Introduction

Most societies tend to be organized hierarchically: some individuals usually exert a greater influence, are more respected or admired than others (^{Sidanius & Pratto, 1999}). Although there are several variables -like race, sex or personality traits- that have been used to explain such power and status differences among the distinct social agents, in this article we will examine the psychological consequences of one of the most relevant variables within the social hierarchization process: socioeconomic status or social class (^{Fiske & Markus, 2012}).

Belonging to a high or low social class brings important psychological consequences. Those who belong to a lower social class tend to be exposed to more pushing problems and social constraints than those in higher classes (^{Fiske & Markus, 2012}; ^{Krauss, Piff, Mendoza-Denton, Rheinschmidt, & Keltner, 2012}). Accordingly, the former generally show a lower sense of control; they tend to think that whatever happens in their lives is to some extent not up to them, but to the circumstances or social context (^{Kraus, Piff, & Keltner, 2009}; ^{Kraus et al., 2012}; ^{Lachman & Weaver, 1998}).

From this standpoint, this article will approach the effect of the subjective social class (SSC) on a variable related to the sense of control: generalized self-efficacy (^{Jerusalem & Schwarzer, 1992}; ^{Schwarzer, 1993}). Although previous studies have already addressed the effects of SSC on the sense of control (i.e., ^{Kraus et al., 2009}), most of them have measured the SSC as an individual difference (i.e., individual’s rank perception about their place within a certain society; ^{Kraus et al., 2009}, Studies 1-2), or it has been experimentally manipulated (i.e., by means of making participants feel that they belong to a lower or higher social class; ^{Kraus, Coté, & Keltner, 2010}, Study 3).

The aim of this article is to go one step beyond that and increase our knowledge about the consequences of social class, by jointly examining the influence of these two forms of operationalizing the SSC. That is to say, we will analyze the effect of the interaction between SSC as an individual difference (SSC-ID) and experimentally manipulated SSC (SSC-EM) on the generalized self-efficacy among the general population.

Likewise, not only will it be researched in this article the way in which the SSC influences the dependent variable (i.e., generalized self-efficacy), but also how the participants develop, through their own words, what it means to have low or high SSC. This is important because, though many times invisible, the human intentions and meanings are built upon the framework of social structures (^{Miles & Huberman, 1994}). Therefore, this study will make use of a mixed research technique: on the one hand, testing quantitatively the hypotheses about the consequences of the SSC on self-efficacy -measured both as individual and manipulated difference-; on the other hand, exploring qualitatively the way in which participants describe differences between social classes.

Objective and subjective social class. When discussing social classes, researchers differentiate between objective social class (OSC) and SSC (i.e., ^{Kraus et al., 2009}; ^{Piff, Stancato, Coté, Mendoza-Denton, & Keltner, 2012}). OSC, also known as socioeconomic status, refers to the amount of material and social resources-such as education and income level, or occupational reputation-which an individual can obtain or handle (^{Han, Chu, Song, & Li, 2015}; ^{Oakes & Rossi, 2003}). On the other hand, SSC indicates the subjective perception that people have about their position in the social hierarchy (^{Adler, Epel, Castellazo, & Ickovics, 2000}).

^{Kraus et al. (2009}) argue that the OSC measurement usually poses several problems. For example, there is discrepancy on what is the best way to combine the several aspects that comprise OSC. Some questions arise, such as: Is educational level more important than income level? Must family income be divided into its number of family members? Besides, these elements tend to relate to third variables in an opposed way; for example, whereas educational level is usually related with more liberal ideologies, income level predicts more conservative ones (^{Trautmann, van de Kuilen, & Zeckhauser, 2013}). It has also been found that the hierarchical position a person believes to have within a group can be a better predictor of several variables-like life quality and wellbeing-than objective reality itself (^{Anderson, Kraus, Galinsky, & Keltner, 2012}). These reasons have led SSC measurement to acquire a special relevance not only in social psychology studies, but also in other disciplines, such as the medical science (i.e., ^{Adler et al., 2002}; ^{Adler, Epel, Castellazzo, & Ickovics 2000}).

