The Relationship Between Friendships and Social Problem-Solving in Adolescence

Abstract

The aim of the study was to examine the relationship between friendship functions and social problem-solving among adolescents in same- and opposite-sex friendships at ages 12 and 16 (N = 304). Social problem-solving was assessed using the Social Problem-Solving Inventory–Revised, while friendship functions were measured using the McGill Friendship Questionnaire. Both instruments demonstrated satisfactory psychometric properties in all age subsamples. The analyses focused on age-, sex-, and friendship-type differences in friendship functions and social problem-solving characteristics. Based on the results, reliable alliance and self-validation were the most important friendship functions in both same- and opposite-sex friendships. Same-sex friendships were characterized more strongly by positive orientation and rational problem-solving, whereas opposite-sex friendships showed higher levels of negative orientation, impulsivity, and avoidance. In addition, several previously identified age- and sex-related characteristics of adolescent friendships were partially confirmed. The findings emphasize the important role of friendships in adolescents’ social functioning and suggest that friendship context is associated with the quality of social problem-solving. At the same time, the interpretation of opposite-sex friendship patterns and profile-specific differences requires caution because of the relatively small subgroup sizes. These findings indicate the need for further research on friendship-specific social problem-solving in adolescence.

1. Introduction

Our relationships play a crucial role in social and emotional development across adolescence (Rose et al., 2016; Walker et al., 2016). Among these relationships, friendships become increasingly important because they provide emotional support, reciprocity, intimacy, and opportunities for the management of interpersonal conflicts. Although several studies have examined adolescent friendships and social problem-solving separately, much less is known about how specific friendship functions are associated with particular social problem-solving characteristics in same- and opposite-sex friendships.

Recent studies (Graber et al., 2016; Walen & Lachman, 2000) suggest that conflicts and disagreements within friendships are not necessarily destructive. Constructive social problem-solving may strengthen the relationship and contribute to adolescents’ social development. However, most previous studies have focused primarily on same-sex friendships, while much less attention has been paid to opposite-sex friendships and their specific problem-solving characteristics.

1.1. The Interpretation of Friendships: Characteristics, Components, and Functions

The studies that analyze the characteristics, components (e.g., Rosenthal & Kobak, 2010), and the age-related peculiarities of a friendship (e.g., Walker et al., 2016) are in close connection to one another because most of them that examine age-related characteristics take adult friendship features as their basis and they examine adolescent friendships from that point of view. This is no coincidence since adolescent friendships show the most similarity to adult ones (Kurzovic, 2015). The interpretation based on age phases (e.g., Sullivan, 1953) is strongly related to the interpretation of friendship based on family socialization (Rose et al., 2016), which shows an overlap with the approach that puts the fulfilment of needs into the focus (Jordan, 2013).

In terms of the universally accepted view, adult friendships, similarly to other social relations, are based on multiple components and involve two people (Hartup, 1996; Hartup et al., 1988; Rose et al., 2016), and they are fundamentally characterized by voluntarism, the lack of formalities, and positive emotional attitude (Graber et al., 2016; Helm, 2013). This relationship can be equipped with several other characteristics (e.g., often spending time together, equality, reciprocity, and support) whose frequency and intensity are dependent on the parties’ age, sex, sociocultural attributes, as well as their micro- and macro-environment (Hartup, 1992; Newcomb & Bagwell, 1996).

Rosenthal and Kobak (2010) see equality among parties and strong commitment to the relationship as the basis of a friendship, which enable the development of mutual assistance and a strong emotional bond. Furman and Rose (2015) also regard equality as well as reciprocity as basic components and, in their absence, intimacy, support, emotional security, and the feeling of belonging to someone cannot develop (Camirand & Poulin, 2019). Davis and Todd (1985) have defined similar components, and conflict can be found among them. According to the researchers, the conflicts that occur in a friendship put it to the test in all cases. The conflicts can be situational (not caused by the parties and something must be done to stop it) or personal (its source is one of the parties and their action directly affects the other one as well as the relationship). The more effectively and sooner the parties manage to solve the conflict, the stronger they can feel the friendship to be. In addition to this, strength can be defined by the acknowledgement of and the degree of commitment to the other one.

According to Rose et al. (2016), one of the motors of change is the system of expectations in relation to friendship, and these expectations are based on situations experienced in other relationships. This underlies the validity of friendships interpreted through family socialization processes (Rose et al., 2016). Considering this, the experience and interpretation of friendship from adolescence onwards is based on those social capabilities (e.g., empathy, collaboration, conflict management, and delay) that the individual acquired from parents and siblings through following their models and during the collective interpretation of situations, family programs, as well as free-time activities. The need fulfilment approach (Helgeson & Lopez, 2010; Jordan, 2013) highlights that friendships contribute to the fulfilment of emotional and interpersonal needs that may not be sufficiently satisfied in family relationships. These functions become especially important during adolescence, when peers gradually gain a more central role in everyday social life.

The friendship functions described by Mendelson and Aboud (2014) closely fit the need fulfilment model: friendships can be interpreted as relationships that enable the fulfilment of different functions. The previously defined components of friendships can also be interpreted as functions; however, the researchers highlight the fact that the six functions that they defined must all be present and they must strengthen one another for a long time—in a relationship that is longer than six months, according to their study. The six base functions are stimulating companionship, help, emotional security, intimacy, loyalty to the other person, and their acknowledgement. If these are continuously characteristic of the relationship, this can not only be socially advantageous to the parties, but it can have a positive influence on educational–professional development as well (feeling socially secure enables calm learning and professional development), based on the research carried out by Mendelson and Aboud (1999) as well as Laird et al. (2001).

1.2. Social Problem-Solving in Adolescent Friendships

Quite a number of models are known in relation to the interpretation of social problem-solving (e.g., Crick & Dodge, 1994; D’Zurilla & Goldfried, 1971; D’Zurilla et al., 2002, 2004; Frauenknecht & Black, 2010; Spivack & Shure, 1976), based on which several measurement tools have been developed for the analysis of the entirety of problem-solving or just some of its subprocesses. The SPSI–R (Social Problem-Solving Inventory–Revised), developed by D’Zurilla et al., was used in this research; therefore, its theoretical background will be presented when describing the characteristics of social problem-solving. According to D’Zurilla et al. (2004), social problem-solving consists of both problem orientation and problem-solving style, and these dimensions are especially relevant in adolescent friendships where disagreements become increasingly emotionally significant.

During the orientation phase, adolescents may approach interpersonal problems with either a positive or a negative motivational–emotional attitude (Maydeu-Olivares & D’Zurilla, 1996). The person with a positive orientation interprets the problem as a challenge, believes in a successful solution, and their feeling of self-efficacy is high. In contrast, negative orientation is characterized by frustration, insecurity, and perceiving interpersonal conflicts as threatening situations. The definition of the actual problem happens in the solution phase—seeking alternative modes of solution, deciding what the most appropriate solution is, and implementing it.

