aspect of an individual’s behaviour, performance and attitude.”
Selection Tests – 5 Important Categories of Selection Test
Tests used in selection can be classified into five categories, namely:
ADVERTISEMENTS:
1. Achievement,
2. Aptitude,
3. Interest,
ADVERTISEMENTS:
4. Personality and
5. Intelligence tests.
1. Achievement Tests:
Achievement tests measure the job knowledge of the applicants in the areas such as marketing, HR and economics. When an applicant claims to know something, an achievement test is given to measure how well he/she knows it. It is for this reason that achievement tests are also known as proficiency tests or performance tests.
Achievement tests are a refinement of the work sample technique and are used to measure what the applicant actually can do and what tasks he/she can perform right now. Adaptability tests, abstract reasoning tests, Stanford-Binet Scales, Miller Analogies Test and Purdue test for machinists and machine operators are some of the examples of such tests.
Trade tests are the most common type of achievement tests given, for example, in a typing test, matter to be typed is provided to the candidate and the time taken by the candidate in typing that matter as also the errors made by him/her are noted down, and thus his/her proficiency in typing is measured.
Similar tests are available in shorthand, calculating machines, operating calculators, simple mechanical equipment, dictating and transcribing apparatus.
2. Aptitude Tests:
Aptitude tests are also known as potential ability tests and specific cognitive tests and are used to measure the latent ability of a candidate to learn a given job if he/she is given the required training. Aptitude tests have special significance where the new recruit has little or very less experience along the lines of the job opening.
Specific aptitude tests have to be designed for jobs that require mechanical, clerical, musical, academic, linguistic, motor capacities and abilities. Bennett Test of Mechanical Comprehension and the Stenquist Mechanical Aptitude Test are some of the examples of popular mechanical aptitude tests.
ADVERTISEMENTS:
Aptitude tests for clerical jobs lay emphasis mainly on arithmetic, handwriting, spellings, vocabulary, checking and so on. Aptitude tests help in detecting peculiarities or defects in the applicant’s sensory or intellectual capacity.
3. Interest Tests:
Interest in a job or task contributes to success on the job. A person interested in his/her job is likely to do better than one who is indifferent or uninterested. Interest tests have been designed to discover a person’s field of interest and to identify the kind of work that will satisfy him/her. They are in a sense inventories of the likes and dislikes of the people of some occupations.
Although the interest tests have been mostly standardised, some organisations have developed tests to suit their own requirements, especially when there is a continuing demand for a large number of new personnel for a single job. Such developed tests are also known as ‘Tailor- made Tests’.
Most widely used interest scales are ‘Kuder Preference Record’, ‘Strong Vocational Interest Blank’, mechanical reasoning tests and so on.
4. Personality Tests:
ADVERTISEMENTS:
These tests aim at measuring the basic make-up or characteristics of an individual. They assess his/her introversion motivation, emotional reactions, emotional maturity, stability, mood, value system, ability to adjust, interpersonal relations, self-image, self- confidence, ambition, tact, optimism, decisiveness, sociability, objectivity, patience, fear, distrust, suspicion, judgement, dominance, impulsiveness, integrity, stability and so on. Industrial psychologists usually emphasise the ‘big five’ personality dimensions as they apply to personnel testing.
Theses dimensions are as follows:
i. Extraversion (sociable, active, energy, zeal and so on)
ii. Emotional stability/neurolism (poor conditions, adjustment, insecurity, hostility, anxiety and so on)
ADVERTISEMENTS:
iii. Agreeableness (trust, caring, gentle and so on)
iv. Conscientiousness (achievement and dependability)
v. Openness to experience (imaginative, unconventional, autonomous).
These are normally conducted for selecting supervisors and higher executives. These are pen and paper tests.
5. Intelligence Tests:
These tests aim at measuring the general level of intelligence of the applicant. This is done by measuring the IQ of the applicant. In addition to this, they also measure a range of abilities such as numerical ability, vocabulary, memory and verbal fluency.
However, intelligence is usually measured with individually administered tests such as the Wechsler Adult Intelligence Scale, Wonderlic Personnel Test, Stanford-Binet Test, California Test of Mental Maturity (Adult Level), Kaufman Adolescent and Adult Intelligence Test, Minnesota Paper Form Board Test (MPFB), Comprehensive Test of Verbal Intelligence, Wide Range Intelligence Test and Slosson Intelligence Test.
