Reasons for disliking a classmate: A comparative study with Spanish and Portuguese students

Socioemotional competence and peers ’ social experiences have a mutually affecting relationship. Exploring students ’ reasons for disliking a classmate within the classroom context has theoretical importance for peer interaction literature and practical importance for assisting teachers in building positive social environments. The goal of the study was to evaluate a system of categorization of reasons for disliking a classmate. For that, and in line with grounded theory, an enlarged heterogeneous sample of fourth and sixth grade Spanish and Portuguese students was used. Data collection was done through a sociometric task followed by a semi-structured interview (Spanish students) or by completing a questionnaire (Portuguese students). 1596 reasons provided by the Spanish students and 638 provide by the Portuguese students were analysed following the grounded theory procedure. Results showed that students tended to mention the same type of reasons; differences within groups relate to reasons ’ salience. Implications are drawn for teaching practice and for exploring the effect of culture and school organization on students ’ appreciation of each other.


Introduction
Socioemotional competence is fundamental for establishing and maintaining positive peer interactions and relationships. Peers related social experiences have as well an influence on emotional and interpersonal development (Denham, Warren, von Salisch, et al., 2011). By school age, most students are aware of a large array of emotions, and of emotional, cognitive, and behavioural regulation strategies and are able to effectively send emotional complex messages, by taking into consideration the characteristics of the interaction and of the people involved in it (Denham, 2005;Denham et al., 2011). In addition, emotions' understanding also becomes more complex as children age (Cuisinier and Pons, 2011). Five to seven-year-olds are able to understand that the experience and expression of emotions are affected by internal states and that it is possible to simulate and to dissimulate emotions. From eight years onwards, children are able to understand the mix and ambivalent nature of emotions, the effect that moral rules have on emotions and the possibility of using mental strategies for controlling emotional experiences.
Developmental changes in experiencing, expressing and understanding emotions have an effect on peers related social experiences. These experiences work as a context for children to learn normative rules of expressing emotions, and to develop deeper understanding of emotions and their role in peers' interactions and relationships (Denham et al., 2011). Thus, emotional competence and interpersonal competence have a complex mutual relationship. Students at different ages experience different emotions regarding their peers, as well as understand these emotional experiences differently, which will affect peers' interactions and relationships. The nature of peers social experiences will facilitate socioemotional and interpersonal competence.
Due to the developmental importance of peers related social experiences for socioemotional and interpersonal competence, many studies have been exploring peer interactions and relationships from different perspectives. One important focus is peer rejection, which is a group construct representing how much an individual is disliked by the peer group (Bagwell, 2004;García Bacete et al., 2014). Peer rejection is a stressful life event (Dodge, Lansford, Burks, et al., 2003) that causes immediate negative emotions, such as anxiety, sadness, and anger (Nesdale, 2008;Sandstrom and Zakriski, 2004). Importantly, rejection might also stimulate certain peer behaviours that inflict social and emotional distress on the rejected child (Ladd, 2005). Indeed, peers might develop public negative forms of behaviour towards the rejected child, such as exclusion from interactions and play, disapproval, and verbal and physical aggression (Lev-Wiesel et al., 2013;Sandstrom and Zakriski, 2004). Depending on the dispositions of the rejected student as well as on his/her interpretation of the behaviour displayed by the peers, the treatment by the peer group will have far more consequences on children's school adjustment (Buhs, 2005). All of these consequences make peer rejection an important topic of study, in particular for substantiating interventions that might help rejected children fit in.
While all of these empirical results make the phenomenon of peer rejection of most theoretical relevance, it is our claim that exploring peer disliking of each other is as well of practical and theoretical importance (Sureda, Garcia Bacete, & Monjas, 2009). Unlike peer rejection, disliking a classmate is not a property on one sole child, it is not a group construct. The experience of disliking a peer is part of the social experience of any child (García Bacete, Carrero, Marande, & Ochoa, 2017). Peer disliking is a negative peer evaluation eliciting negative emotions (García Bacete, Marande, & Mikami, 2019)), that might have an important effect on the attitudes and on the behaviours towards the peers (Hennessy et al., 2007). Thus, it is our claim that knowledge of students' reasons for disliking a classmate across different school grades and contexts will provide significant information for understanding peers related social experiences. This body of knowledge will also be useful for teachers to assist students in developing more socially adequate ways of dealing with those classmates they dislike, and in building positive social environments promotive of learning and interpersonal and emotional development.