The most common form of SSC measurement is the so-called MacArthur ladder (Scale of Subjective Social Status), developed at the University of California (adult version: ^{Adler et al., 2000}; teenage version: ^{Goodman et al., 2001}). By using this scale, participants are shown a ten-rung ladder. Then they are asked to imagine a ladder, which represents the society they live in-or their school, neighborhood, organization they work for. Participants need to locate themselves in one of the ten rungs, according to how they perceive their social status. The higher up in the ladder, the higher social status and more prestige the participant owns. This measurement has proven useful to find out, for example, that SSC relates positively to health (^{Adler et al., 2000}) or to sense of control (^{Kraus et al., 2009}). It is also related negatively to altruistic behavior (^{Dubois, Rucker, & Galinsky, 2015}; ^{Piff et al., 2012}) and empathy (^{Kraus et al., 2010}).

Nonetheless, most of these studies, due to their methodological design, did not allow establishing a cause-effect relationship between the variables. This led ^{Kraus et al. (2010}) to suggest an experimental manipulation of the subjective social class. Hence, they developed a paradigm in which they ask the participants to compare themselves with someone in a high or low part of the social hierarchy. Comparing with others activates cognitive mechanisms of assessment about what has been accomplished or will be possible to achieve (^{Diener, Fujita, Tay, & Biswas-Diener, 2011}; ^{Shmotkin, 1991}; ^{Smith, Diener, & Wedell, 1989}), which changes the perception that people have about their social position and has similar consequences to those obtained when SSC is measured as individual difference (i.e., on empathy, ^{Kraus et al., 2010}).

In this article we examine the joint effect of both variables: subjective social class measured as individual difference (SSC-ID) and experimentally manipulated (SSC-EM). More precisely, we believe that people with a high SSC-ID will be less liable to an experimental manipulation of their SSC. This prediction is congruent with some studies about social power. In these studies, conducted by ^{Johnson and Lammers (2012}), the most powerful people-those with a higher degree of influence and control over others-were found to be less susceptible to social comparison; to put it another way, their self-concept was not usually affected by “upwards” or “downwards” social comparison. This is due to the fact that the most privileged people-those who are powerful or have a high SSC-ID- are very often more self-centered in their own point of view, and they pay low attention to social information (e.g., ^{Galinsky, Magee, Inesi, & Gruenfeld, 2006}). On the contrary, the least privileged people are less prone to using the information about their social context when it comes to defining their self-concept; that is, they are more susceptible to social comparison (^{Johnson & Lammers, 2012}).

Social class and self-efficacy. ^{Kraus et al. (2009}) have shown that people with a high SSC usually show a higher sense of control than those with low SSC. For instance, in the Studies 1 and 2, conducted with university students in the USA, the researchers measured the SSC-ID using the MacArthur ladder (^{Adler, 2000}), and they found that this measure is positively related to the sense of control measured using the ^{Lachman and Weaver Scale (1998}). In Study 1 participants were shown a graphic revealing an increasing trend towards economic inequality at the present time; in Study 2, the graphic showed a gradual decrease of economic disparity. In both cases, the SSC-ID was found to be negatively related to the contextual explanations of economic inequality; for example, by explaining inequality-both its increase and decrease-through factors such as differences in educational opportunities, instead of explaining them through differences in capabilities and endeavor. Importantly, the sense of control mediated the following results: the negative relation between the SSC-ID and contextual explanations was explained, at least partially, by means of differences in perceived sense of control. In other words, the SSC-ID increased the sense of control, which in turn decreased the contextual explanations. This suggests that the more common trend of those with a low SSC-ID to explain situations based on contextual factors can be due, at least partially, to them having a sense of control in a lower degree.

