According to our first hypothesis, judgments about oneself (self and ideal self) will be characterized with positive (i.e., emergent) characteristics, while the different groups with disabilities will be characterized with less desirable characteristics. Table 3 depicts the main results concerning average scores for each element on potency, evaluation, and activity. It also shows the percentage of extreme scores on the implicit (i.e., scores from 1 to three) and the emergent (scores from 5 to 7) poles, and on neutral, ambiguity or undefined (i.e., score of 4) responses. Thus, according to our predictions, the Self and Ideal Self average scores are higher than the scores for the other groups. Ratings on Friends tend to be located between our Self and our Ideal Self, denoting proximity. The different groups with disabilities are farther from ourselves, and the percentage of responses in the emergent pole is less, above all the group with intellectual disabilities.

In addition, ANOVA test revealed that non-significant differences on ratings of elements concerning individuals with disabilities were found, regardless of having or not previous contact with these individuals. Similarly, no significant differences were found based on the year (freshmen, sophomore, etc.) of the students.

Table 3 Table 3 Descriptive statistics on potency, activity and evaluation for the different elements of the Semantic Differential

Continuing with our first hypothesis, we expect that judgments about ourselves will be described as both warm and competent and Table 4 shows that those average scores are in fact, high for Self-scores. According to our second prediction, average scores on Self show the primacy effect of competence over warmth. Friends, enemies, and people with physical disability show that effect as well. On the contrary, the Ideal Self, people with intellectual disabilities, and people with Down Syndrome show the reverse pattern, which could denote distance in the judgments.

As we predicted in our third hypothesis, people with different disability levels are the subject of mixed feelings, and so, while in ratings on individuals with physical disabilities the primacy of competence vs. warmth dimension appears, for the ratings of individuals with intellectual disabilities and with Down syndrome, the pattern is reversed.

As we stated in our fourth hypotheses, the primacy of warmth over competence appears on judgments of distant elements, whereas for oneself and close elements, the opposite pattern appears. Moreover, analysis of differences in mean scores (repeated measures t-test) indicates that self-scores on competence were significantly higher than self-scores on warmth. Conversely, warmth scores for intellectual disability and Down syndrome are significantly higher than competence scores. Therefore, the paternalistic pattern, which in turn increases the possibility of arousing emotions such as overprotection or pity, appears more clearly in these two groups. These results support the fifth hypothesis.

Table 4 Table 4 Competence and warmth scores for the different elements, with Pearson’s correlations, and repeated measures t-Tests **

significant with p < 0.001

Finally, using the Idiogrid software, we calculated mean scores for each element and we performed a Principal Components Analysis (PCA) (promax rotation) to summarize the numerical information from the Grid. The analysis resulted in two components. The first component, with an eigenvalue of 13.27, explains 82.94% of total variance. The second component, with an eigenvalue of 1.54 explains 9.61% of total variance. The cumulative explained variance was 92.55%. Results are summarized in Figure 1. By looking at the first component, all the emergent poles of the constructs are on the right side of the graphic. The ideal self is located at the right side of the first component, in contrast to the Enemy element that is plotted on the opposite side. In consequence, the first component seems to deal with the evaluative (good vs. bad) or warmth (warmth vs. cold) component of social judgments. Meanwhile, the second component seems to deal with the competence dimension, given that the top part of the graphic includes constructs such as adult, competent, able, etc. Most of the elements are well-defined by this two-dimensional space, except Physical disabilities. The elements concerning people with intellectual disabilities and people with Down Syndrome are located toward the inferior pole of the component on competence, denoting low ratings on this component. Taking in mind these results, it is possible to say that the elements plotted in the first quadrant: Self, Friend, and Ideal self are rated as positive and competent. The second quadrant includes the element Enemy that is defined by negative evaluation and high competence. The third quadrant, which includes people with intellectual disabilities, is mostly defined by low competence. The fourth quadrant includes the label Down syndrome which is characterized by positive evaluation and low competence. It is also defined by constructs related to warmth dimension.