Reasons for disliking a classmate
Some theories suggest that children like peers who are similar to them, regarding age, gender, ethnicity, or other characteristics, as similarity facilitates self-validation and self-confidence, and reduces conflict with peers (Bennet and Sani, 2008;Gifford-Smith and Brownell, 2003;Hartup, 2009;Nesdale, 2008). Other theories mention that children seek contact with peers depending on their evaluation of how much that interaction will afford the possibility to satisfy their basic needs, such as companionship, trust, autonomy, and sense of belonging and of self-worth (Asher and McDonald, 2009). While these theories explain the process of disliking or liking someone, they do not explore children's perceptions regarding peers and peers' behaviours. Nevertheless, these perceptions affect how children evaluate each other, justify what behaviours are more acceptable and valued, as well as the expression of certain behaviours in relation to target-peers (Goldstein and Tisak, 2010;Hennessy et al., 2007;Johnson, Heath, Bailey, et al., 2013).
For instance, the study of Coplan, Zheng, Weeks and Chen (2012) showed that kindergarten and early elementary school children tended to make different appraisals and causal attributions regarding the behaviour of shy, unsociable and socially competent hypothetical children. They were also more eager to be friends with the shy hypothetical child than with the unsociable hypothetical child. Focused on children with mental health problems, O'Driscoll et al. (2015) also found that attributions concerning the behaviour of hypothetical peers with ADHA or depression influenced how fair adolescents think it is to exclude a peer with mental health problems. Similarly, Dolphin and Hennessy (2014) found that whenever adolescents made external attributions regarding the cause of depression of a hypothetical classmate, they did not infer responsibility on the behaviour displayed by the peer, they felt sympathy and pity towards the classmate and reported a higher probability of socially accepting him/her.
Most of these studies examined clinical samples or problematic forms of behaviour and used vignette or videos of hypothetical peers for eliciting a response from the participants. However, responses regarding a hypothetical peer might differ from responses in real life situations. Furthermore, any student in a class, and not only those students displaying problematic behaviour or specific conditions, might be disliked by his/her classmates. Thus, these studies lack ecological validity (Dolphin and Hennessy, 2014). In order to overcome these limitations, the current study aims to examine students' reasons for disliking a classmate within the context of regular classroom life and processes.

The current study
This study aims to evaluate a system of categorization regarding reasons for disliking a classmate, which was previously created by some of the authors (García Bacete, Carrero, Marande, & Ochoa, 2017)). Using the constant comparative method (Carrero et al., 2012;Corbin andStrauss, 2007, García Bacete, Carrero, Marande, &Ochoa, 2017) were able to differentiate reasons for disliking a classmate provided by 1st and 2nd grade students.
Within the system of categorization, reasons for disliking a classmate were grouped in three supra-categories: Behaviour, Identity and Indifference, and in categories and sub-categories within these supra-categories. Behaviour supra-category describes reasons for disliking a classmate grounded on the target-peer's behaviour (e.g., on what she/he does, says, or tries to do) and it emerged as an important collection of reasons for disliking a classmate. Indeed, peers tended to dislike those who do not comply with the norms, cause annoyance, dominate or show aggressive behaviour (García Bacete, Carrero, Marande, & Ochoa, 2017). This supra-category adjusts to extant reports on the association of problematic behaviour and peer rejection (e.g., Gifford-Smith and Brownell, 2003;Ladd, 2005;Rubin, Bukowski, and Parker, 2006). Identity supra-category describes reasons grounded on attraction processes. Children tended to dislike those classmates with whom they do not share joint interests and activities, towards whom they hold an overall negative appreciation, or who belong to certain social groups with which they do not identify (García Bacete, Carrero, Marande, Bacete et al. International Journal of Educational Research xxx (xxxx) xxx & Ochoa, 2017). This supra-category adjusts to theories of personal attraction that state that people tend to like those who are similar to them as similarity facilitates self-validation and self-confidence (Gifford-Smith and Browell, 2003;Hartup, 2009) and that belonging to and identifying with a certain group help children develop a more positive image of who they are (Bennett and Sani, 2008;Nesdale, 2008). Finally, Indifference supra-category describes reasons grounded on a lack of familiarity, lack of personal knowledge, reflecting a low interest in getting to know or in initiating contact with the target-peer. This supra-category adjusts to existing reports on peer interaction showing that familiarity is an important dimension for peer interaction and relationships (Farmer et al., 2007;Gifford-Smith and Brownell, 2003). Thus, the final system of categorization showed adjustment to the data, and to theoretical knowledge regarding peer interactions and relationships, which are important criteria for evaluating the final categorization (Carrero et al., 2012;Corbin and Strauss, 2007). However, considering that age is a key variable for conceptualizing peer social experiences (Bukowski, Cillessen and Velásquez, 2012), it is as well important to explore how motives for disliking a classmate change at different ages and whether the system of categorization is useful for using with older students.