In this research, unlike in the studies led by ^{Kraus et al. (2009}), we will ascertain the effect of the SSC on the generalized self-efficacy (^{Jerusalem & Schwarzer, 1992}; ^{Schwarzer, 1993}). Whereas the sense of control relates to the success prospects or to the individual’s belief that they can intentionally produce the desired results and avoid the undesired (^{Skinner, 1996}), the self-efficacy relates to the belief of being able to produce a controlled response, initially linked to the relation between the agent and the type of response (^{Bandura, 1977}).

According to ^{Bandura (1987}, ¹⁹⁸⁹, ¹⁹⁹⁴), individuals’ actions, motivation levels and emotional states are driven by their beliefs about their own capabilities to reach a target, namely, their self-efficacy. These beliefs may be influenced by either direct or vicarious experiences, as well as by social persuasion, and psychological and emotional states. Self-efficacy is considered to be a mediator variable that influences conduct, endeavor and result, regardless the individual’s capabilities (^{Mansilla, 2003}). It also directly affects the estimates about individuals’ current situation and future ambitions, which anticipates their choices and performance in studies, work and personal life by becoming a self-fulfilling prophecy (^{Han et al., 2015}).

Given that ^{Kraus et al. (2009}) have found that social class -measured both as an individual difference or experimentally manipulated- positively affects the sense of control, and given that self-efficacy is a very similar variable to the sense of control (^{Skinner, 1996}), SSC is thought to be going to positively relate to generalized self-efficacy. In short, we hypothesize: First: the higher the subjective social class (SSC-ID), the greater the general self-efficacy level (H₁). Second: those participants assigned to a high SSC-EM condition will show a greater generalized self-efficacy than those assigned to a low SSC-EM condition (H₂). Third: following the results obtained by ^{Johnson and Lammers (2012}), who found that more privileged people are usually less susceptible to social comparison -and that the way to manipulate SSC-EM is using a social comparison paradigm-, we anticipated that people with a low SSC-ID, but not those with a high SSC-ID, will be influenced by the experimental manipulation of the social class (H₃).

Furthermore, we will also use in this study a qualitative methodology so as to explore the way in which people build up the meaning of belonging to a higher or lower social class. Hence, we will analyze the responses given by the participants when asked to describe the characteristics of people who were higher up or lower down in the social ladder.

Results

Quantitative results. Twelve participants were left out of the analysis because one of them did not answer any of the items in the General Self-Efficacy Scale, and eleven did not properly follow the instructions of the experimental manipulation. In order to make the interpretation of results easier, the predictor variables were focused on the measures of each one of them (mean centering), and the dependent variable kept the original response scale. The final sample consisted of 380 participants, from whom 190 (50%) were assigned a High SSCM (G₁) and 190 (50%) were assigned a Low SSCM (G₂).

Hypothesis testing. A positive correlation was found between SSC and generalized self-efficacy (r ₍₃₈₀₎ = .23, p < .001). This way, according to our predictions (H₁), we were able to find that the higher the subjective social class (SSC-ID) the participants showed, the greater general self-efficacy they reported.

Additionally, in order to examine the effect of the manipulated subjective social class on self-efficacy (H₂) we performed a t test for independent samples, and we found that the subjects who were assigned a low SSC-EM reported a higher general self-efficacy (M = 34.02, SD = 4.12) than those who were assigned a high SSC-EM (M = 33.15, SD = 4.79), t ₍₃₇₈₎ = 1.89, p = .059, CI 95% (-.03, 1.77), d = .19), though the difference was not statistically significant. Likewise, the mean difference was found to be contrary to our hypothesis, so the data gathered in this study do not allow to confirm Hypothesis 2: a main effect of SSC-EM on self-efficacy was not found. We will examine below if there is an effect of the interaction between SSC-EM and SSC-ID on this latter variable.