Based on their empirical research conducted among adolescents and adults, D’Zurilla et al. (2004) distinguished between three general social problem-solving styles: rational, impulsive, and avoiding, which can manifest themselves in behavior because of problem-solving thinking. These can be regarded as general forms of problem-solving behavior. Rational problem-solvers focus on relevant facts throughout the problem-solving process. They consider several alternative solutions and evaluate their potential consequences. They show interest in their partner’s perspective and opinions, ask questions when clarification is needed, stand by their own position, communicate assertively, acknowledge when they are wrong, and take responsibility for their actions. Impulsive problem-solvers consider only a limited number of relevant facts, while negative emotions strongly influence both their decision-making and the implementation of solutions. They tend to make decisions quickly and impulsively, often perceive the other person as an opponent, and frequently strive to enforce their own interests while disregarding those of others. Avoidant problem-solvers tend to terminate the problem-solving process prematurely, shift responsibility to others, or avoid addressing the problem altogether. Consequently, they assume the least responsibility for the outcomes of their actions (Chang et al., 2004).

The characteristics of social problem-solving are closely connected to developmental changes in adolescent friendships. Friendships during the kindergarten years are one-sided and fickle, but the friendships among young students are based on collaboration; however, every quarrel and difference in opinion threatens the relationship and, because of this, the parties aim for an agreement (Selman, 1981). From adolescence onwards, friendships become increasingly characterized by reciprocity, emotional security, intimacy, and mutual commitment, which also influence how conflicts and disagreements are managed.

Significant conflicts may emerge when a relationship fails to fulfil the functions for which it was established. In contrast, the fulfilment of these functions evokes positive feelings and cognitions in both parties and contributes to the development of a sense of security. One such function is devotion. From the onset of adolescence, devotion becomes gradually more important in a friendship, which still comes with jealousy at the beginning—more so in friendships among girls. According to Selman (1981), when the relationship is less burdened by jealousy and when reciprocity pairs up with the friends thinking about each other as individuals with autonomous wills and they start treating the other one so, a mature friendship has developed where the members are able to create a dynamic balance of closeness and distance that is satisfactory for both of them. If the degree of jealousy does not decrease in adolescent friendship, this may become the source of quarrels, which can result in the unravelling of the relationship (Lamm et al., 1998).

Self-validation and acknowledgement of the other person’s achievements, behavior, and personal characteristics (e.g., “I like it that you’re conscientious”) also become increasingly important during adolescence. While difference in opinion often goes together with the negative description of the other person during the early school years (they can come to think the opposite of what they learnt about the other person very quickly), the negative characteristics of the other one are defined based on the given situation in adolescence and not generalized, which shows the more and more situation-specific evaluation of the other person (Mendelson & Aboud, 1999).

The quality of emotional support and help provided by the friend is also closely related to social problem-solving within the relationship. Based on Pikó’s research (1996), members of girl friendships state a higher degree of support in adolescence as compared to boys—they need more support from friends both emotionally and cognitively. However, there is no sex-related difference in relation to satisfaction with this (boys do not require the support characteristic of girls, they feel that what they give and receive is enough). The loosening of the friendship is mainly caused by a decrease in emotional support, which can also cause its unravelling.

The weakening of trust and the more frequent denial of help can also first lead to the loosening and then the unravelling of a friendship. Unrequited requests for favors as well as asking for too many of them can often lead to open conflicts and to a spectacular break in the friendship (Zsuzsa, 2004). In the case of betraying a friend (e.g., breaching sworn secrecy), the loosening of the friendship takes a significantly shorter time; after it comes to light, the relationship ends soon. However, this is also dependent on how long the friendship is as well as personal characteristics; for example, one’s ability to forgive (Lamm et al., 1998).

1.3. Friendship Functions and Social Problem-Solving in Adolescence

Friendships become increasingly important during adolescence because they provide emotional support, reciprocity, intimacy, and opportunities for the management of interpersonal conflicts (Rose et al., 2016; Walker et al., 2016). Developmental models of friendship (e.g., Corsaro, 1985; Sullivan, 1953) emphasize that adolescent friendships gradually become more like adult friendships, as reciprocity, trust, mutual respect, and emotional closeness gain increasing importance. One of the most striking features of adolescence is the gradual emotional detachment from parents and the strengthening role of peers and friends in everyday social life (Hartup, 1983; Steinberg & Silverberg, 1986). From early adolescence onwards, friendships are increasingly characterized by balanced mutual dependence, emotional security, intimacy, and loyalty (Selman, 1981; Rose et al., 2016).

These developmental changes also influence how adolescents manage interpersonal disagreements and social problems within friendships. In contrast to childhood, differences in opinion and arguments do not necessarily threaten the relationship during adolescence—discussing emotions, perspectives, and disagreements may even strengthen the friendship (Zsuzsa, 2004). At the same time, friendships provide an important context for the development and practice of social problem-solving skills.

Although friendship functions and social problem-solving have mostly been examined separately in previous research, several theoretical links can be assumed between them. Friendship functions may create an interpersonal context that either supports or hinders adaptive problem-solving processes during adolescence (Mitic et al., 2021; Graber et al., 2016). Emotional security may contribute to lower negative orientation because adolescents who feel emotionally safe in a friendship are less likely to perceive disagreements as threatening situations (Bukowski et al., 1994; Shah et al., 2024). Similarly, reliable alliance and commitment to the relationship may decrease avoidance tendencies, as maintaining the friendship can motivate adolescents to remain engaged in social problem resolution instead of withdrawing from it.

Intimacy, help, and mutual support may facilitate rational problem-solving because these friendship characteristics provide opportunities for perspective-taking, communication, emotional sharing, and the joint consideration of alternative solutions (Bagwell, 2026; Mitic et al., 2021). In contrast, friendships characterized by weaker reciprocity or emotional support may increase impulsive and emotionally driven reactions during disagreements. Self-validation may also play an important role in interpersonal problem-solving. Adolescents who feel accepted and valued by their friends may approach conflicts with greater self-confidence and assertiveness, which can support constructive problem-solving (Shah et al., 2024). Positive peer relationships may, therefore, function as protective factors in adolescents’ social–emotional adjustment and interpersonal functioning (Carapeto et al., 2025).

Research also suggests that emotional security and mutual support contribute to the stability and durability of friendships (Aboud & Mendelson, 1996; Bukowski et al., 1994). Friendships characterized by high emotional security are associated with stronger trust and support, while loyalty and appreciation become increasingly important from middle adolescence onwards (Mendelson & Aboud, 2014). These characteristics may also influence how adolescents interpret and solve interpersonal problems within friendships.