Accordingly, another goal of the study was to further evaluate the system of categorization, by extending the data collection to new age and cultural groups (i.e., 4th and 6th grade Spanish and Portuguese students). Aligned with principles of grounded theory (Carrero et al., 2012;Corbin and Strauss, 2007), the new groups worked as comparative groups aimed at maximizing the variance of data in order to saturate the initial categories. The categorization is considered robust if new information no longer enlarges the properties of the original categories or supports the emergence of new categories (Carrero et al., 2012).
Considering that the supra-categories are clearly supported by data and theoretical knowledge on peer rejection, peer interaction and personal attraction, any deep changes in the properties of the supra-categories were not expected. However, some changes were expected in the categories and sub-categories within the supra-categories. As children grow up, their emotional understanding of peer interactions and of adequate social expression of emotions change (Cuisinier and Pons, 2011;Denham, 2005;Denham et al., 2011), as well as their social goals and motives, and the repertoire of social behaviours (Gifford-Smith and Brownell, 2003;Rubin et al., 2006). Thus, we expected to find differences regarding categories of behaviour for sustaining classmate dislike. In particular, we expected that younger students would report more physical aggressive behaviours while older students more intimidation and verbal aggressive behaviours, since young students tend to resort to physical aggressive behaviour while older students resort more often to other forms of aggression (Gifford-Smith and Brownell, 2003). In addition, given that older students engage more frequently in activities such as conversation and hanging out (rather than play), and have more sophisticated notions of friendship (Ladd, 2005;Rubin, 2002) and a more complex social understanding (Selman, 2003), we expected their reasons for disliking a classmate would be more guided by unshared interests and activities (categories within supra-category Identity) than younger students' reasons.
As this is an exploration study and no previous data exists regarding Portuguese students, no specific hypotheses were derived regarding differences between Spanish and Portuguese students. Nevertheless, cultural characteristics act as distal environments affecting students' development and behaviour (Chen and Rubin, 2011). In addition, the school system organization and classroom composition and social dynamics have been considered important contexts affecting the emergence of certain types of peer's interactions and relationships as well (Chang, 2004;Mikami et al., 2010;Wentzel, 2009). Thus, some differences regarding both cultural groups were expected.

Spanish sample
The sample used in this study was a convenience sample mainly due to accessibilitythe schools involved were situated nearby the universities that collaborated with the study. Participants were 1702 students enrolled in 13 public Primary Education schools of four Spanish cities (Castellón, Sevilla, Palma de Mallorca and Valladolid). Students belonged to 40 first and second grade classrooms (younger students' sample), 10 fourth grade classrooms and 26 sixth grade classrooms (older students' sample). Their ages ranged between 6 and 12 years-old, according to the structure established in the Spanish education system. The study was authorized by the ethics board of the University and the participating school's boards. Participation in the study was voluntary and families provided written informed consent. From this sample, 247 students did not nominate any classmate or did not express any motive for disliking a classmate. From the total of 1702 participants, 4159 reasons provided by 1455 students for disliking a classmate were analysed (Table 1).