Moderation effect. In order to contrast H₃, i.e., proving that the SSC-ID (M) moderates the effect of SSC-EM (X) on the general self-efficacy (Y), we carried out a moderation analysis with 1000 bootstraps, using Model 1 from the macro PROCESS implemented for SPSS (^{Hayes, 2013}).

As shown in Table 1, it was found that the interaction effect between the subjective social class (SSC-ID) and the manipulated social class (SSC-EM) on the general self-efficacy was marginally significant.

Table 1: Results of regression analysis with interaction term between SSC and the experimental condition on the Generalized Self-Efficacy. 

As it can be observed in Figure 1, and in line with H₃, we found that people with a high SSC-ID (1 SD above the mean) presented no significant changes in their self-efficacy as a function of the SSC-EM (b = -.12, t ₍₃₇₃₎ = .20, p = .87, CI 95% (-1.4, 1.14)). The participants with a low SSC-ID (1 SD below the mean), on the contrary, were susceptible to the SSC-EM (b = -1.70, t ₍₃₇₃₎ = -2.67, p < .01, CI 95% (-2.93, -.44)). In these participants the SSC-EM affected their general self-efficacy, being this lower in the case of high SSC-EM than in the case of low SSC-EM.

Figure 1. Subjective social class as individual measure moderates the subjective social class experimentally manipulated.

Qualitative Results. In a complimentary and exploratory way, we analyzed the characteristics which were assigned to people with a lower (G₁) or higher (G₂) social class reducing them in categories and dimensions according to the main emerging issues. Such characteristics were coded according to three analysis categories:

1. Type of characteristics. This regards to the nature of the characteristics that people attribute to the other group (low or high class), according to the degree of control that they are acknowledged, which can be recognized in the individual itself, in its social context, or in both. There are three types: (a) situational: all the characteristics related to material life’s conditions, people’s resources or external elements, including access to services and spaces (i.e., money, opportunities, wealth, education); (b) dispositional: all the characteristics related to the traits which define a person, such as their way of thinking, feeling, etc., including traits related to their capability to exert influence or power; (c) situational and dispositional: when both characteristic types are acknowledged as for people and conditions, resulting in a combination of factors.

2. Valence of the characteristics. This relates to the value judgements that the already mentioned characteristics have associated. They could be: (a) neutral: aspects which do not directly contribute nor affect the human development of people neither in a positive nor negative way; (b) negative: adverse or problematic factors which affect the human development and limit the possibilities of the people to develop their potential, i.e., characteristics which are not approved by society and are considered to be negative, such as envy, arrogance, etc.; (c) positive: aspects which favor the human development and the deployment of people’s capabilities (i.e., access to education, economic and/or material resources), or characteristics which are socially cherished and considered to be positive such as solidarity, happiness, etc.; (d) complementary: positive and negative aspects at the same time, so that a certain aspect can compensate for the lack of another, e.g. “poor but happy” or “rich but stressed”.

3. Theme of the characteristics. This relates to the identification of contents of specific themes in the characteristics stated by the participants (i.e., “not having a car”, “being a person of great faith”, “having low self-esteem”, etc.). These contents were grouped by common themes according to the categories emerged during the analysis (i.e., money, luxuries, travels, leadership, sociability, etc.).

Three expert judges led the coding process. The exercise was conducted jointly, deliberating about the statements in which there was no general consensus in order to ensure a unified criteria as for the assignment of the correspondent categories. The results of this coding are presented in the following contingency table (Table 2).

Table 2: Contingency table between the Type and the Valence of characteristics for each experimental condition. 

Relation between the SSC-EM and the type of characteristics. There was no association found between the SSC-EM and the characteristic type performed on the other groups χ ² (2, N = 372) = .27, p = .87, that is, there were no significant differences between the characteristics assigned to people who belonged to a higher or lower class. Therefore, in both conditions the participants stated characteristics mostly related to situational aspects (59.1%), and to a smaller extent related to dispositional aspects (19.4%) or to the combination of both factors (21.5%).

