bulletWho We are  
bulletWhat We Offer  
bulletWhy We Excel  
bulletContact Us
bullet Free Stuff         


Reliability & Validity
General comments on the subject of reliability and validity

The problem posed by measuring individual potentials arises from the fact that there is no "standard" such as a meter-bar in the metric system.

It thus follows that reliability is an inadequate criterion to determine whether a test is valid. Reliability is necessary but insufficient in itself to guarantee the validity. One can measure very accurately the wrong items!

The RELIABILITY of the U-TEST has been submitted both to the method of parallel testing and to the method of shortened versions (Spearman-Brown).

Correlation coefficients lay between .92 and .94 which is exceptionally high. Essentially these results are due to the experimental selection of the questions. Research in this field has been accomplished by Julia Lewis-Salmen.

The VALIDITY can only be determined by the comparison of the test results with the judgment of observers who know the tested person very well and only on items that can really be observed. The U-TEST attains hereby a correlation coefficient of .94 to .96.

This score is astonishingly high. It is so high as to appear even unbelievable. The best way to verify for yourself this "unbelievably" high score is to test, by means of the U-TEST, a couple of people that you know very well.

The ultimate test of the U-TEST is answered by knowing: is the candidate the way the U-TEST describes and is the personís performance as indicated?

Users have been very astonished by its precision.

You will be able to see the reliability of the results obtained, for yourself.


Capacity Analysis
The U-TEST is essentially an individualís capacity analysis.

Capacity: relates to skills and to potential. The capacity to learn a new job, the capacity to tackle problems, the capacity to perform on a job.

Analysis: denotes a very profound and detailed investigation.

Productivity is the relationship between "efforts made" and "results obtained". The personal qualities determine the behavior and, consequently, the efforts made by the candidate. However, it is not sufficient to make efforts. One should also obtain results. In other words, one must be efficient. The U-TEST measures those factors which are related to "efficiency": accuracy, speed, persistence, correct estimation, etc.. The U-TEST is not just a "character analysis". It allows you, by means of relevant information, to invest in the right people.

The U-TEST is not a psychological test. It assesses specific potentials and shows the effect of these on productivity.

This is the relationship between two variable items. A positive correlation means that if item A increases, then item B will also increase accordingly. A negative correlation means that if A raises, B declines.

Correlation Coefficient
This is a figure indicating the level of correlation. This number is always between 0 and 1. A correlation coefficient equal to 1 signifies a complete relationship, i.e. if the magnitude of one item is known, then the other can be immediately deducted. If the correlation between two items is 0, then this means that there is absolutely no relation between the two.

This is the degree of precision of a test, i.e. the degree to which the variables measured by the test are free from errors of measurement. There are different methods to measure reliability. A first method consists of testing the same person by means of "parallel" tests, i.e. by different tests that are based on the same principles. Differences are then expressed by a correlation coefficient. Another method consists of using shortened versions of the same test and to calculate the correlation coefficient using the Spearman-Brown formula.

This is the appropriateness that the test really measures what it is supposed to measure. Validity is expressed as a correlation coefficient between results predicted by the test and those results of which one can be sure, e.g. by means of observation. 

Back to the U-MAN Edge