Portuguese sample
The sample was a convenient sample of 504 students (52.2 % male), enrolled in 17 public schools of Lisbon's metropolitan area and surrounding districts. Fourth graders (56 % of the sample, 15 classrooms) ranged from 8 to 10 years-old and sixth graders (44 %, 11 classrooms) ranged from 11 to 12 years-old (Table 1). Participation in the study was voluntary. Participant students had parental consent to join the study. In addition, the study was authorized by the Portuguese Data Protection Commission and by the Ministry of Education. Twenty-four students did not nominate any classmate and 78 didn't provide any reason for negatively nominating a classmate. In total, 638 reasons for disliking a classmate presented by 402 students were analysed (Table 1). Most of the students provided two motives to justify their nomination (59 %) while 39 % of the students provided one reason. Bacete et al. International Journal of Educational Research xxx (xxxx) xxx

Data collection in Spain
The data for the present study was collected through individual semi-structured interviews in which 4th and 6th grade boys and girls answered questions extracted from the sociometric questionnaire developed by García Bacete, Carrero, Marande, & Ochoa, 2017: 1) "From all the girls and boys in your class, whom do you like least?", and 2) "Why do you dislike (Name of each of the nominees in question 1)"? From these interviews, we compiled a list with 1882 reasons expressed by students to explain their rejection towards their classmates, using grounded theory to analyse the data (Carrero et al., 2012). In order to avoid overrepresentation of a few over-expansive students, a maximum of 5 reasons per student-nominator was considered to the quantitative analysis. Thus, the total number of reasons analysed were 1596. The same procedure was followed in the previous study with 1 st and 2nd grade Spanish students (García Bacete, Carrero, Marande, & Ochoa, 2017).

Data collection in Portugal
Data were collected by means of a paper sociometric questionnaire, using a peer nomination task. Students were asked to write down the name of the classmates with whom they liked to play or hang out with during recess and the ones with whom they did not like to play or hang out with. No limitation in the number of nominations was used. In addition, students were asked to present reasons to justify their choices. For the current study, analysis focused solely on the 638 reasons related to not liking to play or to hang out with a classmate.

Analytic strategy
To fully understand the reasons why children dislikes some peers, we used Grounded Theory, a methodology developed by Glaser and Strauss (1967), to analyse students' answers. This methodology has shown to be able to generate conceptual categories from data that explain the processes that occur in complex social situations (Carrero et al., 2012).
For analysing the reasons provided by the students, we used the categorization created by García Bacete, Carrero, Marande, & Ochoa, 2017 as a frame of reference. This categorization is formed by three supra-categories (Behaviour, Identity and Indifference), each one constituted by categories and sub-categories (Table 2). This categorization has been recently updated (García Bacete, Marande, & Mikami, 2019). In particular, the category Problematic Social and School Behaviours was further split in two sub-categories: Antinormative Behaviours and Incompetent Behaviours, and answers such as I don't know were no longer lost data, but rather were included in the supra-category Indifference, in the category Do not know -Do not reply.
In line with the grounded theory model (Carrero et al., 2012;Corbin and Strauss, 2007), first all the reasons of 1st and 2nd grade students were re-read and incidents of 4th and 6th grade students were compared to incidents of 1st and 2nd grade students. New incidents were compared to supra-categories, to the categories and to the sub-categories considering their properties. Based on described properties (García Bacete, Carrero, Marande, & Ochoa, 2017, García Bacete, Marande, & Mikami, 2019, decisions were made regarding where to fit the new incidents, namely whether new categories should be created in order to adjust for the new incidents or whether defined properties should be enlarged to include the new incidents. Secondly, in order to better explore the robustness of the categorization, a comparative analysis of younger and older Spanish students, as well as older Spanish and Portuguese students was performed. Chi-square tests were performed to test for significance and adjusted standardized residuals were Bacete et al. International Journal of Educational Research xxx (xxxx) xxx considered for examining which cells are contributing the most to the chi-square value and account for any significant difference between observed and expected values (Field, 2009).

Global evaluation of the categorization
All reasons mentioned by the older (Spanish and Portuguese) students fit into the same supra-categories and categories as the ones identified for the younger Spanish students (Tables 3-5).
Most of the specific reasons were the same or very similar. Nonetheless, there were some differences on how the reasons were expressed, which did not affect its content or meaning; this was particularly evident on the supra-category Identity (Table 4). Spanish students tended to use more often physical and concrete attributes as a reason for disliking a classmate, for instance, small, ugly, big ears, wears glasses. They also expressed more reasons for disliking based on the classmate's group belonging, such as being a girl, being from South America; yet no mention was made to country and nationality in the case of the Portuguese sample. In addition, the Portuguese students tended to mention more reasons for disliking a classmate related to non-physical attributes, such as "I do not identify with him", which were not as evident within the Spanish students (Table 4).