These theoretical connections suggest that friendship functions and social problem-solving are closely interconnected during adolescence. However, relatively little empirical research has examined how specific friendship functions are associated with particular social problem-solving dimensions, especially in same- and opposite-sex friendships. Although both adolescent friendships and social problem-solving have been extensively studied, much less is known about how the specific functions of friendships relate to adaptive (positive orientation and rationality) and non-adaptive problem-solving dimensions (negative orientation and impulsivity and/or avoidance) in different friendship contexts. Few studies have examined whether these relationships differ between same-sex and opposite-sex friendships, as well as across different developmental stages of adolescence. This question is particularly important because friendships gradually become more emotionally intimate, reciprocal, and identity-relevant from early to middle adolescence, which can fundamentally influence how adolescents interpret and manage interpersonal problems.

Therefore, the present study focuses on the developmental and relational boundaries of social problem-solving in adolescent friendships. We examined whether the functions of friendship are related differently to the dimensions of social problem-solving in same-sex and opposite-sex friendships during the early (age 12) and middle (age 16) stages of adolescence. Rather than treating friendship quality and social problem-solving as independent constructs, the study approaches them as interconnected interpersonal processes that shape adolescents’ problem-solving and relationship maintenance.

1.4. Aim and Hypotheses

The aim of this study was to examine how friendship functions relate to adaptive and non-adaptive social problem-solving in same-sex and opposite-sex friendships during early and middle adolescence. Based on developmental theories of friendship and social problem-solving, we hypothesized that (1) stronger friendship functions—particularly emotional security, intimacy, reliable alliance, help, and self-validation—are associated with more adaptive social problem-solving (high positive orientation and rationality) as well as with less adaptive problem-solving (high negative orientation, impulsivity, and avoidance) (Kasik, 2015; Kasik et al., 2018).

We also hypothesized that (2) these relationships vary depending on the type of friendship and developmental stage; that is, adaptive relationships between friendship functions and social problem-solving are expected to be stronger in same-sex friendships than in opposite-sex friendships, as same-sex friendships are generally characterized by greater emotional closeness, reciprocity, and stability during adolescence (Mendelson & Aboud, 1999; Selman, 1981).

Finally, we hypothesized that (3) latent profile analysis would identify distinct profiles characterized by different combinations of friendship functions and dimensions of social problem-solving (Gál et al., 2025; Fejes et al., 2023). Profiles characterized by stronger friendship functions are expected to exhibit more adaptive problem-solving characteristics, while profiles characterized by weaker emotional security and reciprocity are expected to exhibit non-adaptive orientations and styles.

2. Methods

2.1. Participants

The study involved Hungarian 6th grade (12-year-old) and 10th grade (16-year-old) students, and all participants were native Hungarian speakers. We selected these two age groups because findings from previous developmental psychology research suggest that significant changes occur between early and middle adolescence in both the characteristics of friendship and social problem-solving processes.

The study is part of a larger research project aimed at examining various components of social competence, in which students from public schools in 5 Hungarian cities participated (out of 12 classes). The schools participated in the program on a voluntary basis. Here, 3 students were excluded from the analysis due to missing data; therefore, the final sample consisted of 304 adolescents (6th grade: n = 138, M_age = 13.09 years, SD = 0.89, n_boys = 62, n_girls = 76; 10th grade: n = 166, M_age = 16.18 years, SD = 0.52, n_boys = 70, n_girls = 96). Consequently, the results should primarily be interpreted in the context of Hungarian students’ adolescent peer relationships.

Since the study examined same-sex and opposite-sex friendships separately, participants were first asked to indicate whether they had a same-sex and/or opposite-sex friend whom they considered a close friend at the time of data collection. Participants were asked to think of a specific friend while answering the friendship-related questions, rather than thinking about friendships in general. The questionnaires did not require participants to name their “best friend”; however, the instructions emphasized that the answers should refer to a friendship that is personally very important to them. Among the 6th graders, 4 adolescents (1 boy and 3 girls) reported having no same-sex friends, while among the 10th graders, this number was 6 (2 boys and 4 girls).

The lack of friends of the opposite sex was more common: 41 students (32 boys and 9 girls) in 6th grade and 38 students (23 boys and 15 girls) in 10th grade indicated that they had no friends of the opposite sex. No participant indicated a lack of both types of friendships. Consequently, analyses of same-sex friendships were conducted on 134 6th grade and 160 10th grade students, while analyses of opposite-sex friendships were conducted using data from 99 6th grade and 128 10th grade students.

2.2. Measurements

Two measurement tools were used in the study. The Social Problem-Solving Inventory–Revised (SPSI–R, D’Zurilla et al., 2002; Hungarian version: Kasik et al., 2009) was used to examine social problem-solving, and the McGill Friendship Questionnaire (MFQ, Mendelson & Aboud, 2014) was used to analyze the functions of the friendship.

SPSI–R: The questionnaire contains 25 statements, and it measures the characteristics of problem-solving along 5 factors (factor/5 statements): positive orientation (PO) and negative orientation (NO), rational (R), impulsive (I), and avoiding (A) problem-solving styles—the contents of these are reviewed in the study’s theoretical background. The content of the factors was described in the theoretical background. The statements must be evaluated on a scale of 5 (0 = absolutely not typical of me to 4 = absolutely typical of me). Similar to the earlier Hungarian studies (Kasik, 2015, the reliability of the questionnaire was adequate in the case of both age groups, and in the case of all factors in the case of types of friendships (Cronbach-α = NO: 0.77–0.82; PO: 0.79–0.86; R: 0.80–0.86; I: 0.70–0.79; A: 0.76–0.87).