Relation between the SSC-EM and the valence of the characteristics. We found an association between the manipulated subjective social class and the type of characteristics assigned to the other groups χ ² (3, N = 372) = 240.32, p < .001. In this way, the participants with a high SSC-EM reported some characteristics of the people of a lower status in a more negative (78.8%) or complementary way (15.1%), than in a positive way (7%); whereas the contrary trend was observed when it came to describing people of a higher social status, mostly reporting positive (81.2%) or complementary characteristics (14%) rather than negative (4.2%).

Relation between the type of characteristics and their valence. We found a significant link between the type of characteristics and the valence associated to each of them χ ² (6, N = 372) = 102.25, p < .001. However, this association represents differentiated trends for each group. For the high SSC-EM group, which had to describe the characteristics of people from a lower social status, it was found that they described the situational characteristics as mainly negative (75.2%) and no one reported any positive aspect. Among the dispositional characteristics described by this group, there were a similar number of positive and negative valences (35.1%) or they reported a combination of both types (24.3%). For the low SSC-EM group, which had to describe the characteristics of people from a higher social status, it was found that the situational characteristics described were mainly positive (97.2%) and none of them was described as negative. As for the dispositional characteristics, their valence was positive (57.1%), followed by complementary valence (22.9%), and, to a lesser degree, by negative valence (14.3%).

Thematic characteristics. The analysis of the specific content of the characteristics described by each experimental group will be presented now.

High SSC-EM (G ₁ ) and characteristics of participants of low status. The Table 3 lists the answers of the participants who were given instructions to report the characteristics of those located below them in the scale. Most of them perceived the others as people with some sort of education lack, associated to studying in public schools, limited to basic schooling and quitting their studies.

The economic deprivation is another condition that describes people who are perceived as belonging to a lower status. They were thought to lack or have low levels of basic satisfiers such as food, housing and clothing, as well as water and electricity services. Even though these responses imply the concept of “poverty”, the mere word was only mentioned by two of the participants.

The main situational characteristic given to explain the lower position of people was the lack of opportunities, and though this may indicate an awareness of the external obstacles that people in an unfavorable situation have to face, in the second position we could find the lack of ambition, which is related to conformism and the lack of big life goals.

Concerning the world of work, those lower down in the scale are perceived as having a bad job which demands a great deal of effort and commitment, but that is badly paid. No specific occupation was listed as being related to this job insecurity, except for one participant who reported an association between this and working as a construction laborer. Among the indicators of low status, not having a car, using public transport and not having resources to travel, stood out from the participants’ responses.

In certain aspects, opposed perspectives can be observed: there is an equal number of participants who think that those lower down in the scale are happier --because they live with lower pressures or because they give more value to the things they own-- and of those who think they are unhappier; some would describe them as honest, modest and hardworking, while others believe they are untrustworthy, conformist and unsociable; some mentioned they belong to a large and united family, while others highlighted they have family problems or belong to a small family, or have no close relatives at all.

Table 3: Themes of the characteristics assigned to people in a lower socioeconomic status than G_1. 

Low SSC-EM (G ₂ ) and characteristics of participants of high status. As it can be ascertained in Table 4, the participants who compared themselves with those of a higher social status, perceive them as wealthy individuals, people with material resources, luxuries and able to travel frequently. Apart from having abundant material assets and being thought to lead a comfortable life, they are identified as businesspeople, company owners or top managers within an organization.

Concerning their psychological traits, it is important to remark that they were described in more positive than negative terms. They are thought to be perseverant, hardworking and committed, and to have great ambitions and leadership skills. Among the negative characteristics reported, the most significant ones were being arrogant, boastful and spendthrift.

Among the possible explanations the participants offer about why these people belong to a higher social level, the most significant ones are related to their access to a good education (prestigious schools) and, in general, the better opportunities they can access to. As a matter of fact, these two factors were the most mentioned in the downward comparison: bad-quality education and lack of opportunities.

