Method

Results

First, we distributed the colors by pair according to age and verified that 64 (37%) children out of the 172 who answered both the instruments presented at least one of the colors by pair. Next, because a protocol may have more than one colors by pair indicator, we verified the distribution according to the number of colors by pair; the ones with the highest number of indicators included four pairs (Table 2).

The Chi-square test was used to verify whether there were statistically significant differences between the presence and absence of colors by pair according to the sociodemographic variables (sex and age). No significant differences were found for sex (x² = 0.11, p = 0.515, Eta = 0.007) but were found for age (x² = 12.484, p = 0.029, Eta = 0.246). Eleven-year-old children presented more colors by pair than younger ones, while eight-year-old children presented the lowest number of colors by pair.

We also compared the number of colors by pair according to the children’s age, and a statistically significant difference was found (x² = 42.063, p = 0.03, Eta = 0.243). Note that the 11-year-old children presented more than one color per pair and the younger children presented fewer combinations of colors per pair per protocol.

Table 3 presents the variables of the ZSC structural summary that presented statistically significant differences after comparing the presence and absence of each of the four most frequent colors by pair (Vd↓Vm↑, La↓Vi↓, Vm↑Pr↑, and Vd↓Az↑) and the pair suggesting healthy functioning (Vd↓Az↑) between the groups. The number of children with each of these pairs, as shown in Table 3, is smaller than the number of children presenting these pairs in Table 1, considering that not all the children in the database had completed the ZSC protocol. Therefore, we emphasize that the number of participants considered in each group is presented in Table 3.

Table 3 shows that 11 of the ZSC variables presented significant differences (p < 0.05) because some pairs were more frequent than others, and each of them enabled differentiating two or three of the ZSC variables. The analysis of Vd↓Az↑ showed that nine variables presented significant differences.

Discussion

We opted to emphasize the understanding of CPT’s colors by pair in this study because there is a lack of studies addressing validity evidence for this indicator. However, before discussing the results concerning the colors by pair, we note that three of the four most frequent pairs in the CPT database included the red color (Table 1). This color is associated with more excited states, extroversion, impulsiveness, and aggressiveness (^{Villemor-Amaral, 2005}, ²⁰¹⁴). ^{Farah et al. (2014)} also report that red was a predominant color among children, and red is mentioned as one of the three most frequent colors found in children’s protocols (^{Villemor-Amaral, 2014}).

Regarding the understanding of colors by pair, the four most frequent pairs were Vd↓Vm↑, La↓Vi↓, Vm↑Br↑, and Vm↑Pr↑. According to ^{Villemor-Amaral (2005)}, Vd↓Vm↑ refers to irritability and impulsiveness without the individual having the resources to elaborate further, which hinders coordinated and coherent actions. La↓Vi↓ suggests a sluggish and not very productive attitude, while Vm↑Br↑ reveals excitability and impulsiveness in a weakened structure, indicating uncontrolled actions and a disorganized attitude. Finally, Vm↑Pr↑ indicates accentuated conflicts with a profound feeling of dissatisfaction. ^{Villemor Amaral (1978)} noted that, in general, the colors by pair would be related to the examinee’s particular functioning, possibly associated with clinical conditions with a tendency toward emotional disorganization.

The comparisons based on age revealed an increase in the frequency of colors by pair among 11-year-old children (Table 2). ^{Villemor Amaral (1978)} had already reported that two of the four most frequent pairs found in this study, Vd↓Vm↑ and Vm↑Pr↑, are frequently found among adolescents. The increased frequency of colors by pair among preadolescents and adolescents may be associated with conflicts typical of this stage of development. Studies addressing other indicators also found age differences explained by puberty (^{Villemor-Amaral et al., 2012}; ^{Villemor-Amaral & Quirino, 2013}; ^{Cardoso et al., 2018}).

The results concerning ZSC show that few variables presented significant differences due to the CPT’s colors by pair (Table 3). Thus, even though we may explore various potential interpretations, caution is needed regarding the generalization of the results.

Regarding the most frequent colors by pair, the M, Hd, and PHR variables in the ZSC were less frequent in the group that presented the pair Vd↓Vm↑. M is part of the indicators related to processing. It indicates abstract reasoning and superior forms of conceptualization, while Hd and PHR concern interpersonal relationships and can only be interpreted when increased or compared to the remaining answers involving a perception of human content (^{Villemor-Amaral & Primi, 2009}). As previously mentioned, Vd↓Vm↑ suggests irritable and impulsive behavior without resources to elaborate and, due to decreased green, social restriction, possibly resulting in a difficulty keeping coordinated and coherent actions (^{Villemor-Amaral, 2005}). Therefore, decreased M in the group of children presenting Vd↓Vm↑ suggests that the more irritable and impulsive behaviors in the sample are potentially associated with difficulties in interpersonal relationships and decreased abstract reasoning.

Another piece of information provided in Table 3 concerns a decrease in the FC’ and COP indicators among the children presenting La↓Vi↓. FC’ is present in the ZSC affection dimension and suggests not expressed negative affects that are being internalized, while COP indicates a tendency to perceive interpersonal relationships positively and being available to establish cooperative relationships (^{Villemor-Amaral & Primi, 2009}). Considering that La↓Vi↓ suggests low productivity and a sluggish attitude (^{Villemor-Amaral, 2005}), there is evidence of a relationship between more inactive functioning with less tendency toward cooperative behavior and less tendency to internalize negative affect.

