1 Introduction

Over the past decade, a decrease in students’ study wellbeing has become a central concern among teachers, educational policymakers and researchers (e.g., Salmela-Aro et al., 2021; Vinter et al., 2021; Walburg, 2014; Yang & Chen, 2016). The prevalence of burnout has increased to 21% of Finnish upper secondary school students, while in turn, about one-third of them report experiencing study engagement (Finnish Institute for Health and Welfare, 2021). The school closures due to the COVID-19 pandemic further increased the challenge regarding study wellbeing (Branje & Morris, 2021; Salmela-Aro et al., 2021; von Soest et al., 2022) by eroding students’ peer relationships. The reduced study wellbeing has severe short and long-term costs for the students and society, including increased risk of developing depression and other mental health problems, learning loss and even dropping out of school (e.g., Bask & Salmela-Aro, 2013; Fiorilli et al., 2017; Salmela-Aro et al., 2009; Symonds et al., 2016).

The role of peer relationships is crucial for developing and sustaining study wellbeing (Mendoza & King, 2020; Rautanen et al., 2021, 2022; Rubin et al., 2008; Ryan, 2001). For instance, peer support can reduce the risk of study burnout by helping students to overcome study-related challenges (Ulmanen et al., 2022). Peer relationships are also important for balanced psychological development, particularly in early adolencence. However, the quality of peer relationships cannot be taken for granted. They require that students are able to recognise and understand each other’s emotions and tune into them: that is, empathy skills (e.g., Eisenberg et al., 1991; Feshbach & Roe, 1968). Empathy skills are a key for providing well-fitting social support in studying for the peers and in creating a sense of relatedness in a peer group. However, it can be assumed that there are variations between students in their empathy skills (see van Noorden et al., 2015; Tikkanen et al., 2022), their ability to provide support for each other (Ulmanen et al., 2022), the sense of relatedness (Furrer & Skinner, 2003), as well combinations of these variations. The variations are likely to result in differences in experiences of study wellbeing. Still, empirical evidence regarding individual differences in peer relationship experiences and their associations with study wellbeing is limited, and hence, effective means to promote students’ study wellbeing and the quality of peer relationships are lacking. In this study, we contribute to bridging the gap in the literature by exploring the variation in the combinations of peer support, sense of relatedness, cognitive and affective empathy skills, and their association with study wellbeing in terms of study engagement and study burnout among upper secondary education students.

1.1 Students’ study wellbeing

Study wellbeing can be characterised as a combination of positive mental states related to studying, such as satisfaction, self-efficacy and study engagement, and absence of negative ones, such as strain or anxiety related to studying or study burnout, which contribute to successful learning (Korhonen et al., 2014; Widlund et al., 2018). In this study, we focus on upper secondary education students’ study wellbeing in terms of experiences of study engagement and study burnout. Study engagement is a positive study-related state of mind characterised by experiences of vigour, dedication, and absorption (see Schaufeli et al., 2002). It can be considered to be a symbol of an optimal studying experience. In turn, study burnout is caused by prolonged and extensive study-related stress. It comprises three symptoms: emotional exhaustion, cynicism, and sense of inadequacy (Maslach et al., 2001; Salmela-Aro, Kiuru et al., 2009, Salmela-Aro, Savolainen et al., 2009; Schaufeli et al., 2002). Emotional exhaustion manifests as chronic generic fatigue and lack of emotional energy, whereas cynicism is characterised by experiencing studying meaningless. Furthermore, sense of inadequacy is shown in feelings of failing in studying or feeling incompetent as a student. It has been shown that study wellbeing is related to life satisfaction and smooth educational transitions (Gempp & González-Carrasco, 2021; Lewis et al., 2011). In turn, study burnout has been shown to contribute to increased risk of developing depression and dropping out of school (Salmela-Aro, Savolainen et al., 2009; Bask & Salmela-Aro, 2013; Symonds et al., 2016). Both study engagement and study burnout are highly embedded in social interactions at school (Havik & Westergård, 2019; Pietarinen et al., 2014; Ulmanen et al., 2022). Especially during adolescence, peer relationships are crucial resources for fostering favourable development of study wellbeing (Mendoza & King, 2020; Rautanen et al., 2021, 2022; Rubin et al., 2008; Ulmanen et al., 2022; Wentzel et al., 2016 ).