Table 4: Themes of the characteristics assigned to people in a higher status than G_2. 

Discussion

The results of this study provide confirmation about the positive effect of the SSC on the degree of general self-efficacy (^{Henry 2001}; ^{Jerusalem & Schwarzer, 1992}; ^{Kraus et al., 2009}; ^{Schwarzer, 1993}). However, the effect was only found when the SSC was measured as an individual difference (SSC-ID), not when it was experimentally manipulated (SSC-EM). In other words, we corroborated H₁, but not H₂.

On the other hand, the results showed that, according to our prediction (H₃), the SSC-ID moderates the effect of the SSC-EM-the effect of social comparison varies according to the position that people think they hold in the social hierarchy. Hence, people with a low SSC-ID were affected by the experimental manipulation, whereas people with a high SSC-ID were not. This is congruent with the results found by ^{Johnson and Lammers (2012}), which showed that the people who are in a privileged position are usually less susceptible to manipulations in which social comparisons are used. The people who see themselves as belonging to a high class are aware of their position of “privilege” compared to the average, and therefore, regardless the social comparison they are exposed to (with people of a higher or lower status), their scores are usually significantly higher than the score of lower class people in the self-efficacy scale, and they do not show a clear variation between the experimental conditions.

Despite this, the results found in people with a low SSC-ID were contrary to what was expected. By assigning a group of people with a low SSC-EM, they increased, instead of diminishing, their self-efficacy level. This is inconsistent with previous results where it had been proved that assigning participants with an experimental condition in which they perceive themselves as belonging to an upper social class (because they compared themselves with people of a lower status) usually increases several variables related to the sense of control in their social context (i.e., ^{Kraus et al., 2012}).

These results of SSC-EM in people with a low SSC-ID could be explained by means of cultural differences. The qualitative results showed that this group used mostly positive characteristics when describing people of a higher social status, such as being perseverant, hardworking, committed, having great ambitions and leadership skills. In the Mexican context, “overcoming poverty” has a great value; it gives the individual a position of “social respect” (^{Díaz-Guerrero, 2007}, p. 142), within an imaginary dignity in which the myth of social success at all costs is very present (^{Cyrulnik, 2003}).

Accordingly, this group of people might be motivated to deploy some sort of compensatory control strategy (see ^{Kay, Whitson, Gaucher, & Galinsky, 2009}), in which a low perception of control over the social context is compensated by a higher perception of self-control (in this case: self-efficacy). In this way, in the case of the low class where the appearance of a relative lack of resources (less money, fewer opportunities, poorer conditions) was made clear in comparison with other groups, these people could experience a lower sense of control over their social context, and as a consequence, they might try to retrieve the sense of control by means of increasing their self-efficacy perception.

Similar results have been obtained from this standpoint, showing that when facing a decrease in the sense of control on the social context, there is a tendency to use psychological mechanisms to compensate for this lack of control by holding on to certain ideologies which praise personal agency (^{Kay, & Eibach, 2013}), such as meritocracy and the idea of “working hard” to achieve success (^{Goode, & Keefer, 2016}), or in some cases even by rationalizing and justifying the status quo and the inequalities it implies (^{Knight, Tobin, & Hornsey, 2014}). The fact that the type of characteristics attributed to people of a higher or lower class is mainly focused on situational aspects could be an indicator of this phenomenon. Nonetheless, further researches will have to corroborate whether the mechanisms of compensating control could explain the increase of the self-efficacy perception.

We would like to remark that there is an important difference between our study and that by ^{Kraus et al. (2009}), given that we did not include a manipulation check. This is because the measure needed to assert that the manipulation has been successful is the same used to measure the SSC-ID. Consequently, provided that the participants reported their SSC-ID before the manipulation, we were not able to use that measure again. However, we believe that the simple fact that the participants actually compared themselves with people higher up or lower down in the scale than themselves (as it is shown successfully in the qualitative study) is enough evidence that people were subjectively locating themselves in a subjective social class; one that was higher or lower than theirs at the moment of performing the task.

