Advantages of Covert Assessment of Highly Desirable or Undesirable Traits
Overt personality tests (i.e., tests with items that obviously relate to
the traits being measured) may be inadequate in many applied situations.
This is especially the case when the traits being assessed are either
highly desirable (e.g., achieving tendency, emotional empathy,
integrity, affiliative tendency) or highly undesirable (violence risk,
aggressiveness, trait anxiety, depression). This section provides
information on ways to assess such desirable or undesirable traits when it is important to minimize
faking by those being tested.
The PAD Temperament Model, described in the following link, provides covert, indirect,
and camouflaged ways for assessing various aspects of personality and psychopathology.
The three basic scales of the model are Trait
Pleasure-Displeasure (P), Trait Arousability (A), and Trait
Dominance-Submissiveness (D). For covert assessments, participants are
administered the PAD scales which bear little surface resemblance to
traits that require measurement (e.g., violence risk). Next, equations
such as those given in the material below are used to calculate scores for the desired
All the computational work is done in a software package that uses scores on the PAD scales
to derive estimated scores on 41 personality traits (including highly desirable
or undesirable traits). The software includes additional safeguards that detect when respondents attempt to slant their answers to make good impressions. When respondents slant answers, scores on socially desirable traits become inflated or greater than they would have been in the absence of response slanting; similarly, scores on socially undesirable traits are decreased. The PAD Temperament software includes a built-in feature to help detect when such response slanting/faking occurs and warns the tester when results become "unacceptable."
Detailed information about the software is given in the following link:
Computation of Trait Scores Using PAD Scores
Using the software, participants are first administered the three PAD scales. Next, equations
1 through 3 are used to convert participant scores to z-scores (i.e., standardized scores that have mean values of zero and standard deviations of one).
Once standardized PAD scores are available for a particular participant,
covert trait scores are computed using equations that are based on data
reviewed by Mehrabian (1996b). The following 4 equations illustrate the general approach for desirable traits. Each of the following equations is written for standardized Trait Pleasure-Displeasure (P), Trait Arousability (A), and Trait Dominance-Submissiveness (D) scores. Also,
to aid the interpretation of computed scores, coefficients in these
equations are adjusted so the computed trait scores are also
standardized (with mean values of zero and standard deviations of one).
- 1. Trait Pleasure z-score = (Trait Pleasure raw score - 38)/22
- 2. Trait Arousability z-score = (Trait Arousability raw score - 30)/33
- 3. Trait Dominance z-score = (Trait Dominance raw score - 11)/33
Equation 4 is taken from Mehrabian and O'Reilly (1980, equation 20),
except that all coefficients are multiplied by 1.628 to yield
standardized Achieving Tendency scores. Equation 5 is based on findings
by Mehrabian (1997b, equation 5) showing that the Balanced Emotional
Empathy Scale is essentially an equally weighted sum of Trait Pleasure
and Trait Arousability scores. Equation 6 is taken from Mehrabian and
O'Reilly (1980, equation 8), except that, once again, all coefficients
are multiplied by 1.895 to yield standardized Affiliative Tendency
scores. Finally, equation 7 is based on findings obtained by Mehrabian
(1996a, equation 3C) showing that Conscientiousness is a nearly equally
weighted sum of Trait Pleasure and Trait Dominance scores.
- 4. Achieving Tendency = 1.628*(.13*P + .60*D) = .21*P + .98*D
- 5. Emotional Empathy = .707*(P + A)
- 6. Affiliative Tendency = 1.895*(.47*P + .24*A) = .89*P + .45*A
- 7. Conscientiousness = .707*(P + D)
Computation of undesirable trait scores is illustrated by the following
Equation 8 is founded on findings relating the PAD scales to the
Mehrabian Trait Anxiety, the Spielberger, Gorsuch, and Lushene Trait
Anxiety, and the Eysenck Neuroticism scales (Mehrabian, 1996b, equations
34-36). The coefficient for P in equation 8 is the average of the three
P coefficients given by Mehrabian (1996b, equations 34-36). The
coefficient for A in equation 8 is similarly the average of the three A
coefficients in the aforementioned equations. Finally, the coefficient
for D in equation 8 is the average of the D coefficients in the same
three equations. The multiplying factor of 1.662 is used so that
computed Trait Anxiety scores are standardized for easy interpretation.
- 8. Trait Anxiety = 1.662*(-.39*P +.37*A -.27*D) = -.65*P +.61*A -.48*D
- 9. Depression = 1.838*(-.38*P +.17*A -.35*D) = -.70*P +.31*A -.64*D
- 10. Aggressiveness = 2.01*(-.36*P +.20*A +.28*D) = -.72*P +.40*A +.56*D
- 11. Violence Risk = 1.549*(-.63*P + .14*D) = -.98*P + .22*D
Equation 9 is based on findings relating an extremely broad-based and
multi-factor measure of Depression to the PAD scales (Mehrabian, 1997b,
equation 9). Equation 10 is taken from Mehrabian and O'Reilly (1980,
equation 30). Finally, equation 11 is taken from Mehrabian (1997b,
Overall, equations 4 through 11 illustrate the PAD-based covert
assessment of personality traits. Details of the PAD Temperament Model
are given in the section below. Information regarding covert assessment
of other traits can be obtained from Albert Mehrabian (see contact
information at the end of this page).
Mehrabian, A. (1987). Eating characteristics and temperament:
General measures and interrelationships. Springer-Verlag, New York.
Mehrabian, A. (1991). Outline of a general emotion-based theory of
temperament. In J. Strelau and A. Angleitner (Eds.), Explorations in
temperament: International perspectives on theory and measurement
(pp. 75-86). Plenum Press, New York.
Mehrabian, A. (1995a). Framework for a comprehensive description and
measurement of emotional states. Genetic, Social, and General
Psychology Monographs, 121, 339-361.
Mehrabian, A. (1995b). Relationships among three general approaches to
personality description. Journal of Psychology, 129, 565-581.
Mehrabian, A. (1996a). Analysis of the Big-five personality factors
in terms of the PAD Temperament Model. Australian Journal of
Psychology, 48, 86-92.
Mehrabian, A. (1996b). Pleasure-arousal-dominance: A general
framework for describing and measuring individual differences in
temperament. Current Psychology, 14, 261-292.
Mehrabian, A. (1997a). Comparison of the PAD and PANAS as models for
describing emotions and for differentiating anxiety from depression.
Journal of Psychopathology and Behavioral Assessment, 19,
Mehrabian, A. (1997b). Relations among personality scales of
aggression, violence, and empathy: Validational evidence bearing on the
Risk of Eruptive Violence Scale. Aggressive Behavior, 23,
Mehrabian, A. (2000). Beyond IQ: Broad-based measurement of individual
success potential or "emotional intelligence." Genetic, Social, and
General Psychology Monographs, 126, 133-239.
Mehrabian, A., & O'Reilly, E. (1980). Analysis of personality measures
in terms of basic dimensions of temperament. Journal of Personality
and Social Psychology, 38, 492-503.