1.2 Peer relationships

Functional peer relationships at school can be identified by the presence of students’ sense of relatedness, empathy between peers, and reciprocal peer support for studying (see Acosta et al., 2019; Garandeau et al., 2022; Van Ryzin & Roseth, 2018). Hence, these factors can be seen as indicators of functional peer relationships. The sense of relatedness entails the extent to which students feel accepted by and connected with their peers (Allen et al., 2018; Goodenow & Grady, 1993). Students who feel that they are acknoweldged, respected and cared about by the peers that they care about (Ryan & Deci, 2000; Van der Kaap-Deeder et al., 2020) have been found to have better mental health, to be happier and better adjusted at school, and to display higher levels of prosocial behaviours compared to their peers without such school experience (Allen et al., 2018; Arslan et al., 2020; Pittman & Richmond, 2007). It has also been shown that a high sense of relatedness is associated with reduced rates of truancy, increased odds of school completion and better academic achievement (Allen et al., 2018; Demanet & Van Houtte, 2012; Sari, 2012). Accordingly, the experiences of relatedness provide a cornerstone for positive peer relationships at school.

Yet, ability to relate with others calls for an understanding of another person’s emotions, i.e., empathy skills (e.g., Eisenberg et al., 1991; Feshbach & Roe, 1968), and adequate social response. In the literature, two dimensions of empathy has been distinguished: cognitive empathy referring to an ability to recognise and understand what other people think and feel, and affective empathy refering to an ability to resonate, react and tune into others’ emotional states and share the emotional experience of another person (Eisenberg, 2004). It has been shown that empathy is associated with several positive interpersonal outcomes including prosocial behaviours such as co-operation, supporting others and having friends at school, as well as reduced negative behaviours such as bullying and aggressive behaviour (e.g., Belacchi & Farina, 2012; de Wied et al., 2006; Garandeau et al., 2022; Laghi et al., 2018; Stavrinides et al., 2010; Van der Graaff et al., 2018; Van Lissa et al., 2016). Miklikowska and others (2022) recently showed that adolescents are more likely to want to make friends and spend time with peers whose empathy skills are at similar levels as their own, but also that friendships provide a context for observation and modelling of empathic behaviour. This suggests that interactions with peers, such as peer support, is an important arena for learning the empathy skills, but the empathy skills are also needed to establish well-fitted reciprocal relationship with peers.

Although empathy is essential for building experience of relatedness with peers, also adequate behavioural response is required to establish and sustain functional peer relationships. At school, this involves perceiving the peers as a resource for studying. Accordingly, peer support for studying is a central element of well functioning peer relationships at school. Peer support for studying refers to a study-related social resource available in the peer group that helps overcoming challenges in studying (Rautanen et al., 2021, 2022; see also Cohen et al., 2000). Such support is both emotional and informational in nature involving encouragement, caring, as well as direct advice and problem solving (e.g., Liu et al., 2016; Malecki & Demaray, 2003; Wentzel et al., 2016). Also the dynamics of peer support may vary, which shapes the peer support experience: students can receive peer support from their peers, give support to others or engage in reciprocal support (Feeney & Collins, 2015). It can be assumed that reciprocal peer support in studying provides opportunities for adolescents not only to develop a strong sense of relatedness in the peer group but also to be successful in their studying (Furrer & Skinner, 2003; Ulmanen et al., 2016, 2023). It has been shown that reciprocal peer support for studying contributes to experiences of study engagement among primary school students (Rautanen et al., 2021).