ADVERTISEMENTS:
Some organisations also conduct the polygraphy and honesty tests.
Motor and Physical Abilities Tests:
Such tests usually measure the speed and accuracy of simple judgement as well as the speed of fingers, hands and arms movement. Some of the popular tests in this regard include the Stromberg Dexterity Test, the Crawford Small Parts Dexterity Test, the Minnesota Rate of Manipulation Test and so on.
Tests are also divided into the following three categories:
1. Objective tests – These tests measure neurotic tendencies, self-sufficiency, dominance, submission and self-control.
2. Projective tests – The candidate is asked to project his/her own interpretation into certain standard stimulus situations. The way he/she responds to these stimuli depends on his/her own values, motives and personality (e.g., Rorschach Blot Test and Thematic Apperception Test).
ADVERTISEMENTS:
3. Situation tests – These tests measure the applicant’s reaction when he/she is placed in a peculiar situation. Normally, in a leaderless group, problem is posed and solution is to be found out. A situation test requires the examinees to respond to situations representative of jobs. Video-based simulation is a situational test in which an examinee responds to video simulations of a realistic job situation.
Selection Tests – Validation
Tests have become quite common means for determining the suitability of candidates for employment. The assumption in using tests is that those who score above the cut-off point are suitable for employment while those scoring below this point are unsuitable for employment. This assumption is valid only when tests themselves have validity.
Validity of a test is the degree to which its prediction is supported by evidence, that is, higher scorers are better job performers. Validation is the process of ensuring validity in a test. Validity of a test is necessary because of two reasons-
First, validity in a test is necessary to improve selection decisions. If the test is not valid, decisions based on the test are no better than decisions without it.
Second, validity in a test ensures equity, that is, all applicants have been judged on fair basis; there is no unfair discrimination against any candidate or group of candidates. There are three types of validity that must be ensured in tests- content validity, construct validity, and criterion-related validity.
1. Content Validity:
ADVERTISEMENTS:
Content validity is the degree to which a test measures the knowledge, skills, and abilities used on the job. All candidates for that job are given the same test so that applicants can be properly compared. A simple example of a content-valid test is a typing test.
Such a test can approximate the work to be done on the job; the applicants can be given a typical sample of typing, and their performance can be evaluated on that sample. Assuming that the work on the test constitutes an accurate sample of work on the job, the test is content valid.
2. Construct Validity:
Construct validity is the degree to which a test measures a particular trait related to successful performance on the job. A trait is usually abstract in nature such as intelligence or emotion, and is called construct. Unlike content validity, construct validity is quite complex and difficult because it deals with abstract.
3. Criterion-Related Validity:
Criterion-related validity is the degree to which a particular test accurately predicts the level of performance or important elements of work behaviour. This validity shows the relationship between some predictor (say, test score) and a criterion of job success (say, production output, managerial effectiveness, etc.).
ADVERTISEMENTS:
The relationship between predictor and criterion is worked out in the form of validity coefficient which can vary, theoretically, from +1.0 to -1.0, with former showing the highest positive coefficient and latter showing the highest negative coefficient. The stronger the positive relationship, the more valid is the test.
Predictive Versus Concurrent Validation:
For establishing validity in a test, either of two methods can be applied- predictive validity and concurrent validity. Though the basic objective in both the cases is same, that is, determining the level of validity of a test, their procedure is quite different.
4. Predictive Validity:
Predictive validity is the degree to which there is a relationship between test score and some future desirable behaviour on the job. Predictive validity validates a test by using prospective applicants as a study group. The test is administered with its unknown validity on all prospective candidates.
At this level, the candidates are selected for employment on the basis of some previously used selection criteria. However, their test scores are kept as a safe record. After a lapse of some time, usually a year or so, these employees are evaluated on the basis of their actual job performance.
The evaluation scores are compared with test scores of each employee to judge the relationship between test as a predictor and actual job performance as a criterion. If the coefficient is high, the test is considered as valid. If the coefficient is low, the test is worked out again. Though this method is costly and time-consuming, it is preferable as it demonstrates relationship between the predictor and criterion on actual basis.
5. Concurrent Validity:
An alternative to predictive validity is concurrent validity which is the degree to which a test score is related to job behaviour measure that is available at the same time. Concurrent validation method validates a test using the present employees as the subjects. Test scores are immediately compared with the actual performance data of the employees to find out the relationship between the predictor and criterion.