MFQ: MFQ contains 30 statements, and it measures the characteristics of the functions of friendships along 6 factors (factor/6 statements): stimulating companionship (SC), help (H), intimacy (I), reliable alliance (RA), self-validation (SV), and emotional security (ES). The statements must be evaluated on a scale of 8 (1 = never to 8 = always). Similar to the data from the adaptation (Kasik et al., 2024), the reliability of the questionnaire was adequate in both age groups, and in the case of all factors in the case of types of friendships (Cronbach-α = SC: 0.70–0.79; H: 0.74–0.82; I: 0.72–0.80; RA: 0.77–0.85; SV: 0.73–0.79; ES: 0.81–0.85). The factor of stimulating companionship measures how much the individual desires to perform more activities with their friend (e.g., learning and having fun), and how much they think these mutual activities induce positive feelings and how much they aspire to maximize the time spent together (Buhrmester, 1990). The statements in the factor of help (which are separate from emotional security, intimacy, and the creation of an intimate relationship) are about how much the parties provide help to each other by taking necessity and possibility into consideration, and how much they think that help should be reciprocated. The statements in the factor of intimacy measure how much the parties accept each other’s feelings and thoughts and how much they support the other one, which creates an opportunity for the parties to openly and honestly express their feelings and thoughts to themselves, each other, and the ones outside of the friendship. The factors of emotional security measure how suitable the friendship is for it to create joy and relief to each other and for their secrets to be kept safe—in a problematic or a non-problematic context (Wright, 1974). The factor of reliable alliance measures the strength of staying in the friendship, which requires mutual loyalty and constant availability. It also reveals what the degree of effort is to find the most efficient solution of sustainment in the case of problems, let downs, and arguments in the relationship (Selman, 1981). The factors of self-validation are about how much the members of the relationship strengthen each other’s self-image, how they express criticism, and how rivalry based on inspirational, positive, and upward comparison manifests itself between the members of the relationship (Bukowski et al., 1994).

2.3. Data Analysis

We assessed the reliability of the questionnaires for all age groups and scales using Cronbach’s alpha (see the data for the measurements). Prior to the analyses, we examined the distribution of the variables using measures of skewness, kurtosis, and visual inspection (

Table 1. Results of the normality test (N = 304).

Factors (Type of Friends)	Skewness	Kurtosis
SPSI–R (same-sex friends)
Negative orientation	0.69	0.02
Positive orientation	−0.05	−0.41
Rationality	−0.47	−0.08
Impulsivity	0.08	−0.52
Avoidance	0.74	−0.18
SPSI–R (opposite-sex friends)
Negative orientation	0.47	−0.56
Positive orientation	−0.79	0.88
Rationality	−0.54	−0.35
Impulsivity	0.03	−0.28
Avoidance	1.22	2.21
MFQ (same-sex friends)
Stimulating companionship	−1.28	1.72
Help	−0.90	0.21
Intimacy	−1.56	2.02
Reliable alliance	−2.14	5.40
Self-validation	−0.63	−0.28
Emotional security	−0.70	−0.18
MFQ (opposite-sex friends)
Stimulating companionship	−1.81	3.83
Help	−0.97	0.68
Intimacy	−1.23	0.89
Reliable alliance	−1.57	2.30
Self-validation	−0.97	0.68
Emotional security	−0.87	0.02

). Since some friendship-related variables exhibited a ceiling effect, we took these characteristics into account when interpreting the results. Nevertheless, the general distribution characteristics and sample size allowed for the use of parametric procedures. The values for skewness and kurtosis were within acceptable limits for most variables. Although we identified a slight ceiling effect in two friendship functions—same-sex friends: reliable alliance (kurtosis 5.40), and opposite-sex friends: stimulating companionship (kurtosis 3.83)—the distributions were normal given the sample size, so we performed parametric analyses.

To examine developmental and sex differences in friendship functions and social problem-solving, we conducted multivariate and univariate parametric analyses. These analyses allowed for comparisons across age groups, sex groups, and friendship contexts (same-sex and opposite-sex friendships) at the variable level.

Since the primary goal of the study was not only to examine relationships between variables but also to identify naturally occurring interpersonal patterns, latent profile analysis (LPA) was conducted to identify profiles of friendship functions and social problem-solving in same-sex and opposite-sex friendships. Profile solutions were evaluated using information criteria, classification indices, and theoretical interpretability. For MFQ indicators, entropy values indicated good to excellent classification accuracy (0.831–0.995), supported by high average posterior probabilities (0.929–0.999). The selected solutions showed well-separated profiles in most cases, although the 12-year-old opposite-sex friendship model included a small class (3.2%), which was interpreted with caution. For the 16-year-old same-sex friendship model, classification quality was acceptable (entropy = 0.831), while the opposite-sex model indicated high separation (entropy = 0.945).

For SPSI–R indicators, two-profile solutions were retained across age groups based on parsimony and interpretability. Classification quality ranged from moderate to high (entropy = 0.671–0.964), with consistently high posterior probabilities (0.908–0.993). Class sizes were generally adequate (37.3–44.8%), supporting stable solutions. No stable profile solution emerged for the 16-year-old opposite-sex friendship model, which was, therefore, excluded from further analysis.

We employed cross-tabulation analyses to explore the relationships between friendship function profiles and social problem-solving profiles. These analyses allowed us to examine whether certain friendship patterns are associated with adaptive or less adaptive forms of social problem-solving.

We used structural equation modeling (SEM) to examine the relationships between the friendship functions (6 factors) and social problem-solving (5 factors). Although we initially assessed social problem-solving based on five factors (positive and negative orientation, rationality, impulsivity, and avoidance), in the SEM analyses we modeled these factors as two correlated latent constructs: adaptive social problem-solving (positive orientation and rationality) and non-adaptive social problem-solving (negative orientation, impulsivity, and avoidance). We estimated separate models for same-sex and opposite-sex friendships to examine whether the context of friendship modifies the relationship between friendship functions and social problem-solving.

2.4. Data Collection and Ethical Approval

The questionnaires were completed in two phases. In the first phase, participants answered questions regarding same-sex friendships, and approximately 10 days later they completed the questionnaires again regarding opposite-sex friendships. The questionnaires were administered during regular school hours under teacher supervision. Prior to data collection, teachers received preparatory information about the study’s objectives, procedures, and ethical requirements. During administration, teachers were present only to ensure the organizational conditions of data collection and did not have access to the completed questionnaires or the students’ responses. Participants were informed that their answers would remain confidential and would not be shared with teachers, parents, or peers.

Participation was voluntary and anonymous. Prior to data collection, permission was obtained from school administrators, and written parental consent was collected in accordance with national regulations governing psychological and educational research involving minors. In addition to parental consent, students provided informed assent before participation and were informed that they could refuse participation or withdraw from the study at any time without any consequences. No personally identifying information was collected.

The study was conducted in accordance with the Declaration of Helsinki and institutional ethical guidelines and received approval from the Institutional Review Board (Ethics Committee) of the University of Szeged Doctoral School of Education (approval number: 2021/5; approval date: 9 March 2022).

3. Results

3.1. Differences in Social Problem-Solving According to Age and Sex

First, we examined whether social problem-solving differed as a function of age and the nature of the friendship (same-sex or opposite-sex friendships). The multivariate test revealed a significant effect of age (Wilks’ Λ = 0.867, F(10, 137) = 2.11, p = 0.028), suggesting differences between younger and older adolescents. However, univariate analyses (

Table 2. Age differences according to SPSI–R factors in same-sex and opposite-sex friends (N = 304).