To sum up, experiences of relatedness, empathy skills and reciprocal peer support are distinct but interrelated constructs that provide a basis for functional peer relationships at school that can be assumed to contribute further to study wellbeing. In general, research has shown that school-related peer support is linked to higher levels of study engagement and reduced risk of experiencing study burnout symptoms (Rautanen et al., 2021; Ulmanen et al., 2016, 2022, 2023). However, there is also evidence suggesting that a high level of peer support combined with a lack of teacher support can compromise study wellbeing (Ulmanen et al., 2022, 2023). Also, results concerning the association between students’ sense of relatedness and study wellbeing are partly mixed. While a sense of relatedness has been identified as being an important factor in eliciting positive emotions in studying (Furrer & Skinner, 2003; Ulmanen et al., 2016) and to reduce study-related feelings of inadequacy, cynicism, and emotional exhaustion (Dopmeijer et al., 2022), there is also evidence that study-related burnout symptoms can spread in close relationships (Kiuru et al., 2008). Likewise, it has been shown that adolescents with well-developed empathy skills are more likely to experience high levels of psychological wellbeing (Vinayak & Judge, 2018), but there is also evidence that students reporting high levels of empathy are at a higher risk of study burnout compared to their less empathic peers (Farina et al., 2020). More recently, it was shown that affective empathy skills were linked with better study wellbeing, while cognitive empathy without the affective component contributed to lower study wellbeing among upper secondary education students (Tikkanen et al., 2022). Further studies on the role of empathy in adolescent peer relationships and students’ own perceptions of empathy has been called for (Portt et al., 2020).

Due to the complex interplay of innate predispositions, environmental influences, and personal experiences, it can be assumed that individual variations in peer relationship experiences regarding relatedness, empathy skills and peer support exist among the students (see Bronfenbrenner & Morris, 2007; Davis, 1983; Rubin et al., 2008). For example, some students might perceive all dimensions of peer relationship as well-developed and functional, while others might score high in empathy skills, but low in peer support or sense of relatedness or vice versa. Furthermore, those differences are likely to further contribute to the experiences of study engagement and study burnout. Yet, the empirical evidence on this is mixed. In this study, we took up the challenge and explored peer relationship experience profiles to see which can be detected among upper secondary students and whether these profiles differ in study engagement and study burnout.

2 Aim

The aim of this study was to explore the variations in upper secondary school students’ peer relationship experience in terms of sense of relatedness, perceived cognitive and affective empathy skills, and peer support. In addition, we explored whether the variations in peer relationship experiences were related to their experiences of study engagement and study burnout.

The following research questions were addressed:

  1. 1.

    What peer relationship experience profiles can be detected among upper secondary school students?

  2. 2.

    What differences are there between the profiles in experiences of study engagement and study burnout?

3 Methods

3.1 Finnish upper secondary education

In Finland, children start the compulsory comprehensive school at the age of seven. The comprehensive school consists of six years of primary school (grades 1–6) and three years of lower secondary schools (grades 7–9). At the age of sixteen, the students apply for entry to compulsory secondary education, either for the academic track, to attend vocational school, or a combination of these. There is also an opportunity to attend a voluntary tenth grade before continuing into secondary education. In Finland, the academic track is the most common choice for the students followed by vocational track. For example in 2020, 93% of students continued studying in either track. More specifically, 54% of comprehensive school graduates continued their studying in the academic track, whereas 39% entered the vocational school (Statistics Finland, 2020).

3.2 Participants

The participants in this study were 280 upper secondary students from all over Finland, including 217 students from the academic track (78%) and 60 students from the vocational track (21%). They were studying their second year of upper secondary education and were about 17 years old. Majority of the participants were girls (69%, n = 192) and 28% were boys (n = 79). In addition, 3% (n = 9) disclosed “other” as their gender or did not want to specify it. In terms of the representativeness of the sample, students from the academic track and girls were overrepresented.

