THE PROBLEM WITH PERSONALITY TESTS
What’s wrong with this scene?
You walk into a psychologists office and you ask him to help you get your life on track. Almost by habit, he reaches into his top right hand desk drawer and pulls out a folder with some forms and hands you a copy. It’s a questionnaire! Professionally, they are called “Self-Report Tests” (also referred to as “inventories” or “instruments”). There may be anywhere from 25 to a 100 or more responses required, and when the results are summed up and sorted out, your true self will be magically revealed to you. This routine has been going on for around 125 years. Various statistical methods (mostly factor analysis) have been developed to tease out significant relationships between the questions. The goal is to create a more accurate test, employing the words we commonly use to describe people (also called the lexical approach). Don’t bother asking the psychologist to guarantee that the horoscope-like description of you is really you.
Do you see the problem yet? The problem is that the test is asking YOU the questions. The answers come from, YOU, the very person who has been frustrated your whole life because YOU don’t have a clue about what kind of person YOU are. Suddenly, you are the expert that is supposed to know the precise meanings and implications of the various descriptive words, ideas and behaviors and in addition to that you are supposed to know exactly, the degree to which these words apply to you. You are supposed to rank the various descriptions, or assign some value on a scale of one to five to each item.
On top of that, remember, humans are very tricky when asked questions about themselves. Look up the etymology or historical background of the word “personality”. It comes from the word “persona”, which refers to the mask that actors wore in the ancient theater. Too often, our personality is what we want people to think we are or what we want to imagine we are. So we are deceptive in our social interactions and we tend to lie in varying degrees on psychological tests.
Then there is the question of how well people understand the descriptive terms used on a test. What’s the difference between “anger” and “frustration?” Or is there a difference between being “careful” and “cautious” or “fearful” and “worried?” We don’t have a dictionary in our heads, let alone a thesaurus.
It gets worse. What about the reliability problem? This has to do with the fact that people score differently when given the same test after a several month interval. The publishers of various tests boast about test-retest correlations between 70-90%. This would be considered “good” to “excellent.” But there is no way of knowing just how reliable a particular person’s ratings are without retaking the test. And if you do retake the test which score is the correct one? This means that there is a good chance that your health, attitude, anxiety level, amount of sleep, and/or personal problems will make your results inaccurate.
We haven’t even discussed the validity of what is being measured. Are the various types real? What are we talking about? Are we really explaining personality or are we presenting a model of how words commonly used to describe personality are related. Some tests might even be considered word usage tests more than personality tests. And finally, is this the best we can do after all of these years of research? There is the claim that the problems with self-report tests can be managed. That hardly inspires confidence. Management of tests merely adds another layer of things that have to be measured. What is clearly lacking is agreement on a general theory of the basic dimensions of personality.
Topic: The Problem With Personality Tests (Read 926 times)
June 16, 2008, 11:00:39 AM