Figure 1 Figure 1 Principal Components Analysis (PCA) promax rotation

The combined utilization of two models on social perception, together with the Idiogrid software have allowed us to graphically present social representation and assigned status of different groups of people. Several findings are noteworthy as well. First, individuals without disabilities tend to rate people with disabilities as more similar to those who share the label, than to those who do not. Also, regarding the debate on competence vs. warmth, people with a cognitive condition (i.e., intellectual disability or Down syndrome) are considered more warm than competent, whereas those who do not share such condition are considered more competent than warm. The label associated with a cognitive condition appears to have a halo effect resulting in a paternalistic and stereotyped view that reduces the perception of interpersonal differences between diverse groups. Considering that stereotypes develop to justify prejudice (Crandall, Bahns, Warner, & Schaller, 2011; Salas et al., 2017), and that different disability labels evoke negative expectancies and bias (Foster & Ysseldyke, 1976), it is important to utilize “sound” and “simple” graphic representations of our social world to further promote our understanding of the role played by our bias, prejudices, and expectancies, as a starting point to develop strategies to remove those stereotypes.

A more detailed analysis according to the proposed model by Osgood et al. (1957) allows recognition that, while individuals with physical disabilities are subject to relatively better ratings, individuals with intellectual disabilities receive the lowest scores, and individuals with Down syndrome obtain medium-high ratings. Added to this is the fact that, while the group of people with physical disabilities achieve a high percentage of responses that denote uncertainty or ambiguity, the Down syndrome element obtains the highest percentage of responses in the emerging pole for potency, activity and evaluation. Finally, people with intellectual disabilities obtain the highest percentage of responses in the implicit pole for the activity domain. These findings suggest that people with Down syndrome are subject to relatively more favorable attitudes than other disability conditions. In this group, the responses denoting ambiguity are relatively low, indicating a greater willingness to characterize this group as very similar to each other. In other words, contrary to one of the classic premises of the semantic differential, where a score of four is considered neutral or “meaningless” (Clevenger, Lazier, & Clark, 1968), we consider this score as indicative of efforts to reduce overgeneralization and simplification, by stating that “it depends” on the specific characteristics of the individual under consideration. However, the attribution of features to the Down syndrome group, such as sympathy or kindness, are labels under which paternalistic attitudes are hidden and can ultimately reduce their rights. If we also consider that, in the group of people with physical disabilities, there is a primacy of competence vs. warmth compared to what happens to people with intellectual disabilities and Down syndrome, in which the dimension warmth prevails over competence, it seems clear that goodness over capability is a reasonably widespread perception. Additionally, although the Ideal Self stresses warmth over competence as well, the differences appear even more evidently after the PCA. Thus, while the Ideal Self obtains high ratings in evaluation and competence, people with intellectual disabilities and Down syndrome are rated as having low competence.

An additional noteworthy result relates to the fact that, although the three-dimensional model proposed by Osgood et al. (1957) can be effectively and reliably used to identify social perceptions and stereotypes as well as concomitant profiles and even emotions. The bidimensional model resulting from the Stereotype Content Model proposed by Fiske et al. (2002) fits the perception of differences better when rating the interpersonal world of those well-known and even unknown protagonists. This finding agrees with Sayans-Jiménez et al., (2017) where bi-factor models showed greater utility to directly and more easily explore the stereotype content including its evaluative content. Thus, Principal Components Analysis yielded two factors with high explanatory power, the first one dealing with the evaluative domain regarding warm vs. cold characteristics where the second one dealing with the competence dimension. As this dimension decreases, the elements are considered less capable and competent but, at the same time, holders of personal characteristics more related to goodness and lacking danger and menace. These results agree with previous findings with different assessment techniques (Carlsson & Björklund, 2010; Cuddy et al., 2008; Cuddy et al., 2009; de Paula Couto & Koller, 2012; Durante, Capozza, & Fiske, 2010; Nier et al., 2013), and serve to reclaim a more egalitarian vision focused on rights, equal opportunities and nondiscrimination against people with disabilities. Otherwise, we will be helping to maintain stereotypes that in turn perpetuate inequality.

Finally, some cautions should be noted relating to the sampling selection and size and the fact that participants in the second study were all college students. Further studies should focus on other possible variables that could be influencing current results. Likewise, the utilization of additional measures will help analyze additional psychometric properties (e.g., convergent validity) of the developed measure. Also, the inclusion of additional elements may help us analyze social perception, stereotypes, and prejudices toward many other disadvantaged groups who see how their status quo is justified for the sake of their own health or personal safety or that of those around them.