SPSI–R Factor	12-Year-Olds M (SD)	16-Year-Olds M (SD)	F(1, 146)	p	η2
Same-sex friends
Negative orientation	1.26 (0.88)	1.40 (0.91)	4.92	0.028	0.006
Positive orientation	2.41 (0.65)	2.64 (0.55)	1.70	0.194	0.012
Rationality	2.62 (0.87)	2.71 (0.80)	0.08	0.782	<0.001
Impulsivity	1.81 (0.67)	1.65 (0.77)	0.34	0.561	0.002
Avoidance	1.26 (1.00)	1.13 (0.93)	0.00	0.974	<0.001
Opposite-sex friends
Negative orientation	1.08 (0.87)	1.46 (0.95)	5.69	0.018	0.040
Positive orientation	2.45 (0.75)	2.58 (0.65)	1.97	0.162	0.013
Rationality	2.53 (1.00)	2.75 (0.91)	1.66	0.200	0.011
Impulsivity	1.73 (0.80)	1.68 (0.81)	0.15	0.698	0.001
Avoidance	1.02 (1.01)	1.15 (1.09)	0.35	0.553	0.002

) showed that this age-related difference was limited to negative orientation, even though, based on another Hungarian study—both cross-sectional and longitudinal (Kasik, 2015)—not only negative orientation but also impulsivity and avoidance increase with age. According to this study, negative orientation is characteristic of older adolescents in both same-sex and opposite-sex friendships. There were no significant age differences in positive orientation or in the styles (rational, impulsive, and avoidant). Based on a multivariate analysis of context sensitivity (variables measuring differences between same-sex and opposite-sex friends), there was no significant age effect (Wilks’ Λ = 0.945, F(5, 142) = 1.65, p = 0.150), suggesting that the extent to which adolescents distinguish between same-sex and opposite-sex friendship situations remains relatively stable across age groups.

Among 12-year-olds, the multivariate effect of sex was not significant (Wilks’ Λ = 0.838, F(10, 39) = 0.76, p = 0.669). Based on univariate analyses (

Table 3. Sex characteristics based on SPSI–R among 12- (n = 138) and 16-year-olds (n = 166).

SPSI–R Factor	Boys M (SD)	Girls M (SD)	F(1, df2)	p	η2
Same-sex friends (12-year-olds)
Negative orientation	1.12 (0.88)	1.39 (0.87)	2.73 (1, 108)	0.101	0.025
Positive orientation	2.52 (0.66)	2.31 (0.63)	2.66 (1, 101)	0.106	0.026
Rationality	2.55 (0.93)	2.69 (0.81)	0.65 (1, 110)	0.422	0.006
Impulsivity	1.79 (0.77)	1.83 (0.56)	0.12 (1, 110)	0.735	0.001
Avoidance	1.43 (1.10)	1.08 (0.86)	3.53 (1, 107)	0.063	0.032
Opposite-sex friends (12-year-olds)
Negative orientation	0.85 (0.82)	1.29 (0.87)	4.89 (1, 70)	0.030	0.065
Positive orientation	2.43 (0.85)	2.48 (0.66)	0.07 (1, 66)	0.799	0.001
Rationality	2.54 (1.06)	2.53 (0.97)	0.00 (1, 72)	0.982	<0.001
Impulsivity	1.63 (0.97)	1.83 (0.56)	1.17 (1, 76)	0.282	0.015
Avoidance	1.05 (1.07)	0.95 (0.97)	0.19 (1, 70)	0.663	0.003
Same-sex friends (16-year-olds)
Negative orientation	1.28 (0.89)	1.48 (0.91)	1.93 (1, 171)	0.167	0.011
Positive orientation	2.68 (0.46)	2.61 (0.60)	0.70 (1, 165)	0.403	0.004
Rationality	2.67 (0.77)	2.72 (0.82)	0.16 (1, 170)	0.691	0.001
Impulsivity	1.76 (0.80)	1.57 (0.73)	2.77 (1, 175)	0.098	0.016
Avoidance	1.28 (0.99)	1.04 (0.88)	2.87 (1, 172)	0.092	0.016
Opposite-sex friends (16-year-olds)
Negative orientation	1.48 (0.93)	1.47 (0.98)	0.00 (1, 122)	0.971	<0.001
Positive orientation	2.58 (0.63)	2.58 (0.67)	0.00 (1, 124)	0.956	<0.001
Rationality	2.65 (0.88)	2.79 (0.92)	0.67 (1, 127)	0.414	0.005
Impulsivity	1.76 (0.84)	1.62 (0.79)	1.01 (1, 127)	0.317	0.008
Avoidance	1.45 (1.18)	0.93 (0.98)	7.25 (1, 124)	0.008	0.055

), girls showed a higher negative orientation only in opposite-sex friendships. There were no sex differences in the other factors of social problem-solving. Among 16-year-olds, the effect of sex in the multivariate analysis was also not significant (Wilks’ Λ = 0.856, F(10, 86) = 1.45, p = 0.174). The univariate analyses (Table 3) identified only one significant difference: avoidance was more characteristic of boys in the case of friendships with the opposite sex. There were no further sex differences.

3.2. Differences in Friendship Functions According to Age and Sex

Multivariate analysis of friendship functions showed a significant age effect (Wilks’ Λ = 0.628, F(12, 89) = 4.40, p < 0.001), suggesting that 12-year-olds and 16-year-olds judge friendship functions differently. Based on univariate analyses (

Table 4. Age differences according to MFQ factors in same-sex and opposite-sex friends (N = 304).

MFQ Factor	12-Year-Olds M (SD)	16-Year-Olds M (SD)	F(1, 100)	p	η2
Same-sex friends
Stimulating companionship	7.15 (0.79)	7.15 (0.79)	0.01	0.933	<0.001
Help	6.68 (1.10)	6.68 (1.10)	1.58	0.212	0.016
Intimacy	6.95 (1.20)	6.95 (1.20)	0.00	0.994	<0.001
Reliable alliance	7.46 (0.79)	7.46 (0.79)	0.07	0.797	0.001
Self-validation	6.12 (1.22)	6.12 (1.22)	2.23	0.138	0.022
Emotional security	6.27 (1.39)	6.27 (1.39)	1.75	0.189	0.017
Opposite-sex friends
Stimulating companionship	6.95 (1.20)	6.95 (1.20)	0.29	0.594	0.003
Help	6.36 (1.39)	6.36 (1.39)	0.05	0.817	0.001
Intimacy	6.46 (1.63)	6.95 (1.20)	8.31	0.005	0.077
Reliable alliance	6.86 (1.41)	7.46 (0.79)	2.82	0.096	0.027
Self-validation	6.09 (1.43)	6.12 (1.22)	4.09	0.046	0.039
Emotional security	6.19 (1.64)	6.27 (1.39)	0.84	0.360	0.008

), there was a significant difference in intimacy and self-validation, with 16-year-olds scoring higher on both factors. In the case of same-sex friendship, there was no significant difference across functions.