Table 3 also shows a decrease in the Y, es, and esAj indicators among children presenting Vm↑Br↑. These indicators belong to the ZSC control and stress tolerance. The Y indicator suggests an intense load of emotional stress, and when es and esAj are increased, they suggest intense stimulation above that an examinee can tolerate when making decisions or implementing actions (^{Villemor-Amaral & Primi, 2009}). Even though decreased Y, es, and esAj were related to the presence of the pair Vm↑Br↑, which indicates excitability and impulsiveness, potential uncontrolled actions, and disorganized behavior (^{Villemor-Amaral, 2005}), these ZSC indicators would only have interpretative meaning when in increased levels.

Table 3 shows decreased P, D, and DAj in the group of children presenting the pair Vm↑Pr↑. A decrease in P answers, incorporated in the ZSC mediation, suggests difficulty perceiving the environment’s conventional aspects, while decreased D and DAj, which belong to control and stress tolerance, indicate difficulty managing stressful situations that hinder decision-making (^{Villemor-Amaral & Primi, 2009}). Furthermore, because the pair Vm↑Pr↑ reflects accentuated internal conflicts and feelings of dissatisfaction (^{Villemor Amaral, 1978}), these colors suggest a tendency toward decreased perceptions of the environment’s conventional aspects and difficulty dealing with stressful situations among the children who, according to the CPT, expressed dissatisfaction and internal conflicts.

The Chi-square test, used to compare the two groups according to the Vd↓Az↑ pair, revealed nine variables with statistically significant differences (Table 3). The only variable with decreased values in this group was F%, while the remaining variables appeared at increased levels.

The F%, es, esAj, and C’ variables belong to the ZSC control and stress tolerance. Decreased F% suggests excessive openness to experiences and a very broad focus of attention, which indicates that the individual may sometimes fail to process information in its entirety. The es, esAj, and C’ variables indicate that the intensity and frequency of a stimulus are more demanding than the actions an individual can prepare and implement at a given time. Hence, the individual tends to refrain from emotionally expressing unpleasant affect (^{Villemor-Amaral & Primi, 2009}).

The interpretation of an increase in blend% would depend on the remaining ZSC indicators, considering that, when combined with good resources, blend% suggests that the individual can entirely consider stimuli and has a good ability to work. Nonetheless, the combination of this indicator with insufficient resources suggests control difficulties and disorganization (^{Villemor-Amaral & Primi, 2009}).

The PHR, H, and Hd indicators need to be analyzed together because they are all associated with interpersonal relationships and self-image. An increase in these variables suggests an interest in others associated with difficulty establishing relationships due to a partial and suspicious perception of others, with conduct less adapted than expected in interpersonal relationships (^{Villemor-Amaral & Primi, 2009}).

The interpretation of both CPT and ZSC variables suggests that establishing social restrictions to attempt to control and maintain more authentic relationships of Vd↓Az↑ in the CPT (^{Villemor Amaral, 1978}; ^{Villemor-Amaral, 2005}) may be related to the containment of unpleasant affect and restricted affective contact due to a partial and suspicious perception of others, as indicated by ZSC (^{Villemor-Amaral & Primi, 2009}). Hence, after obtaining a theoretical understanding of the results, we need to relate these findings to our proposed hypotheses.

As for the hypothesis related to the children’s age, we expected that older children would present a more significant number of colors by pair. The reason is that previous studies (^{Villemor-Amaral et al., 2012}; ^{Villemor-Amaral & Quirino, 2013}; ^{Cardoso et al., 2018}) report that preadolescents presented evidence of emotional conflicts. The results found here corroborated this interpretation.

Even though the hypothesis that children who presented Vd↓Az↑ would also more frequently present variables in the ZCS that suggest healthy development was not confirmed, interpretation of both tests’ indicators provided complementary information about how these children relate. Likewise, the hypothesis that children more frequently presenting the remaining colors by pair would also present more ZCS variables suggesting less healthy development was not confirmed. These results suggest that, even though both the CPT and ZSC assess aspects of emotional dynamics and cognitive functioning, the ways these tests assess these aspects differ. Hence, further studies are needed to understand the correspondences and differences between these two methods.

Regarding this study’s limitations, although various studies have investigated Zulliger’s validity to be used among children, no manual or scientific publication provides national normative tables establishing the expected scores among children. Therefore, the analyses considered the raw scores of each indicator without effectively understanding what decreased or increased scores represented.

Note that the database used in this study was composed of the protocols of children without a history of psychological/psychiatric care or school failure. Perhaps, establishing criteria with specific emotional or cognitive functioning characteristics would result in a more significant number of children presenting colors by pair. Therefore, we suggest that future studies establish clinical criteria when considering colors by pair - for instance, composing a group of children considered aggressive and another group composed of non-aggressive children to compare the colors by pair associated with irritability and explosive emotional discharges.

Studies like this support understanding the possibilities and limitations of the integrated use of different projective methods. Additionally, studies can also contribute to the instruments’ psychometric quality (^{Ambiel & Carvalho, 2017}; ^{CFP, 2018}) and promote a reflection upon the extension and practical use of these resources to measure children’s adaptation (^{Borges & Baptista, 2018}). Therefore, even though the meaning of each of the colors by pair was not identified when the combined use of the ZCS variables was considered, this study contributes to the literature on colors by pair by showing the importance of new studies in understanding the psychometric quality of this seldom investigated indicator.