3.3 Data collection

The data used in this study were collected between May and June 2021. The data collection was part of a larger national research project comprising two phases. In the first phase, case schools were selected using cluster sampling: the case schools, chosen from six different school districts, were situated all over the country, in both low and high socio-economic status areas, representing both rural and urban areas (Pietarinen et al., 2021). The annual data collection for the first phase began in 2017 when the participants were in the seventh grade of comprehensive school and ended when they were in the ninth grade. Participation in the study was voluntary, and informed written consent was obtained from the students’ parents. The second phase of the project started in 2021, when the participants had made the transition to upper secondary education. This study is part of the second phase. For this study, we contacted those students who had given their contact information and permission for the follow up (N = 761) at the end first phase. They were sent a link to an online survey via SMS and e-mail. As an incentive to participate in the study, four €100 gift cards were raffled among the participants. Participation in the study was voluntary and students gave their written consent to participate. According to local guidelines (Finnish National Board on Research Integrity, 2019), an ethics review is required when (a) research involves intervention in the physical integrity of research participants, (b) deviates from the principle of informed consent, (c) involves participants under the age of 15 being studied without parental consent, (d) exposes participants to exceptionally strong stimuli, risks causing long-term mental harm beyond that encountered in normal life or (e) signifies a security risk to subjects. None of these conditions were encountered in this study, and therefore, no ethics review was required.

3.4 Measurement

Students’ empathy skills were assessed using the Empathy towards peers scale, consisting of two factors: cognitive empathy (CE) and affective empathy (AE). The cognitive empathy factor included five items measuring students’ ability to recognise and understand their peer’s emotions (e.g., “It is easy for me to put myself in my peers’ shoes if they tell me they are disappointed” and “I recognise the feelings expressed by my peers”). The affective empathy scale included four items measuring students’ ability to share emotions or tune into their peer’s emotional experiences (e.g., “I am worried if my peers have concerns” and “I react easily to the enthusiasm of my peers”). The Empathy towards peers scale has been validated in a previous study (Tikkanen et al., 2022).

Peer support was assessed with the School-related peer support scale including ten items that measured providing and receiving social support for studying (e.g., “I support my friends in their studying”, “My classmates’ encouragement inspires me in my studying”). The school-related peer support scale was adapted and modified from Rautanen et al. (2022) to fit the upper secondary education context.

The Sense of relatedness scale included four items measuring students’ sense of intimacy and connection with students’ classmates (e.g. “The classmates that I like, also like me”, “I feel close to and connected with the classmates who are important to me”). The scale has been adapted and translated from the Basic Psychological Need Satisfaction scale (Van der Kaap-Deeder et al., 2020; see self-determination theory, Ryan & Deci, 2000). The items were formulated to fit the upper secondary school context. The Finnish scale was tested in a pilot study in which 22 upper secondary education students completed the survey and commented on the items.

Study engagement was assessed using the Study engagement scale (9 items), which measures energy, dedication, and absorption in studying (e.g., “When I study, I feel like I am bursting with energy” and “I can get carried away by my studying.”) (Modified from Schaufeli et al., 2002; Rautanen et al., 2021).

Study burnout was assessed using the Study burnout scale, consisting of three factors: emotional exhaustion (three items e.g., “I spend a lot of time worrying about my studies outside of school time”), cynicism (two items e.g., “Going to school feels like a waste of time to me”) and sense of inadequacy (two items e.g., “I feel inadequate in relation to my studies”) (Inspired by Salmela-Aro et al., 2009a, b; Schaufeli et al., 2002; and based on several qualitative studies e.g., Pyhältö et al., 2010).

Empathy towards peers, peer support and sense of relatedness, study engagement and study burnout were all measured using a 7-point Likert scale (1 = I totally disagree, 7 = I totally agree).

