Zuckerman-Kuhlman-Aluja Personality Questionnaire (ZKA-PQ) and Cloninger’s Temperament and Character Inventory Revised (TCI-R): A comparative study

: The aim of this study was to analyze the psychometric properties (normal distribution values, reliabilities and factor structure) of Zuckerman-Kuhlman-Aluja Personality Questionnaire (ZKA-PQ) and the Temperament and Character Inventory revised (TCI-R), the similarities and differences between both facet-factor questionnaires. The total sample consisted of 482 subjects (53.1 % men and 46.9 % women) from diverse age. Results showed somewhat better psychometric properties, like reliability and facet-factor structure, for the ZKA-PQ than the TCI-R. The expected five factor facet structure of the ZKA-PQ was clear found. However, the seven factor structure of TCI-R was not clear and it did not show a clear distinction between Temperament and Character factors. When ZKA-PQ and TCI-R variables are analysed together, the ZKA-PQ factors are related to the Character as well as the Temperament factors. In some cases they represent the opposite poles of ZKA-PQ factors, e.g. Neuroticism versus Self-Directiveness and Aggression versus Cooperativeness. Some are directly and highly related to ZKA-PQ factors, e.g. Sensation Seeking and Novelty Seeking, Extraversion and Reward Dependence


INTRODUCTION
7 information. This sample is different from that used in the ZKA-PQ original study . (Restlessness) and AC4 (Work Energy). Factor scores are the sums of the four facets for each factor. The Spanish version of the ZKA-PQ  was used in the present study.