Younger Spanish students mentioned a higher proportion of reasons related to aggressive behaviour (28.8 % versus 23.6 % of the older students' answers, asr = -3.7). On the contrary, older students expressed more reasons related to dominance than their younger counterparts (respectively, 10.2 % and 3.4 %, asr = 8.9) (Table 6). Indeed, more reasons such as the following were mentioned by older students: "She always wants to boss around"; "She tells me not to play with Maria" (Table 3).
In addition, regarding aggression, younger students' most mentioned reasons for disliking a classmate were related to physical aggressiveness (21.3 % of all reasons versus 6.1 % in 4th and 6th graders (asr= -13.1), while older students mentioned a higher proportion of verbal and gestural aggressiveness (12.8 % and 4.1 % respectively, asr = 10.3), such as insulting, shouting, mocking or threatening (Table 3 and 6). Regarding the sub-category problematic social and school behaviours, younger students mentioned Bacete et al. International Journal of Educational Research xxx (xxxx) xxx Table 3 Description of the behaviour supra-category, its categories and sub-categories, its properties and incidents. more antinormative social and school behaviours than the older ones (18 % and 11.3 % respectively, asr=-5.8). For instance, "He is a liar", "He cheats", "She does not well-behave". Furthermore, they mentioned less reasons related to incompetent social and academic functioning (respectively, 4.5 % versus 9.7 %, asr = 6.6), such as being rude or not doing the homework (Table 3 and 6). Within the supra-category Identity, younger and older Spanish students differed significantly in the sub-category Preference. First and 2nd grade students tended to express more reasons for disliking the target-classmate in grounds of a positive choice for interacting with other classmates (e.g., "I play with my friends"; "He is always with David") ( Table 4) than their older counterparts (respectively, 3.7 % and 1.3 % of the total expressed reasons, asr = -4.6) ( Table 6). Bacete et al. International Journal of Educational Research xxx (xxxx) xxx Table 4 Description of the Identity supra-category, its categories and sub-categories, its properties and incidents.

Subcategories Preference Dislike
Properties Idiosyncratic nominator's preferences, based on a global affective frame, or on specific interests and activities, and expressed in an unilateral or reciprocal way Idiosyncratic nominator's dislikes, conflict, exclusion, negative reputation, based on a global affective frame, or on specific interests and activities, expressed in a unilateral or reciprocal way Dislike of other children based on their belonging to a social group or category or on their participation in activities typical of those from the same group, in the absence of other reasons more specific. These reasons actually express stereotypes and prejudices against those who are not alike, or who belong to other group    Significant differences were also found regarding the categories within Indifference. Reasons for disliking a classmate grounded on unfamiliarity were more frequent in younger students than in 4th and 6th grade students (7.2 % and 4.1 % respectively, asr = −4.1) ( Table 6). For instance, "I don't know him well", "We don't play together" (Table 5).
A closer look at the categories within Behaviour indicated that Spanish students more frequently mentioned aggressive behaviours and dominant behaviours as a reason for disliking a classmate (respectively, 23.6 % versus 14.9 %, asr = -4.6 and 10.2 % versus 2.8 %, asr = -5.7). In addition, within the category Aggressive Behaviour, Spanish students reported more often forms of verbal aggressive behaviour, as a reason for disliking a classmate than the Portuguese students (12.8 % versus 5.6 %, asr = -4.9). Additionally, Spanish and Portuguese students' reasons significantly differed regarding the supra-category Identity. Spanish students mentioned less than Portuguese students, reasons related to personal attraction processes (Personal Identity) (respectively, 13 % and 22.6 % of the total reasons expressed, asr = 5.6) and more reasons related to group belonging and stereotyping processes (Social Identity) (respectively, 4.4 % and 2.0 %, asr = -2.7) ( Table 6).