3.5 Analysis

As a preliminary analysis, we conducted confirmatory factor analyses (CFAs) to ensure that the factor structures of each construct fit the data. All models showed adequate fit with the data (see Appendix 1 for the results). Then a person-centred approach (more specifically, latent profile analysis (LPA) was used to identify subgroups of upper secondary school students based on various patterns of empathy skills (i.e., cognitive and affective empathy), experiences of peer support and sense of relatedness. LPA provides statistical criteria for model comparisons in selecting the best-fitting model, and opportunity to include predictor and outcome variables for the profile memberships (e.g., Vermunt & Magidson, 2002; Morin et al., 2018). The analyses were carried out using Mplus 8.6 software. As the histograms and statistical parameters of kurtosis, skewness, and standard deviations of the study variables suggested that the data were not normally distributed, we used the MLR (maximum likelihood robust) estimator, which produces maximum likelihood estimates with standard errors and χ2test statistics that are robust to non-normality (Muthén & Muthén, 1998–2017). Within-class variances were held constant across classes and the indicator variables were allowed to correlate in the overall model. The Akaike (AIC), the Bayesian (BIC), adjusted Bayesian (aBIC) information-based measures of fit, and a Vuong-Lo-Mendell-Rubin (VLMR) and Lo-Mendell-Rubin (aLRT), and bootstrapped (BLRT) likelihood ratio tests were used in choosing the best-fitting model (Berlin et al., 2014; Nylund et al., 2007) along with the theoretical meaningfulness of the profile solution. In addition, we used the average latent class probabilities and entropy values to assess the clarity of the various profile solutions. Furthermore, we evaluated the profile sizes, and only accepted profiles that comprised more than 5% of participants, which has been suggested to be a threshold in profile selection (see Goodman et al., 2022; Stanley et al., 2017; Spurk et al., 2020).

After selecting the profile solution, we analysed whether all the profiles differed from each other in terms of the indicator variables (i.e., peer support for studying, sense of relatedness, cognitive empathy, and affective empathy) using Mplus Model constraint option. To examine, whether the students with various peer relationship experience profiles differed from each other in terms of study engagement and study burnout symptoms, we used the auxiliary Mplus command (Muthén & Muthén, 1998–2017). The analyses were carried out with BCH procedure, which takes into account the misclassification and does not affect the class formations (Asparouhov & Muthén, 2014). In practice, the mean variables of study engagement, exhaustion, cynicism, and sense of inadequacy were included as antecedents of the latent class variable while accounting for the measurement error in classification. BCH procedure performs a weighted multiple group analysis, where the groups correspond to the latent profiles (Asparouhov & Muthén, 2021).

4 Results

Overall, the upper secondary education students reported high levels of both cognitive and affective empathy (see Table 1 for the descriptive statistics). In other words, they found it easy to recognise and identify their peers’ emotions and to tune into those experiences. Furthermore, the students perceived the peer support for studying as sufficient, indicating a willingness to help their peers with studying and an ease in seeking help from them. On average, the students also felt a strong sense of relatedness within their peer groups, reporting feelings of being respected, accepted and liked by their peers. However, there seemed to be a significant variation between the individuals.

The dimensions of functional peer relationships were found to be associated with each other. This means that the students with high levels of cognitive empathy, also tended to score high in affective empathy. Moreover, the students who perceived their empathy skills as well-developed, were more likely to engage in reciprocal peer support practices, and feel a strong sense of relatedness, compared to students who reported lower levels of cognitive and affective empathy towards peers.

The dimensions of functional peer support were related to higher levels of study engagement (see Table 1). Furthermore, the students’ perceptions of peer support for studying as well as sense of relatedness were associated with reduced levels of all burnout symptoms, that is, exhaustion, cynicism and sense of inadequacy. Students’ reports of cognitive empathy were related to reduced levels of cynicism, while the other associations between empathy skills and burnout symptoms were not statistically significant.