Among 12-year-olds, the multivariate profile of friendship functions differed significantly by sex (Wilks’ Λ = 0.496, F(12, 36) = 3.05, p = 0.005). According to univariate analyses (

Table 5. Sex characteristics based on MFQ among 12- (n = 138) and 16-year-olds (n = 166).

MFQ Factor	Boys M (SD)	Girls M (SD)	F(1, df2)	p	η2
Same-sex friends (12-year-olds)			F(1, 47)
Stimulating companionship	6.90 (0.96)	7.31 (0.75)	3.05	0.087	0.061
Help	6.19 (1.21)	7.00 (0.98)	12.97	<0.001	0.216
Intimacy	6.46 (1.58)	7.33 (0.96)	7.64	0.008	0.140
Reliable alliance	7.22 (1.05)	7.52 (0.71)	0.04	0.835	0.001
Self-validation	5.60 (1.32)	6.10 (1.28)	0.17	0.684	0.004
Emotional security	5.69 (1.54)	6.96 (1.07)	17.97	<0.001	0.277
Opposite-sex friends (12-year-olds)			F(1, 47)
Stimulating companionship	6.72 (1.32)	7.16 (0.85)	1.58	0.215	0.032
Help	6.23 (1.48)	6.52 (1.30)	0.14	0.707	0.003
Intimacy	5.93 (1.91)	6.27 (1.90)	1.02	0.317	0.021
Reliable alliance	6.40 (1.59)	6.89 (1.34)	0.40	0.529	0.008
Self-validation	5.79 (1.68)	5.63 (1.56)	0.37	0.543	0.008
Emotional security	5.98 (1.65)	6.04 (1.90)	0.06	0.809	0.001
Same-sex friends (16-year-olds)			F(1, df2)
Stimulating companionship	6.94 (0.85)	7.33 (0.58)	13.77 (1, 181)	<0.01	0.071
Help	6.40 (1.27)	6.96 (0.78)	12.85 (1, 173)	<0.001	0.069
Intimacy	6.65 (1.24)	7.22 (0.85)	13.43 (1, 177)	<0.001	0.071
Reliable alliance	7.31 (0.81)	7.66 (0.54)	12.26 (1, 173)	0.298	0.066
Self-validation	5.88 (1.12)	6.59 (1.00)	19.24 (1, 171)	<0.001	0.101
Emotional security	5.68 (1.49)	6.66 (1.05)	26.62 (1, 177)	<0.001	0.131
Opposite-sex friends (16-year-olds)			F(1, df2)
Stimulating companionship	6.45 (1.51)	7.34 (0.81)	19.60 (1, 136)	<0.001	0.126
Help	6.23 (1.57)	6.43 (1.26)	0.67 (1, 128)	0.414	0.005
Intimacy	6.50 (1.64)	6.76 (1.26)	1.02 (1, 134)	0.313	0.008
Reliable alliance	6.55 (1.63)	7.31 (0.96)	11.41 (1, 133)	0.001	0.079
Self-validation	6.22 (1.48)	6.37 (1.12)	0.46 (1, 130)	0.498	0.004
Emotional security	6.15 (1.51)	6.63 (1.36)	1.90 (1, 67)	0.172	0.028

), girls scored higher on the factors of help, intimacy, and emotional security in same-sex friendships. In opposite-sex friendships, there was no significant difference by sex across the functions. Among 16-year-olds, sex differences showed an even more pronounced multivariate effect (Wilks’ Λ = 0.458, F(12, 39) = 3.85, p < 0.001). Based on univariate analyses (Table 5), girls achieved higher scores in almost all functions in same-sex friendships, with reliable alliance the only factor where there was no significant difference. In opposite-sex friendships, girls also achieved significantly higher scores on the stimulating companionship and reliable alliance factors.

3.3. Friendship Profiles by Age and Friendship Type

Using LPA, we created profiles for both age groups based on the functions of friendship and friendship type (same-sex and opposite-sex). For 12-year-olds (

Table 6. Profiles according to friendship functions (M)—12-year-olds.

MFQ Factor	Profile 1	Profile 2	Profile 3	Profile 4
Same-sex friends (n = 89)	n = 29	n = 34	n = 26	-
Stimulating companionship	7.81	7.25	6.42
Help	7.67	7.04	5.37
Intimacy	7.78	7.39	5.48
Reliable alliance	7.96	7.45	7.08
Self-validation	6.92	5.98	4.74
Emotional security	7.73	6.70	4.97
Opposite-sex friends (n = 62)	n = 20	n = 8	n = 19	n = 15
Stimulating companionship	7.51	7.98	6.79	5.85
Help	7.50	7.80	5.96	4.65
Intimacy	7.45	7.98	6.13	3.24
Reliable alliance	7.39	7.95	6.63	5.07
Self-validation	6.91	7.65	5.40	4.08
Emotional security	7.47	8.00	5.58	3.65

), three profiles emerged regarding same-sex friendships (Log-likelihood = −524.53, BIC = −1219.62, ICL = −1225.48). In the case of highly supportive, emotionally intense friendships (Profile 1), the roles of intimacy, reliable alliance, and emotional security were prominent. In the case of balanced, less intense friendships (Profile 2), the values for all dimensions were relatively high but more moderate than in Profile 1, and the levels of self-validation and emotional security were significantly lower than in Profile 1. Functional, less intimate friendships (Profile 3) were characterized by low emotional security, self-validation, and intimacy, but reliable alliance was also relatively high here.

In the case of 12-year-olds’ friendships with the opposite sex (Table 6), four profiles emerged (Log-likelihood = −456.98, BIC = −1041.90, ICL = −1046.13), which indicates greater variability in friendship patterns than in same-sex friendships. In the case of balanced, emotionally secure friendships (Profile 1), all values were high but not extreme. In the case of emotionally extremely intense friendships (Profile 2), all values were extremely high across the dimensions of friendship. Since we did not examine romantic commitment, these profiles cannot be definitively considered evidence of a romantic transition. In the case of less intimate friendships (Profile 3), values for all dimensions were moderate. In the case of functional, less intimate friendships (Profile 4), emotional security and intimacy were low, as in same-sex friendships (see Profile 3 above).

Among 16-year-olds (

Table 7. Profiles according to friendship functions (M)—16-year-olds.