Regarding the sub-categories, within Aggressive Behaviour, Spanish students mentioned more verbal aggressiveness than Portuguese students (respectively, 12.8 % and 5.6 %, asr =-4.9). On the contrary, within the category Personal Identity, the Portuguese students expressed more reasons for disliking the classmate related to an overall negative appreciation of a classmate or a lack of common interest and activities, than their Spanish counterparts (respectively, 20.4 % and 11.7 %, asr = 5.3) (Table 6). Indeed, more reasons such as the following were mentioned by Portuguese students: "I don't like playing with him", "She is not my type", "I don't identify with him" (Table 4).adjusted standardized residuals

Discussion
The goal of the current study was to evaluate a system of categorization of reasons for disliking a classmate. Two samples of 4th and 6th grade Spanish and Portuguese students were used in order to maximize the variability and thus to saturate the initial cate Bacete et al. International Journal of Educational Research xxx (xxxx) xxx gories created using reasons given by 1st and 2nd grade Spanish students. It was expected that the previous categorization would not suffer great changes since it was supported on data and theoretical knowledge on the field of peer experiences and interpersonal relationships. Indeed, the results are aligned with the expectations as no new categories emerged from the data and properties of the supra-categories, categories and sub-categories remained broadly the same. Differences were mostly related to the expression of reasons and not with the nature of the reasons. These differences are particularly salient regarding the supra-category Identity.
In order to explore more deeply the categorization, we have conducted two analyses, one comparing younger and older Spanish students, and the other comparing Spanish and Portuguese students. Based on the literature, it was expected that younger students would express more often reasons related to physical aggressive behaviours than older students would, whilst more intimidation and verbal aggression as reasons for peer disliking were expected in the older students (Gifford-Smith and Brownell, 2003). We also expected that older students would report more frequently reasons based on personal attraction processes, peers' global appreciation, and shared interests and activities (Rubin, 2002;Rubin et al., 2015).
Consistently with initial hypotheses, 1 st and 2nd grade Spanish students did express more reasons related to physical aggressive behaviour than 4th and 6th grade students. Some studies on developmental aggression have showed that younger children have not yet developed control over how to express their emotions and regulate their behaviours when they are subjected to strong emotions (e.g., angry at hearing a negative comment, frustration at losing in a game) (Moore et al., 2019) and thus, they still may lack sophisticated ways to express these strong emotions and use physical aggressive behaviour more often than older children (Tremblay, 2000). Hence, episodes of physical aggression are more frequent in younger students and so may have a stronger impact on students' appreciation of each other. Furthermore, we observed an increase in reasons for peer disliking due to verbal aggression or dominance behaviours. Verbal aggression is a way to hurt somebody much subtler than physical aggression. However, achieving the goal of really hurting the others requires a large lexicon of vocabulary and great verbal competencies. This may be one of the reasons why 1 st and 2nd graders possibly use less verbal aggression and more the unsophisticated physical aggression. Another explanation to the increased verbal aggression as reason for disliking a peer may be related to group processes. As children grow older, peers become more influential in their social life, and belonging to and remaining stable in a peer group becomes of utmost importance to them (Nesdale and Dalton, 2011). They obey then to group processes, hierarchy, and norms of their group, which may lead to the use of verbal aggression against out-group peers, as well as dominance behaviours.
Regarding problematic social and school behaviours, older students tended to ground their reasons on the target-classmate's incompetent social and academic functioning, while younger students more often mentioned antinormative and inadequate social and academic behaviours as reasons for disliking a classmate. These differences might be explained using a developmental perspective. Peers' social experiences play an important role in students' social and emotional development. Children and adolescents seek in these experiences the possibility of establishing mutual exchanges, companionship, trust, intimacy, and sense of belonging and of self-worth (Asher and McDonald, 2009;Hartup, 2009). As students grow older, they become increasingly competent in emotional functioning (Denham et al., 2011), and their socio-cognitive abilities develop in ways that facilitate empathy, perspective taking and coordination (Selman, 2003). Thus, those students who do not function competently might be easily noticed and their incompetent behaviour may constitute a reason for their classmates not wanting to establish a relationship or interaction with them (Asher and McDonald, 2009). Contrarily, incompetent social and academic functioning might not be as important for younger students whose overriding goal is to find a peer to play with, and not so much the search for intimacy, self-disclosure, companionship, which requires more competent functioning (Denham et al., 2011;Rubin, 2002).