Table 1 The descriptive statistics of the study variables and correlations between them
Full size table

4.1 The peer relationship experience profiles

The LPAs were run with one to six profiles (Table 2). According to VLMR, aLRT and BLRT, adding a new class increased the model fit to three profiles. Similarly, BIC and aBIC showed a clear decrease to three profiles, after which the BIC started to increase and the decrease in aBIC levelled off. The three-profile solution was considered to be the most parsimonious model, and it had a clear theoretical interpretation. In other words, all three profiles provided novel information about how the elements of peer relationships can combine within individuals. Therefore, the three-profile solution was selected. The profiles were considered to have sufficiently large memberships (> 5% of the cases, see e.g., Goodman et al., 2022; Stanley et al., 2017). The latent class probabilities (> 0.80) showed adequate separation between the profiles, reflecting the clarity of the profile solution (see Table 2). To ensure the accuracy of the profiles, we assessed the posterior probabilities of each participant. Specifically, out of all participants, only six individuals had posterior probabilities below 0.6, with two participants most likely belonging to the smallest profile, two to the second smallest profile, and one to the largest profile. The examination indicated a clear classification into the profiles. Further, the entropy values, which are used to assess the accuracy with which models classify individuals into their most likely class ranging from 0 to 1, with higher scores representing better distinction between latent profiles, were examined. The entropy value was above 0.80, indicating a high level of precision in the profiling.

Table 2 Information criteria and likelihood ratio tests for the 1–6 profile solutions
Full size table

The first peer relationship experience profile was largest with an 81% share (n = 227). The students with this profile reported being satisfied with peer support and had a strong sense of relatedness in their peer group (see Fig. 1). More specifically, the students in this profile wanted their peers to do well at school, supported their peers in studying and had the courage to ask the peers study assistance. They also felt that the peers who they cared about also cared about them. In addition, they had well-developed skills in both cognitive and affective empathy, meaning that they possessed the ability to accurately perceive and understand their peers’ emotions as well as to emotionally resonate with them. This profile was named functional peer relationship profile.

Fig. 1
figure 1

The peer relationship experience profiles (Abbreviations: SUP = peer support for studying, REL = sense of relatedness, CE = cognitive empathy, AE = affective empathy). Note. The means within a row sharing the same subscript do not differ statistically significantly (p < .05) from each other

Full size image
Fig. 2
figure 2

The differences in study engagement and study burnout between the profiles. (Abbreviations: Eng = study engagement, Exh = exhaustion, Cyn = cynicism, Inad = sense of inadequacy)

Full size image

In the second profile, high levels of peer support and empathy skills were combined with lower levels in sense of relatedness. Compared to the students in other profiles, the students in this profile scored significantly lower in sense of relatedness. In other words, the students in this profile exhibited a dichotomy in the functionality of peer relationships; on one hand, they demonstrated well-developed empathy skills and engagement in reciprocal peer support, while on the other hand, they displayed low sense of relatedness. This means that although the students in this profile were willing and felt competent in supporting their peers as well as found it easy to recognise and tune into each other’s’ emotions, they did not feel like being cared about by the peers they cared about. The profile was named isolated profile to reflect the contraposition of social proficiency and felt social isolation. This profile comprised 11% (n = 29) of the participants.

The third profile comprised 8% (n = 22) of the participants, and it was a combination of relatively low levels of affective empathy skills, well-developed cognitive empathy skills, high satisfaction with peer support, and strong sense of relatedness. In other words, the students within the profile felt being able and willing to help others and having courage to ask help from others. They also felt that their peers cared about them. Despite reporting that they were competent in recognising and understanding the emotions expressed by their peers, they felt it more difficult to be happy for their peers and were less likely to worry if their peers had concerns than students with other profiles. Hence, the students in this profile demonstrated a nuanced set of empathy skills: while they possessed well-developed cognitive empathy skills, indicating a strong capacity to comprehend others’ perspectives and emotions, their ability to emotionally respond to peers’ emotional states appeared to be less pronounced. For example, these students may understand when and why a peer is excited or distressed, but they may not share those emotions. Hence, this profile was named emotionally unrelated profile.