It must be said that, providing that we did not perform a manipulation check, we have to be careful about arguing that we indeed manipulated the subjective social class; we can only be certain of having manipulated social comparison (upwards vs. downwards). It will be important for future research to examine for greater evidence of the success of the subjective social class manipulation. Measuring the SSC-ID in a different timing to that of the manipulation could be a possibility. This would allow using a manipulation check after the manipulation, which would increase the reliance on the experimental results.

On the other hand, the qualitative results also brought some important findings to light. For instance, we might wonder what is behind the non-appearance of the word “poor”, or the word “poverty” only appearing twice, even though the perspective about the people of low class was one of great restrictions and deprivation. A possible answer to this might be social desirability (^{Edwards, 1957}, ¹⁹⁹⁰) and the idea that the participants could consider the use of such words as “distasteful”, given their negative connotation; or it could even be due to the word being strongly linked to political propaganda or religious discourse. Another possibility is the omission of the word because it implies a dichotomy between “rich and poor” and because the participants do not feel that they belong in the category of “the rich”--or they feel uncomfortable being categorized like that.

Regarding the personal traits of those in an upper level, the main differences perceived are related to perseverance and ambition. Such traits are easily controllable, at least when compared to others like talent or intelligence. In other words, high-class people are acknowledged by things that are thought to be within their control, i.e., requiring tenacity, effort and, in theory, anyone could develop. According to this, the gap between low-class and high-class people would be mostly due to motivation and self-efficacy. Some could argue that the participants who were assigned a low SSC-EM, when it comes to describing upper-class people as individuals who have believed in themselves and hoped to be where they are at, think they are capable of being on their way to achieve it by reporting a greater self-efficacy; and that comparing themselves to others could imply an assimilation effect by means of changing their self-concept towards the subject in comparison (^{Lockwood & Kunda, 1997}).

There is also consensus about the fact that upper-class people are businesspeople, company owners, and top managers of organizations. On the contrary, lower-class people are not linked to any kind of job or social role; they are just described as working a lot and being badly paid. It is possible that the participants have a clearer idea of the social roles played by upper-class people, because they hold a position that is well desired and spread on the media, in comparison to lower-class people, who are described by some participants as serious, reserved, and about whom there is little or no knowledge nor contact, or intention of having so. It is possible that, for this same reason, upper-class people are described using more positive than negative terms, whereas lower-class people are described both positively and negatively, i.e., there is not a well-defined generalized picture of them.

On the whole, the perspective offered in both cases seems to go to the ends of the social scale: those who are below think those who are above are indeed very high above them (businesspeople or company owners who live a great life), and those who are high up in the scale think those below them are really down below (they have no house, food or clothes). Not falling into a stereotyped vision about the social groups seems to be difficult. In the case of high-class people, it is especially remarkable that they do not mention independent professionals such as lawyers, doctors, engineers, etc.; as if being a top executive or company manager were the only ways to get a high social position. In the case of low-class people, few responses imply a humble life led in combination to lacking the most indispensable things.

Conclusions

The SSC is an important factor in the perception of self-efficacy and it moderates the potential effects of social comparison. The SSC-ID showed a direct link with self-efficacy: the participants who reported a high SSC-ID showed a greater perception of self-efficacy, and they were less prone to be influenced by social comparison; whereas the participants with a low SSC-ID reported a lower self-efficacy and were susceptible to social comparison. On the other hand, we did not replicate the main effect of SSC-EM on self-efficacy.

In the future, it would be important to analyze whether, on the one hand, in some cases the self-efficacy may be more influenced by some cultural characteristics (i.e., meritocratic and individualistic contexts) which would need to be examined in future research.