In what regards reasons related to the supra-category Identity, in particular, the less prevalence of reasons based on personal attraction processes provided by the older students, may suggest that the peers relationships become more stable (Rubin et al., 2015) and thus it is not necessary to highlight the differences or similitudes of activities or interests shared with peers.
Regarding the supra-category Indifference, the less reference to reasons grounded on unfamiliarity, may be explained by the fact that the older students have been together for a longer time and in the school have been given the opportunity to interact in some way and so, it would be difficult to provide justifications such as not having play or talk with before. On the other hand, the slightly increment on reasons related to Do not know/do not reply might be tentatively explained by socially shared norms that value the expression of positive emotions, but not negative ones (Fay et al., 2012). Thus, it is more acceptable not to display any reason for disliking a classmate than to identify it. In addition, not sharing any reason for disliking a classmate may prevent negative emotions to arise.
Comparing older Portuguese students with older Spanish students, we observed that the order of the predominance of the three supra-categories is maintained. In both countries, reasons based on classmate's behaviour are the most frequently mentioned, followed by reasons based on identities, and thirdly, those based on indifference. Nevertheless, regarding categories and sub-categories, several significant differences were found.
Regarding the supra-category Behaviour, we observe a similar trend in both groups in all categories and subcategories (except dominance) though with much lower percentages of aggressive behaviours in the Portuguese sample. Yet differences between both groups were found within the supra-categories Identity and Indifference. In particular, we observed a high increase in reasons related to dislike/exclusion in the Portuguese sample compared to the stability of the Spanish sample, and in Unfamiliarity compared to the decrease in these reasons provided by the Spanish students. Due to the exploratory nature of this study, no hypotheses were formulated. Nevertheless, the distinct procedures of data collection might have introduced some difference in students' answers. In addition, as a tentative explanation, these differences might as well derive from differences between both countries' classroom social dynamics and schools' overall culture and organization.
While the questionnaire in the Spanish study used a general criterion, asking students to nominate those classmates one dislikes (García Bacete & González, 2010), in the Portuguese study the question asked was: with whom students do not like to play or hang out with during recess. The specificity of the question used in the Portuguese study limits the answers to what happens during a Bacete et al. International Journal of Educational Research xxx (xxxx) xxx non-formal period, the recess. Informal activities and recess, in particular, are organized following criteria of choice and preference, of play relationships and friendship ties. The question itself, "not like to play or hang out with", makes it explicit and directs the answers towards shared interests, activities, games and friendships. Classmates express then their own personal distastes, dislikes, conflicts or exclusions, or react to those they believe the rejected child feels towards them. This may explain the strong increase in the dislike/exclusion categories compared to Spanish data (I don't like him, he is unfriendly, I don't like his friends, I don't like what he plays, I don't feel like being with him, when he sees me he goes away,….) or the neutral or absence of contacts of the category Unfamiliarity (we don't play together, I don't go with them, we are not friends, we don't talk). Moreover, Spanish students were asked to provide reasons for each one of the nominated classmates, while Portuguese students were asked why they did not like nominated classmates in general, without addressing each nominated classmate. As a result, Spanish students might have been challenged to look for more specific reasons, and thus to express more concrete reasons grounded on specific attributes and group belonging of the nominated classmates, while Portuguese students might have been conditioned to express more global reasons for disliking, reasons that would capture the attributes of several (or most of) the nominated classmates and thus more abstract attributes. In other words, in the Spanish study, an individual perspective was obtained while in the Portuguese study, the perspective was collective (a group of classmates, either small or large). When a child nominates several peers, it might be difficult to provide reasons related to social identity, because of the problem of attributing the same negative stereotypes to all nominated peers. Indirectly, these circumstances could also explain the absence of an increase in the categories Verbal Aggression and Dominance Behaviours, which can be specific reasons for disliking a specific peer but not a valid answer towards multiple disliked peers. Differences between Spanish and Portuguese students might also be due to differences in the countries' specific classroom social dynamics, and schools' overall culture and organization. Classroom social dynamics impose constrains (or offer opportunities) for students to access a set of resources (such as information regarding norms and values), through their relationships with peers and teachers (Plank, 2000;Tseng and Seidman, 2007). Within these