Aspects regarding reliability
This method of evaluation, as also all the methods based on specialised electronic devices is a distinct, autonomous method having maxim reliability and validity resulting from the conceptual characteristics, from fabric, from functionality and from the utilisation manner, having also a personal standardisation of the indicators. For this reason, it can not be used as alternative (substitutive) to other examining methods.
Pursuant to Standards for Educational and Psychological Tests, 1999, the reliability is defined as the “degree in which the test scores are consistent and repeatable, namely the degree in which these are not affected by the errors of measurement ".
It is known that one of the methods of verification of the reliability of a measurement is the one of the analysis of internal consistency. Generally, the coefficients of the internal consistency are calculated based on the scores obtained at a one time test. The Alpha - Cronbach coefficient belongs also to this category. It indicates the concordance of different parts of the test and in order to reflect a high reliability test it has to have a value very closed to 1. (Under 0.7 Alpha - Cronbach we can not talk about an evaluation method!).
In order to exemplify the degree of reliability of the method used by the BioSpecter Psychometric System we hereby present sequences of the stages of the SPSS analysis and the report of internal consistency for 50 items and a sample of 160 persons.
It can be noticed that at the above mentioned analysis of the internal consistency, the BioSpecter Psychometric System has achieved an Alpha-Cronbach coefficient of 0,984 showing a very high reliability!
Aspects regarding validity
Pursuant to Standards for Educational and Psychological Tests, 1999, the validity of a test is the degree in which the test is measuring what it is supposed to be measured. It means that what the test scores represent it has to known exactly and the measurements need to be correct. On these two aspects, we have previously shown, that the BioSpecter Psychometric System provides both the correctness of the measurement by eliminating the human factor and especially, by correlations between the measurement areas and the psychophysiological functions that projectively explores, the exact image of the significations of the determined scores.
The APA standards(1999) have in view the following categories of validity of the inferences done based on the test scores: the conceptual validity, content validity, the predictive and concurrent validity.
BioSpecter Psychometric System respond to the criterion of content validity!
Unlike the classical evaluation tests, the BioSpecter Psychometric System presents by the algorithm of the implemented method, the selection of content aspects toward which it directs the evaluation. This facility is possible due to the matricial analysis system both of the aspects related to abilities and skills but especially the ones that can exactly characterise the compatibility with elements of an activity profiles register and activity and professions of these profiles. The selection is automatically done by a fractal method using the whole range of psychophysiological indicators from the physical, instinctual area till the deep mental areas. By the constitution of the implemented algorithm is hereby totally eliminated the doubt on the significant of the items used.
BioSpecter Psychometric System respond to the criterion of construct validity!
As in the case of the previous criterion, from design, the algorithm of the Biospecter method provides the construct validity, the set of conducts related to each programmed construct being aprioristically identified and characterised.
The response of the BioSpecter Psychometric System to the criteria an procedures used for validation:
1.Age/development in time of the subject: uses the age of the evaluated person as variable in the determination of the correlations between the electric potentials and aptitude aspects;
2.Correlation to the other test: uses scales comparable to the scales of the classical evaluation methods. Due this reason, it is possible to achieve correlations to these tests, considering that Biospecter Psychometric System does not measure the experience, but only the psychophysiological aspects;
3.Factorial analysis: refers only to multiple choice tests, it does not apply to the BioSpecter Psychometric System;
4.Internal consistency: is provided by the analysis of items correlation;
5.Convergent and discriminant validation: is resolved by construction, inclusively the discriminant validation for the personality report;
6.Experimental method: responds to any tests that experimentally provoke a variable using a certain design or experimental pattern, under the compliance with the System’s operating principles;
BioSpecter Psychometric System respond to the criteria of predictive and concurrent validity!
The possibility to identify types of compatibilities of activity and profession profile, grants to the BioSpecter Psychometric System the ability to determine in case of persons dedicated to a certain profession, the specific aspects of those professions deducted from the compatibility criteria. By this extreme ability, BioSpecter Psychometric System aprioristically to a possible employment can identify the skills to the evaluated person! This way, the problem of time between test administration and obtaining the information it no longer exist. Such a determination is illustrated also in the following table, which contains exactly the elements of a concurrent validity where persons found themselves on a work position have been tested and they have been requested data related to real professional aspects:
The two sets of data have been subject to the correlation calculus. It reflects the fact that between the two sets there is a significant relation, namely the test has certainly a predictive value, conditioned on the test validity, more precisely, the parallelism between the results of the compatibility determination and the real professional aspects. This parallelism is estimated using the correlation coefficient.
Such as the case of reliability, the BioSpecter Psychometric System fulfils the conditions of a high validity, being classified in the expert system class, due to the implemented technical solution.
The limitations of the BioSpecter Psychometric System
 In the purpose that it can not be exclusively used for the purpose to determine a psychological profile. The psychological testing can have in view, with regard to the mental health criterion, not only the normality, but also the pathological manifestations.
Copyright © Biospecter System Ltd. 2013 | All rights reserved.