4.2 Differences in study engagement and study burnout between the profiles

Some differences were detected in study engagement (χ2(2) = 24.42, p < .001), cynicism (χ2(2) = 10.99, p = .004), and sense of inadequacy (χ2(2) = 22.31, p < .001) between the profiles (see Fig. 2). More specifically, the students with the functional peer relationship profile reported highest levels of study engagement. In other words, they reported feeling vigorous, energetic and dedicated to their studying more often than students with either isolated or emotionally unrelated profiles. Similarly, the students with the functional peer relationship profile reported the lowest levels of cynicism compared to other profiles, that is, they were less likely to feel that going to school is a waste of time compared to students with emotionally unrelated (p = .045) and isolated profiles (p = .004). The students with the functional peer relationship profile also reported lower levels of sense of inadequacy than students in the isolated profile (p < .001), meaning that they felt like they were failing in studying less often. However, no differences were detected in study engagement, cynicism or inadequacy between the isolated and emotionally unrelated profiles. No differences were detected in experiences of exhaustion between any of the profiles (χ2(2) = 3.52, p <.172).

5 Discussion

5.1 Methodological limitations and reflections

The study has some methodological limitations that need to be considered when interpreting the results. First, the data were cross-sectional, not allowing us to draw causal conclusions between the dimensions of peer relationships, study engagement and burnout. Therefore, longitudinal studies are needed to analyse the developmental trajectories of the indicator variables (peer support, sense of relatedness, cognitive empathy, and affective empathy) and their dynamics as well as their potential effects to study wellbeing. Furthermore, although the respondents were from all over the country, the data were not representative in terms of gender and study tracks. Because of this limitation and the limited sample size, we could not explore whether background variables such as gender and study track would predict membership of the profiles that could have provided support for the validity of the profiles. However, the validity of the profiles was supported by the finding that students with the fuctional peer relationship profile reporting highest values in study engagement and lowest levels of cynicism, which has been also found in variable-centred studies (e.g., Dopmeijer et al., 2022; Rautanen et al., 2021; Vinayak & Judge, 2018). The very small size of the emotionally unrelated profile, and the differences in profile sample sizes interfere the profile comparisons in some of the analyses. Therefore, further studies with more representative data sets are needed to validate the profiles in various educational settings. Second, the data were collected during the COVID-19 pandemic, which might have affected the results. Third, the criteria for selecting the number of profiles are ambiguous (Nylund et al., 2007). Therefore, further studies exploring whether similar profiles can be found among students in other contexts and age groups is important. Last, as our study was conducted in one country only, studies investigating students’ peer relationships in terms of empathy skills, peer support, sense of relatedness in other socio-cultural contexts would increase the understanding of contextual and invariant elements of functional peer relationships.

5.2 Findings in the light of previous studies

The aim of the study was to expand the understanding of the individual variation in upper secondary education students’ experiences of peer relationships in terms of sense of relatedness, cognitive and affective empathy skills, and peer support, and whether the variation in peer relationships was related to their experiences of study engagement and study burnout. The results showed that although majority of the students fell within the functional peer relationship profile, there were also students with the isolated profile who did not feel as strong a sense of relatedness in the peer group, and emotionally unrelated profile holders, who found it difficult to tune into their peers’ emotions. The students in isolated and emotionally unrelated profiles could be considered as risk profiles in terms of study wellbeing because they reported lower levels of study engagement and higher levels of study burnout symptoms than students within functional peer relationship profile. The study is among the first to have taken a person-centred approach to studying peer relationship experience as a multidimensional construct.