contexts, certain attitudes in relation to classmates will be more valued, more legitimate and certain behaviours and students' attributes will be more or less salient and used for making an appreciation regarding each classmate (Chang, 2004;Farmer, Leung, Weiss, 2011;Mikami et al., 2010;Wentzel, 2009;Plank, 2000). While classroom management, organization and composition might affect classroom social dynamics (Plank, 2000, Farmer et al., 2011, some studies have also been pointing out how organizational and cultural characteristics of the school environment play an important role in shaping classrooms' social dynamics (Chen and Rubin, 2011;Pertegal et al., 2015). Thus, differences found between Spanish and Portuguese students might be as well the result of specific ways of organizing the school or the result of cultural differences regarding the expression of behaviour and its relevance for peer interaction and relationships. Further studies will be needed to explore how culture and the organization of the educational response, by means of classroom functioning and dynamics, might have an influence on evaluative moves of peers in relation to each other. In particular, an important difference that we should consider refers to the fact that 6th graders in Spain are attending their last year in elementary school, while in Portugal they attend their second year in secondary school. This makes Portuguese 6th grade students to be perceived as "older", more like teenagers, than Spanish 6th graders, who are still perceived as children. By being perceived as "older" this may increase the focus on Personal Identity aspects. As mentioned previously, older students' groups have a strong influence on their members, especially on their personal identity, which in turn has an influence on their reasons for disliking a peer. These groups have, in general, a smaller number of members and share more interests in common than groups of primary school children. Additionally, upon the administration of the questionnaire, the Spanish 6th grade students had been together for up to 6 school years while the Portuguese 6th graders might have been in the same class only since the previous school year. This fact could also help to understand the higher proportion of Unfamiliarity reasons in the Portuguese sample, compared to the Spanish one.

Conclusion
Our study was not conducted without some limitations, namely the methodological aspects that accompanied the data collection, as specified previously, such as specific questions asked to the participants and restrictions on answers. This might have influenced some of the results and the comparability between studies and such methodological aspects should be more meticulously planned in future studies. Also, the fact that the data was collected on one single moment might introduce some questions regarding the stability of the reasons provided and even the students nominated. Therefore, the collection of data on multiple times, in future studies, would likely enrich the literature on peer relationships and motives for peer disliking.
Despite these limitations, the findings of this study support the initial system of categorization, which can be considered a useful instrument for analysing peer disliking in different classroom settings. Reasons for disliking based on target-classmate's behaviours, preferences, identity of nominated and nominees, and indifference between both, continue to explain peer disliking in younger and older students and in Spanish and Portuguese students. In sum, the system of categorization is applicable to a wide range of ages, and to different cultures and countries, contributing to its consolidation as a versatile tool to understand peer disliking.
Nonetheless, it seems crucial to highlight that whatever the reasons for disliking a classmate, certain forms of behaving towards peers, such as exclusion or physical aggression, should not be considered legitimate as they can cause great harm to disliked peers (Lev-Wiesel et al., 2013;Sandstrom and Zakriski, 2004). Teachers play an important role in the development of socioemotional competencies, by promoting actively appropriate interactions between peers and facilitating the emergence of constructive and productive peer relationships (García Bacete, Marande, & Mikami, 2019;García Bacete et al., 2013;Colie, 2020;Theurel and Gentaz, 2015). In particular, Morcom (2014) showed that teaching practices with 8 to 11-year-old students, explicitly focused on values such as, mutual respect, appreciation of others, attentive listening, participation and positive learning mind set, facilitated emotional and social learning as well as peer group friendships. Therefore, consistently with this study and other studies (e.g., Killen et al., 2009;Wentzel, 2009), it is expected that knowing and understanding the reasons that make peers dislike each other will allow Bacete et al. International Journal of Educational Research xxx (xxxx) xxx teachers to make sense of classmates' interactions and relationships, which will be a good starting point for teachers to work out elaborated and sophisticated ways of interaction, promoting a positive learning environment, in particular one environment where students interact positively with peers whom they do not like, with whom do not identify with, or who display non-normative forms of behaviour. In addition, the findings highlight the importance of developing cultural comparative studies (Chen and Rubin, 2011). Such studies will add on to the literature, as they will help disentangle the effects of cultural dimensions on students' experiences with peers from the effects of classroom's organization and functioning.