Reliability determines whether a given test is valuable. This is what distinguishes a diagnostic test from a horoscope. Unfortunately, it is reliability that is often lacking, even in professional solutions based on scientific tests.
What does it mean that a test is reliable?
The reliability of a test tells us about the repeatability and accuracy of its measurement. So if we test several or several dozen people many times with the same test, the results should be consistent.
Let’s imagine two archery targets: on both we can see 30 hits made by the same person. On the first target the hits are scattered over the entire target, but on the second the hits are all grouped together around the center. The second target symbolizes a reliable test. We can see that the person hit the bullseye many times and it wasn’t accidental.
The reliability of psychological tests is most often measured using a coefficient called Cronbach’s Alpha. The higher it is, the better, and its limiting value is assumed to be 0.7-0.75. The value of this coefficient tells us the percentage of variability of the test results depending on the actual variance and percentage of error variance (Hornowska, 2017).
Why is test reliability important?
A psychometric test must have three main parameters to be valid: accuracy, reliability and appropriate standards (e.g.: APA – American Psychological Association). The credibility of a test is based on these parameters, and a high-level of reliability is the basis for choosing a given diagnostic tool.
Diagnostic tests of dubious quality, i.e. those that are not scientifically verified, introduce errors into the decision-making process instead of helping to make rational decisions. The costs are borne by both the respondents and the researching parties, i.e. businesses, and both educational and medical institutions.
What can the use of unverified tests lead to?
Case Study:
A recruitment process is launched based on selecting the best candidate for the position. The company in question did not entirely think through the choice of the screening test they used, but decided to hire the candidate who did best on it. It gradually became clear that this person’s profile did not correspond to the tasks and responsibilities connected to the position they were hired for.
The employee experienced enormous stress and suffered a drop in motivation. The work of the entire team he joined was affected as well. In the end, the company had to repeat the recruitment process anyway, which was not by any means the end of the story for the unprofessionally selected employee. The financial and moral consequences in this case clearly underline the importance of choosing the right tools.
Few people realize that tests such as the MBTI, which is widely used in the business world, is not a scientifically recognized test. This is due to low parameters (Pittenger, 1993).
In the wilds of the Internet you can also find tons of quizzes and tests with little more than horoscope value or of the teen magazine pop-psychology fun variety. These tests are especially harmful because many of them actually use the term “psychological test.”
How do we approach test reliability with Empatyzer?
For us, reliability is absolutely fundamental because Empatyzer is not just an application. It’s an approach based on scientific evidence and tests that meet all the parameters mentioned above. For personality tests, we use the scientifically proven IPIP test (in its Polish adaptation), among others. Thanks to this, we have a guarantee that the results we get from it will allow us to build an effective tool that solves many problems.
Empatyzer will help users communicate better from day one. We are confident that this will have a positive impact on many areas of both education and business.