MFQ Factor	Profile 1	Profile 2	Profile 3
Same-sex friends (n = 159)	n = 83	n = 36	n = 40
Stimulating companionship	7.35	7.73	6.41
Help	6.98	7.61	5.38
Intimacy	7.39	7.79	5.67
Reliable alliance	7.62	7.94	6.94
Self-validation	6.45	7.36	5.14
Emotional security	6.54	7.65	4.38
Opposite-sex friends (n = 60)	n = 22	n = 29	n = 9
Stimulating companionship	7.55	6.90	5.36
Help	7.50	6.02	4.60
Intimacy	7.69	6.90	4.78
Reliable alliance	7.77	7.08	5.49
Self-validation	7.37	6.15	4.33
Emotional security	7.64	6.10	4.02

), three profiles emerged for same-sex friendships (Log-likelihood = −959.20, BIC = −2060.32, ICL = −2076.27). In the case of balanced, stable friendships (Profile 1), all values were high but not extreme. This profile represented the most common friendship pattern in this age group, and its proportion was lower among 12-year-olds. In the case of close, supportive friendships (Profile 2), all values were high, particularly closeness and emotional security. Less intimate friendships (Profile 3) were characterized by low levels of help, intimacy, and emotional security, as was also observed in the case of opposite-sex friendships among 12-year-olds (Profile 3).

Among 16-year-olds (Table 7), three profiles emerged for opposite-sex friendships (Log-likelihood = −461.43, BIC = −1017.03, ICL = −1019.61). For close, emotionally secure friendships (Profile 1), all values were high but not extremely high. For moderately intimate friendships (Profile 2), all values were moderate or high. Less intimate friendships (Profile 3) were characterized by low levels of help, self-validation, and emotional security, as were same-sex friendships (Profile 3) and, among younger students, opposite-sex friendships (Profile 3).

3.4. Social Problem-Solving Profiles by Age and Friendship Type

In the case of same-sex friendships among 12-year-olds (

Table 8. Characteristics of social problem-solving by factor (M)—12- and 16-year-olds.

SPSI–R Factor	Profile 1	Profile 2
Same-sex friends (n = 59)	n = 42	n = 17
Negative orientation	1.49	0.51
Positive orientation	2.21	2.99
Rationality	2.50	3.10
Impulsivity	1.90	1.46
Avoidance	1.60	0.17
Opposite-sex friends (n = 59)	n = 42	n = 17
Negative orientation	1.30	0.58
Positive orientation	2.18	3.04
Rationality	2.07	3.54
Impulsivity	1.87	1.31
Avoidance	1.35	0.09
Same-sex friends (n = 151)	n = 38	n = 113
Negative orientation	2.24	1.07
Positive orientation	2.39	2.72
Rationality	2.85	2.68
Impulsivity	1.76	1.58
Avoidance	2.35	0.65
Opposite-sex friends (n = 113)	M	–
Negative orientation	1.47
Positive orientation	2.56
Rationality	2.70
Impulsivity	1.68
Avoidance	1.17

), two profiles emerged (Log-likelihood = −527.46, BIC = −1150.76, ICL = −1154.21). Less adaptive, emotionally charged problem-solvers (Profile 1) were characterized by low positive orientation, but high negative orientation, impulsivity, and avoidance. Adaptive problem-solvers (Profile 2) were characterized by high positive orientation and rationality, but low negative orientation and impulsivity, and very low avoidance.

In the case of friendships with the opposite sex among 12-year-olds (Table 8), two profiles emerged (Log-likelihood = −328.02, BIC = −741.66, ICL = −745.23). Less adaptive problem-solvers (Profile 1) could also be distinguished, who were characterized by high negative orientation, impulsivity, and avoidance, as well as low rationality. Highly adaptive problem-solvers (Profile 2) were characterized by high positive orientation, very high rationality, low negative orientation and impulsivity, and extremely low avoidance.

In the case of same-sex friendships among 16-year-olds (Table 8), two profiles were also distinguished (Log-likelihood = −767.86, BIC = −1671.18, ICL = −1682.38). Negatively oriented and avoidant problem-solvers (Profile 1) were characterized by very high negative orientation and avoidance, less high impulsivity, and low rationality. Adaptive problem-solvers (Profile 2) were characterized by high positive orientation, moderately high rationality, moderate impulsivity, and low negative orientation and avoidance.

In the case of friendships with the opposite sex among 16-year-olds (Table 8), we did not identify separate profiles, and the factors did not differentiate in cluster structure (Log-likelihood = −655.34, BIC = −1405.24). Students were characterized by lower negative orientation, impulsivity, and avoidance, as well as higher positive orientation and rationality, which can be considered adaptive when compared to previous profiles.

3.5. The Relationship Between Friendship Functions and Social Problem-Solving Profiles

We performed a cross-tabulation analysis to explore the relationship between the profiles (

Table 9. Cross-tabulation of friendship functions and profiles according to social problem-solving (N = 258).

MFQ Profile	SPSI–R Profile 1 (Less Adaptive)	SPSI–R Profile 2 (Adaptive)
12-years-old, same-sex friends
1: highly supportive, emotionally intense	15	7
2: balanced, less intense	27	4
3: functional, less intimate	14	7
12-years-old, opposite-sex friends
1: balanced, emotionally secure	9	5
2: idealized	5	1
3: less intimate	12	4
4: functional, less intimate	10	3
16-years-old, same-sex friends
1: balance, stable	18	51
2: intimate, supportive	9	21
3: less intimate	8	29
12-years-old, opposite-sex friends
1: intimate, emotionally secure	-	-
2: moderately intimate	-	-
3: less intimate	-	-

). Among 12-year-olds, there was no significant relationship between the two patterns in the case of same-sex friendships (χ²(2) = 3.79, p = 0.15), but there was a tendency for MFQ Profile 2 (balanced, less intense) to be mainly found in SPSI–R Profile 1 (less adaptive), and for MFQ Profile 1 (highly supportive, emotionally intense) and MFQ Profile 3 (functional, less intimate). In the case of opposite-sex friendships among 12-year-olds, there was also no significant relationship between the patterns (χ²(2) = 1.00, p = 0.80)—due to the small sample size, the χ² approximation was not stable. The differentiation of friendship functions does not structure social problem-solving. Among 16-year-olds, there was also no significant relationship between friendship functions and patterns according to social problem-solving in the case of same-sex friendships (χ²(2) = 0.75, p = 0.69). The proportion of SPSI–R Profile 2 (adaptive) was similar in all MFQ profiles (~70–75%), meaning that the structure of friendship does not directly predict belonging to the social problem-solving profile in this case either.

