Around the world, hundreds of millions of psychometric tests are used every year for educational purposes, placement of employees and certification of professionals. But many have started to wonder about how an increasing scepticism for the quality of these tests and their possible bias has been addressed.
Aptitude Test Fairness
The question as to whether mental tests are fair, both in their technical application and in their content has stimulated much discussion and research. Psychometric testing is now big business with many companies and large retail chains utilizing this form of testing. With so much hanging on the outcome of the tests, those who use them are understandably right in demanding evidence that the tests are reliable and fair to the individuals who are being tested. What is it then, that they should look for to have this question answered?
Both the British Psychological Society and the American Psychological Association have issued guidelines for the design and use of psychological testing. Their concern is not only directed to technical criteria but also to those who market and administer the tests. The responsibility to ensure fair and proper use of tests therefore lies with a number of different parties. There are four fundamental requirements for all psychometric tests: reliability, validity, objectivity and standardisation.
Personality and Aptitude Test Reliability
Reliability refers to the consistency of the measure provided by the test. Psychologists use three main techniques to assess test reliability. The first method is the Test-Retest Method, which involves a correlation of results from the same test which has been given to the same people after a period of time. This is all very well in tests which won’t be affected by practice. There are some people, however, who may remember the answers from the first round of testing, while others may have forgotten the answers. This variable can result in skewed test results, particularly when it comes to the issue of repeatability and consistency. There will also have been different experiences between each test which may affect and skew the results.
Parallel Test Reliability in Psychometric Testing
An alternative method is the “parallel test reliability” (sometimes called “alternate forms method”) as used in the 1937 version of Stanford-Binet. This involves two tests composed of different items that are designed to measure exactly the same trait in exactly the same way. Correlation between the scores could well be affected in tests with items that can be improved by practice, or where familiarity makes them easier. There can also be problems in constructing truly equivalent forms of the test. Other issues to consider with regard to test reliability are the time difference between the tests and also, will the same investigator score the second test and if not, could this make a difference to the results?
Psychometric Tests – Techniques to Check Their Reliability
One other technique is the split-half method where after the test has been administered, items are divided into two equal parts, each scored separately and the results then correlated. But tests can be divided in many ways so it might not be easy to split the test in two equivalent halves, especially if the test contains group items of different difficulty levels. Other considerations should include the effects of possible over-anxiety at the start of the test and fatigue as the test wears on.
Psychometric Assessments and How they are Administered
Because personality and psychometric tests have become the focus litigation in some cases, the development and administration of these tests is guided by professional standards such as the Standards for Teacher Competence in Educational Assessment of Students, the Code of Fair Testing Practices in Education and the Code of Professional Responsibilities in Educational Measurement as well as government regulations such as Equal Employment Opportunity Commissions Regulation on Employment Testing. Most well established companies who provide psychometric tests operate at the highest ethical and professional standards. Even so, with many new companies trying to break into this lucrative market, there is a diverse range of competence and integrity throughout the testing industry.
Sources:
- Annastasi A. Psychological Testing. Sixth Edition Macmillan Publishing Co., London. 1988.
- Cronback, L. Essentials of Psychological Testing. Harper International Edition, New York and London. 1970.
- Edenborough R. Using Psychometrics. Kogan Page Ltd, London. 1974.
- Hayes, N. Foundations of Psychology. Routledge, London. 1994.
- Helmstadter G. Principles of Psychological Measurement. Methuen & Co. London. 1996.