The results showed that all participants estimated the peer support to be adequate and scored high in cognitive empathy skills. Despite this, there was variation in the sense of relatedness and affective empathy skills. More specifically, the results showed that the students in the emotionally unrelated profile felt their peers’ support to be functional, had developed a strong sense of relatedness, and were able to recognise and understand their peers’ emotions, but they could not tune into or resonate with their peers’ emotional states to the same extent as the students in other profiles. Because the previous findings have suggested that cognitive empathy is a precondition for affective empathy (see Feshbach & Roe, 1968; Lamm et al., 2010), it might be that the students in the emotionally unrelated profile were at an earlier stage in learning empathy skills than students within other profiles. In addition, the results revealed an isolated profile with a significantly weaker sense of relatedness than the students in other profiles. However, they reported elevated levels in all other elements of functional peer relationships. In other words, although they helped and were helped by their peers in studying and understood their peers’ emotional states and could even relate to those experiences, they did not feel that they were appreciated and cared about by their peers to a high extent. One reason for the finding might be that they were more likely to give support to their peers than to receive it, which had resulted in their need for relatedness not being met to the same degree as it was for other students (see Feeney & Collins, 2015). The findings imply that high levels of peer support and cognitive empathy skills are linked with each other in all students, which can either mean that being able to understand peers’ emotional experiences is a precondition for providing and receiving functional peer support (see Eisenberg et al., 1991; Feshbach & Roe, 1968), or that peer support provides an arena in which such skill can be learned (Miklikowska et al., 2022). However, the results suggest that the students could experience high levels of peer support and cognitive empathy without the presence of high affective empathy and sense of relatedness. This implies that sense of relatedness does not automatically result from peer support and cognitive empathy.

The results showed that the peer relationship experience profiles displayed different combinations of the elements of peer relationship, that is, cognitive and affective empathy, sense of relatedness and peer support for studying. Furthermore, these varying combinations seemed to differ in terms of study wellbeing. The students in the functional peer relationship profile displayed highest levels of study engagement and lowest levels of cynicism and inadequacy. However, no significant differences in exhaustion were detected between the profiles. The results indicate that cynicism and inadequacy are embedded in social interrelationships between students (see also Tikkanen et al., 2021) to a larger extent than exhaustion, which is likely to be related to general workload. Our results further suggest that peer support and cognitive empathy are necessary for the peer relationship experience to support study wellbeing. However, affective empathy skills and a sense of relatedness also need to be in place for a peer relationship experience to optimally support study wellbeing. In other words, both weak affective empathy skills and low sense of relatedness seem to be linked with higher risk of burnout and reduced study engagement, although the mechanisms are likely to differ.

The study provides several directions for future studies. First, further longitudinal studies are needed to gain a more fine-graded understanding of the development and interaction between functional peer relationship elements. For example, whether the empathy skills are primarily learned through peer support or whether cognitive empathy skills are an essential precondition for the peer support to work should be explored. Longitudinal designs would also allow us to map the development of peer relationships. Second, as the sample size did not allow us to examine who the students in the profiles are, it would be important to investigate, with more comprehensive data sets, whether gender, study track or their combinations explain the profile memberships. Third, whether the profiles differ in terms of study success, and future study and career paths needs to be explored. Last, the mechanisms explaining the development of study wellbeing in various profiles should be explored in future with large, longitudinal data sets.

5.3 Conclusions

The results suggested that differences exist in the anatomy of peer relationship between the students, especially in their affective empathy skills and sense of relatedness. The students who scored lower in either affective empathy or sense of relatedness had higher risk of experiencing study burnout and had fewer study engagement experiences despite scoring high in peer support and cognitive empathy. The results indicate that in order to capture the function of peer relationships, it should be understood and studied as a multidimensional construct.

5.4 Practical implications

The findings have implications for enhancing upper secondary education students’ study wellbeing.

As the results showed that differences existed in students’ affective empathy skills and sense of relatedness in the peer group, various kinds of support are likely to be needed in upper secondary education. For example, in the everyday life of school, it is beneficial that the teachers help the students to (a) practise tuning into a peer’s emotional state in a relevant manner and collectively discussing how to react to them, and (b) intentionally learning to recognise and practise reciprocity in social support. In particular, the students who find it difficult to tune into others’ emotions might benefit from training those skills as a part of their studying in peer groups. The results also showed that the levels of exhaustion did not differ between the profiles, implying that peer relationships, no matter how functional they are, cannot help to reduce such experiences. Hence, the education policy makers as well as upper secondary education teachers should consider the workload related to studying carefully.