Descriptive statistics
Means, standard deviations, normal distribution values, alphas and t-test (and Cohen's d) comparing genders in ZKA-PQ and TCI-R are shown in Tables 1 and 2, respectively. ZKA-PQ showed appropriate skewness indexes and reliability coefficients in the present sample (+1). Only one facet (AC3: Restlessness) had an alpha below 0.60). Most were similar to those reported in the original study in a Spanish population . T-tests showed sex differences with males significantly higher than females on three Sensation Seeking facets, and females higher on all Neuroticism and three Extraversion facets. For the TCI-R, skewness and kurtosis indexes were good and most reliability coefficients were adequate except for 9 of the 29 facets with very low alpha coefficients (<0.60). Females scored significantly higher than males on Harm-Avoidance, Reward Dependence and Cooperativeness.   Table 4 shows the factor analysis of the Temperament and Character facets of the TCI-R.
As previously reported in the literature, no clear distinction between Temperament and Character factors can be observed. Firstly, only two facets have loadings larger than 0.40 on the first factor (NS1 and NS2). All facets of Harm Avoidance and three facets of Self-Directiveness had their highest loadings on a second factor. Harm Avoidance is a temperament trait and Self-Directiveness is a character trait yet in the factor analysis they simply represent the opposite poles of a temperament dimension most appropriately called Neuroticism. The third and fourth factors are mainly defined by Reward Dependence and Persistence facets, respectively. The fifth factor is more heterogeneous being composed of four facets of the Cooperativeness character dimension, three facets of temperament (NS4, RD4, and PS3) and one facet of Self-Directiveness (SD4; Selfacceptance). The Self-Transcendence facets defined the sixth factor. Finally, no facet had a high loading on the last factor, the largest loading was for C1 (Social Acceptance; 0.389). As can be seen, the structure of four temperament and three character factors is not supported. Some factors reflect, specially the second and the fifth ones, a mixture of Temperament and Character facets and others (the first and seventh ones) are somewhat difficult to interpret. Five of the 7 factors have at least one of their facets loading more highly on another factor other than the one they are supposed to define. Overall the facet structure shows a weakness, particularly of discriminant validity.
In the 5 factor solution (The Kaiser-Meyer-Olkin Measure of Sampling Adequacy was 0.887, and Bartlett's Test of Sphericity: Approx. Chi-Square: 13012.82; d.f.: 1176; p <0.001.) the first factor of the 5 factor analysis includes all 4 of the Neuroticism facets from the ZKA-PQ and these are the highest loading facets of the factor. However, it also includes two Aggression facets (AG3, anger, and AG4, hostility) from the ZKA-PQ; two Harm Avoidance Scales (HA 1, worry, HA4, fatigability) from the TCI-R, and (negative loadings) 3 Self-Directive facets (SD 1,3,& 5) from the TCI-R. This bipolar factor may be called Neuroticism vs. Self-Directiveness.
The third factor contains 3 of the 4 facets of the Extraversion factor from the ZKA-PQ, 3 of the 4 Reward Dependence facets from the TCI-R, and all 5 of the Cooperativeness facets from the TCI-R. In terms of the highest loading facets it could be labelled Extraversion or sociability, but it also includes many elements of Cooperativeness or what is called "Agreeableness" in the FFM.
The fourth factor is formed by all 4 of the Activity facets from the ZKA-PQ and the 4 Persistence facets of the TCI-R. It may be called Activity/Persistence or Energy. The fifth factor consisted of the 3 Self-Transcendence facets from the TCI-R.
Although the analysis confirmed 4 of the 5 factors of the ZKA-PQ, aggression facets were split between behavioural expressions (physical and verbal aggression) which loaded on the Sensation Seeking factor and emotional expressions (anger and hostility) that loaded on the Neuroticism factor (del Barrio, Aluja, Spielberger, 2004).  Avoidance facets of the TCI-R, and negative loadings from 3 of the 5 Self-Directiveness facets from the TCI-R. The 2 Harm Avoidance facets loading this factor were anticipatory worry (HA 1) and Fatigability (HA4). Shyness (HA3) loaded negatively on the Extraversion factor and fear of uncertainty (HA2) loaded negatively on the Sensation Seeking factor. In Cloninger's model Self-Directiveness is a Character rather than a Temperament trait but in this analysis it appears to represent the stable pole of Neuroticism.
The Extraversion factor is more narrowly defined in the 7 factor solution including facets of Warmth and Sociability from both tests. All 4 ZKA-PQ Extraversion facets, including Exhibitionism, are included.
The Sensation Seeking factor is again defined by all 4 of its facets from the ZKA-PQ plus 2 Novelty Seeking and one Harm Avoidance facet from the TCI-R. Aggressiveness facets do not load on this factor as they did in the 5 factor analysis. The Activity factor from the ZKA-PQ and the Persistence factor from the TCI-R are closely related. What they have in common is a strong energetic work motivation with high standards for achievement. They should be useful in applied research on predicting work effectiveness.
The fifth factor contains all of the Sensation Seeking facets from the ZKA-PQ and two of the Novelty Seeking facets (NS1, Exploratory Excitability, and NS4, Disorderliness) from the TCI-R. As noted previously, fear of uncertainty (HA2) from the Harm Avoidance factor of the TCI-R, loads negatively on this factor. The sixth factor consists solely of the 3 Self-Transcendence facets from the TCI-R. The seventh and weakest factor consists solely of two impulsivity scales: NS2 and NS3 from the TCI-R and a secondary loading of Boredom Susceptibility/Impulsivity (SS4) from the ZKA-PQ.
In regard to the psychometric properties, the ZKA-PQ dimension with the highest reliability coefficient was Aggressiveness (0.91) and the lowest was Sensation Seeking (0.86).
The coefficients for the dimensions of the TCI-R range between 0.87 and 0.63, for Persistence and Novelty Seeking, respectively. In deserves to be mentioned that the low reliability of Novelty Seeking as well as the lack of significant gender differences on this trait generally observed in many other studies assessing similar constructs (i.e., Costa, Terracciano & McCrae, 2001;Ball, Farnil, & Wangeman, 1984;Zuckerman, 1994) argue against the validity of this scale. In general, the ZKA-PQ is more reliable than TCI-R considering that the mean reliability coefficient of all dimensions and facets of the ZKA-PQ and TCI-R was 0.76 and 0.67, respectively.
On regard to the factor structure, the five factor solution for the ZKA-PQ resembles the original one reported for American and Spanish samples  with almost no secondary loadings for any facet. On the contrary, the factor structure for the TCI-R did not support the distinction between temperament and character since some factors were formed by a combination of facets of both kinds. It is highlighted that the last two factors were difficult to interpret given the lack of relevant loadings on them. Therefore, results of the current study do not justify Cloninger's distinction between Temperament and Character traits.
As expected, the most intense relationships between both models are reported for the Cloninger's Temperament dimensions. Neuroticism is highly related with Harm Avoidance, Cloninger's Temperament trait Persistency with Activity in the ZKA-PQ, and Novelty Seeking with Sensation Seeking. Focusing on the character traits, the Self-Directiveness factor is the reverse of Neuroticism and the Cooperativeness dimension is the opposite pole of the Aggressiveness factor in the ZKA-PQ. Only Self-transcendence is unique to the TCI-R and distinctive from other personality traits in the ZKA-PQ and the remaining variables of the TCI-R.
It should be remarked that, although a Self-transcendence or spiritual trait may be meaningful and valid in some contexts, there is no evidence of it as a basic personality trait in other systems.
Factorial validity is also better for the ZKA-PQ. Only one of the 20 ZKA-PQ facets (Exhibitionism, EX3) loaded higher (slightly) on another factor (Sensation Seeking) than the one to which it had been assigned (Extraversion). In fact, only 3 facets had moderate secondary loadings on a factor other than the one to which they were supposed to belong. In contrast 5 of the 7 TCI-R factors had at least one facet loading more highly on another factor than the one to which they were assigned.
Bearing in mind the better psychometric properties and factor structure of ZKA-PQ, this could be view as the preferred choice in practical and research setting. This statement may be particular appropriate in, for instance, the measurement of Sensation Seeking trait. Since construct validity of the Zuckerman's Sensation seeking scale as well as its psychometric properties are better than the corresponding scale in the Cloninger's system (Novelty Seeking), the SS trait may be more useful to predict disorders related with impulsivity and the need for new and risky activities as, for instance, Antisocial and other Cluster B personality disorders (Zuckerman, 1999) than do Novelty Seeking. On the other hand, research findings may advocate for using the TCI-R in different contexts. In this way, Self-Directiveness trait has been observed to be especially useful in the description of personality disorders (Svrakic, et al., 2002). Since this trait is not well covered by the Zuckerman's model, it may add some incremental power to the Zuckerman's dimensions in the prediction of psychiatric disorders. Another example is the relevance of two TCI-R character traits, Self-directedness and Cooperativeness, for well-being (Cloninger & Zohar, 2011).
A limitation of this study is the nature of the sample. Although it is far from being composed by university students only, the proportion of highly educated people is large, so it is not entirely representative of the general population.
Summing up, there is certainly much in common between ZKA-PQ and TCI-R factors using facets: Neuroticism and Harm-Avoidance, Aggression and Cooperativeness, Sensation Seeking and Novelty Seeking, Extraversion and Reward Dependence, and Activity and Persistence. However, considering that the actual correlations between the factors vary from 0.51 to 0.69, there is enough room for some factor variance not shared and specific to each test.
Helpful studies to understand this not shared variance may analyse each instrument conjointly with other personality measures. Since the FFM supposes a somewhat different view of personality compared to this psychobiological approach, and the NEO-PI-R is also composed by facets, this instrument could be of special interest. Some papers have already addressed this issue (De Fruyt, et al., 2010;García et al., in press).