If the relationship coefficient is high, it shows that the test is valid. However, the concurrent validity leaves many questions to be answered. It has been challenged on the premise that the present employees know the job already and that a learning process takes time. Only considering the time and resources saved, a well-conducted concurrent validity may be preferable to predictive validity.
The above two validation methods provide a means of obtaining validation information taking small samples either of applicants or of present employees in a particular job. As such, they may not reflect the diversity of future employees. To some extent, synthetic validity is the right choice.
6. Synthetic Validity:
Synthetic validity involves taking parts of several similar jobs rather than one job to validate a test. Thus, validity evidence is obtained by relating predictor scores to the same job dimension across several jobs, combining the employees into a larger sample. For example, a typing ability test might be validated against typing performance not only for typing clerks but other personnel involved in typing work in an organization.
Selection Tests – Reliability (With Methods)
Apart from validity, tests must have reliability, that is, they must have consistency in measurement whatever they measure. This implies that an applicant’s performance on any given test should produce same score each time the test is administered to him. If an applicant scores significantly different on the same test at two points of time, the test cannot be considered reliable.
Reliability of a test has implications in measuring a person’s personality traits which do not change with time and not in knowledge and skills which change with time. To measure reliability of a test, test-retest method, alternative form method, and split halves method can be applied.
1. Test-Retest Method:
In test-retest method, a test is administered to an individual at two different time periods and the scores obtained are compared. If the scores at two points of time are similar, the test is said to be reliable.
2. Alternative Form Method:
In alternative form method, two similar forms of a test are administered to an individual. Their scores are recorded and compared to each other. If the two scores are similar, the test is treated as reliable.
3. Split Halves Method:
In this method, a test is divided into two similar and equal parts. The scores of both the parts are compared. If scores of both the parts are similar, the test is treated as reliable. This method measures the internal consistency of the test.
Selection Tests – Process for Developing a Testing Programme: Step by Step Process
When an organization decides to use tests as an instrument for selection process, it must develop a test programme which includes the choice of different tests and administering those tests so that the test results are optimal. Development of testing programme and its implementation requires careful planning, analysis, and experiment which calls for specific skills.
If the HR department is not adequately prepared for this, it is advisable to engage industrial psychologist for the purpose.
In developing testing programme, the following steps are required:
Step # 1. Defining Job Performance Criteria and Predictors:
The basic objective of selection tests is to find a match between a job and job holder so that the job is performed effectively. For this purpose, the job holder must possess certain traits or qualities. However, what those qualities are and the degree to which these qualities must be possessed by the prospective job holder must be ascertained precisely.
Job description and job specification are the instruments which provide information for this. If the job specification has been prepared in detail indicating various traits, these can be taken as predictors of job performance. It implies that those who possess these traits tend to show better performance than those who do not.
Step # 2. Choosing Tests for Measuring Traits:
After determining the traits that must be possessed by the job holders, the type of tests that are likely to measure these traits must be chosen. In choosing the tests, it must be ensured that these are valid and reliable, that is, they measure the traits specified and measure these accurately.
In order to use the tests for selection purpose, it is desirable to administer these tests on the present employees to determine their validity. Apart from validity and reliability, other factors which should be considered in the choice of tests are ease of administration, cost involved, and evaluation of results.
Step # 3. Administering Tests:
When the tests are chosen and their validity and reliability are established, these tests can be administered to test candidates’ traits. In administering the tests, care should be taken that physical and psychological situations are conducive and the candidates are at ease to show their usual performance.
Step # 4. Analyzing the Results of the Tests:
When testing is over, the results of the tests should be calculated. Most of the tests are objective type with each answer of a question predicts the possession or non-possession of a particular trait. These answers may be assigned points to obtain total score. If facilities exist, calculation of results should be computerized to quicken the process. After obtaining the test scores, these may be analyzed, interpreted, and used in selection process.
In using the test results, there may be situations when a candidate may have consistently high scores on all the tests or inversely, he may have low scores consistently. Not much problem comes in this situation as higher scorers may be considered for selection while lower scorers may be weeded out. In another situation, a candidate may have outperformed in some tests but underperformed in others.
