Is MoralDNA® a trustworthy psychometric tool?
Yes. Here is some background information about MoralDNA® and some technical information that can be assessed by psychometric experts.
Structure
MoralDNA® consists of 3 parts.
Part A
Measures the 3 ethics, in a Personal context.
Part B
Measures the same 3 ethics, in a Professional context.
Part C
Measures 10 moral values.
Theory
MoralDNA® uses a simplified Kohlberg model.
Kohlberg’s developmental theory mentions 6 stages of moral development.
Our model refers only to adolescents and adults. We measure our preferences for each of the 3 classical moral philosophies of Consequentialism, Virtue Ethics and Deontology, for which we use the more accessible terms People, Values and Rules.
- People (Consequentialism), where decisions are based on empathy and the well-being of other people affected by our decisions.
- Values (Virtue Ethics), where critical reasoning based on internalized values regardless of rules or consequences is employed.
- Rules (Deontology), where decisions are based on calculating individual consequences of obeying or disobeying law and regulations.
We measure these 3 ethics in 2 different contexts: Personal Life and Work.
We then measure the 10 moral values. Our current model focuses on 10 moral values:
Wisdom, Fairness, Courage, Self-Control, Trust, Hope, Humility, Love, Honesty & Excellence.
History
MoralDNA® started in 2008 and has gone through continuous improvements during all those years:
Scoring
Parts A & B are scored using a Bi-Factor IRT (Item Response Theory) model with an underlying polytomous Graded Response 2PL model (2 Parameter Logistic), using EAP (Expected A Posteriori) score extraction.
Part C is scored using ten IRT 2PL polytomous Graded Response models.
Reliability & Validity
When treated as a classical test, Cronbach A estimates range between .76 and .89 for the 3 ethics in both contexts and between .67 and .83 for the 10 values. Test-retest reliability was 0.67 uncorrected and 0.84 when corrected for measurement attenuation.
A significant 65% of respondents fully agree with the results, while only 2.5% stated that the results did not represent them at all. In most of the correlational studies presented above, results were always correlated with the expected direction as indicated by previous research.