In such a case, taking aggregate of all the tests for the purpose of comparison may pose a problem. In order to solve this problem, either a cut-off point is taken as the basis to ensure that the candidate has at least the minimum degree of all the desirable traits or weight may be provided on the basis of relative importance of different desirable traits. In the weight system, the test scores are multiplied by the weights to obtain aggregate which may become comparable.
Selection Tests – 3 Main Advantages
Various steps of selection process including selection tests are meant to solicit information about the candidates so as to arrive at a decision to select the most desirable candidates out of the several available. Since only some information is available from other sources like application forms, references, etc. selection tests are used to solicit more information about the candidates.
The use of selection tests has the following advantages:
1. Selection tests are standardized and unbiased methods of soliciting information about the prospective employees. Thus, a person who does not get selected on the basis of selection tests cannot argue for the partiality in selection process. It is to be noted that in many organizations, impartiality in selection process is of prime importance like public sector organizations.
2. Selection tests can be used to weed out the large number of candidates who may not be considered for employment in the organization. Normally, organizations receive applications from large number of candidates seeking jobs in the organizations. They all meet the basic requirements of the jobs, but all cannot be called for interview because it is a very time-consuming process.
Selection tests will provide the cut-off point above which candidates can be called for interview. Thus, the tests will save lot of time and money.
3. The tests are able to uncover the qualities and potentials of prospective employees which cannot be known by other methods including personal interview. Since the people are taken in the organization not only for the present jobs but they are promoted over the period of time, tests provide good opportunities to test their potentials for such promotions also.
Thus, tests are a more reliable source for predicting the overall suitability of candidates for the employment.
Selection Tests – Limitations
Selection tests may provide some useful information but they suffer from some limitations also. Two such limitations are quite important which suggest that use of tests should be supplemented by other means of soliciting information about the candidates-
1. Selection tests cannot make a hundred per cent prediction of an individual’s on-the- job success. At best, they just reveal that those who have scored above the cut-off point (other things being equal) will be more successful than those who have scored below the cut-off point. Therefore, it is desirable to use tests only as supplementary method of selection.
2. These are suitable when there is large number of candidates for limited number of jobs or positions. If the number of candidates is small, the use of tests is not economical. In such a case, it is desirable to select persons on the basis of interview only.
Besides these two limitations, many people criticize tests on the basis that these discriminate against the deprived classes because affluent classes can know better how to defeat the tests. However, this limitation can be overcome by suitable design and administration of tests.
A further criticism is put against the tests that these invade privacy of people because these tests put many questions on the personal life of the candidates.
Selection Tests – 6 Main Precautions in Using Selection Tests
Various limitations of selection tests suggest that these should be used cautiously. Test results can improve decision making regarding selection of employees, and as a result, organizations can select the best possible candidates if the following precautions are taken in using the tests-
1. A test can be effective only when it has validity. The validity of a test is the degree to which it measures what it intends to measure. Thus, a valid test is one which accurately predicts the criteria of job success. In the absence of validity of test, it may give wrong result and the persons selected on its basis may be even inferior to those who have been rejected.
2. Another feature which is important in a test is its reliability. The reliability of a test is the consistency with which it yields the same scores throughout a series of measurements. Thus, if a test has high reliability, a person who is tested a second or third time with the same test and under the same conditions will obtain the same result.
3. Norms should be developed as a source of reference on all tests used in selection. Norms are standardized scores that help translate raw scores into a comparative statement. Some companies use minimum marks obtained in a test while others use cut-off score for selecting candidates for further selection process. Norms can be fixed on the basis of success or failure of employees in the organization selected on the basis of similar tests.
4. Tests should not be used merely to decide cut-off point for weeding out the candidates. Instead, the test administrators should ensure that tests have validity and therefore, these can play important role in the selection process. In order to take maximum advantage of tests, proper weight can be given to scores in the tests. This may help in avoiding personal biases in the selection considerably.
5. Test administration, scoring, and interpretation require technical competence and training in testing. Therefore, tests should be handled by properly trained and competent people. In fact, some of the standardized tests insist that their use be restricted to registered approved users only so that these are not used for the selection which may give horrible results.
6. The tests should be used as an additional factor in selection process and should not be treated as the sole basis for selecting or rejecting a candidate. If other sources of information like application forms, references, and interviews are used along with test scores, decision making regarding the selection of candidates is improved. This is the reason that organizations use a combination of different methods in selection process.