Since we did not identify any SPSI–R patterns in the case of opposite-sex friendships among 16-year-olds, we conducted a factor analysis (dimensionality) in this group, which showed that the MFQ factors had no significant multivariate effect on the SPSI–R factors (Wilks’ Λ = 0.747, F(10, 80) = 1.26, p = 0.268), meaning that there was no difference in the social problem-solving pattern between the three MFQ profiles. Based on the univariate results, none of the SPSI–R factors differed significantly between the MFQ profiles (negative orientation: F(2, 44) = 1.37, p = 0.266, η² = 0.06; positive orientation: F(2, 44) = 2.05, p = 0.140, η² = 0.09; rationality: F(2, 44) = 0.06, p = 0.938, η² = 0.003; impulsivity: F(2, 44) = 0.16, p = 0.856, η² = 0.007; avoidance: F(2, 44) = 2.17, p = 0.126, η² = 0.09). A moderate effect can be seen in the case of positive orientation and avoidance, but this was not significant either. Rationality was practically completely independent of the patterns according to friendship functions.

We examined whether there were differences in the structure of the relationships between the patterns across age groups, using SEM. In the initial analysis plan, we estimated a full structural model in which the six factors of friendship simultaneously predicted the five social problem-solving factors. However, this model contained 45 free parameters, while the number of items in the younger samples was low (same sex: n = 74; opposite sex: n = 49). The model had a degree of freedom of 0 (df = 0), resulting in a saturated model. As a result, the fit indices (CFI = 1.00; RMSEA = 0.00; SRMR = 0.00) are not interpretable, as the model perfectly reproduced the covariance matrix.

So, since we did not obtain a model for 12-year-olds—neither for same-sex nor opposite-sex friends—we used a reduced latent model. We modeled the six factors of the MFQ as indicators of a single latent “friendship function” factor, while the five SPSI–R factors were modeled as indicators of a single latent “problem-solving” factor. In the structural model, “problem-solving” was regressed on the “friendship function” factor. We examined the relationship in several steps: (1) multigroup model by age (same-sex friendship), (2) multigroup model by age (opposite-sex friendship), and (3) competing structural model (both friendship types together).

In the case of same-sex friendships, the model showed an excellent fit (χ²(2) = 0.18, p = 0.913, CFI = 1.00, TLI = 1.16, RMSEA = 0.00, SRMR = 0.009). The two structural paths (adaptive problem-solving: β (12-years-old) = 0.26, β (16-years-old) = 0.27, p < 0.001; non-adaptive problem-solving: β (12-years-old) = −0.26, β (16-years-old) = −0.21, p < 0.001) did not differ by age. Based on these findings, emotionally stable, supportive same-sex friendships were associated with higher adaptive and lower non-adaptive problem-solving at both ages.

In the case of opposite-sex friendships, the model showed an acceptable fit, but weaker than the model obtained for same-sex friendships (χ²(2) = 4.31, p = 0.116, CFI = 0.882, robust CFI = 0.900, RMSEA = 0.10, SRMR = 0.054). The two structural paths (adaptive problem-solving: β (12-years-old) = 0.22, β (16-years-old) = 0.24, p = 0.001; non-adaptive problem-solving: β (12-years-old) = −0.24, β (16-years-old) = −0.20, p = 0.003) did not differ according to age. In opposite-sex friendships, the relationship between friendship functions and problem-solving was the same as in same-sex friendships, but the effect was weaker.

In the case of the competing structural model, we examined the relative predictive power by simultaneously including both types of friendship. In the case of adaptive problem-solving, β = 0.219 (p = 0.001) for same-sex friendships and β = 0.091 (p = 0.167) for opposite-sex friendships. In the case of non-adaptive problem-solving, β = −0.2168 (p = 0.025) for same-sex friendships and β = −0.129 (p = 0.106) for opposite-sex friendships. The correlation between the predictors (same-sex and opposite-sex friendships) was moderate (r = 0.497, p < 0.001).

4. Discussion

The aim of our study was to examine the characteristics of friendship functions measured by the MFQ (stimulating companionship, help, intimacy, reliable alliance, self-validation, and emotional security) and the social problem-solving dimensions measured by the SPSI–R (negative and positive orientation, rationality, impulsivity, and avoidance) among 12- and 16-year-old Hungarian students, how these friendship functions relate to social problem-solving, and whether these relationships differ between same-sex and opposite-sex friendships. The results partially supported the hypotheses and, in many cases, provided a more nuanced picture of the relationship between friendship functions and social problem-solving than we had originally anticipated.

Based on the normality test, the distributions of most factors met the conditions for parametric statistical analysis. Among the friendship functions, we observed a slight negative skew for persistence and stimulating companionship, while we observed a slight ceiling effect for reliable alliance and inspiring companionship, suggesting that most students rated these characteristics the highest. This ceiling effect is not surprising in the case of adolescent friendships, as young people generally view their close friends in an extremely positive light (Bukowski et al., 1994; Mendelson & Aboud, 2014). Previous Hungarian studies have also revealed a similar trend, indicating that Hungarian adolescents primarily describe friendships as supportive, security-providing relationships (Zsuzsa, 2004; Pikó, 1996). However, the slight ceiling effect may also suggest that the functional dimensions of friendship appear as normative expectations for adolescents, which is presumably universal in nature and not a Hungarian peculiarity. At the same time, these ceiling effects may have reduced variability in certain friendship dimensions, implying that the quality of the relationship is implicitly valued. This is consistent with the need satisfaction model (Helgeson & Lopez, 2010; Jordan, 2013), which posits that friendship is one of the primary contexts for the satisfaction of basic psychological needs—such as belonging, support, and emotional security—which, compared to the family, may be important during adolescence in terms of striving for autonomy.

Multivariate analysis revealed a significant effect of age, suggesting that the problem-solving characteristics of 12- and 16-year-old adolescents differ from one another overall. However, univariate results suggested that these differences were primarily limited to negative orientation: older adolescents exhibited higher negative orientation with both same-sex and opposite-sex friends. This finding supports the hypothesis that negative orientation is stronger during middle adolescence (Kasik, 2015). This phenomenon aligns well with Hungarian and international cross-sectional and longitudinal studies that have examined the characteristics of social problem-solving (D’Zurilla et al., 2002; Kasik, 2015; Kasik et al., 2026), which suggest that negative orientation—which involves a sense of threat posed by problems and lower self-efficacy in problem-solving—increases during adolescence. In mid-adolescence, self-reflection, dependence on social evaluation, and sensitivity to negative emotions intensify (Steinberg, 2014), leading to problematic situations being interpreted more frequently as threatening or stressful. For the other SPSI–R factors, we found no significant age-related differences. Therefore, the results do not support widespread developmental changes in social problem-solving during adolescence—this may also stem from the composition of the sample.

Although multivariate analyses did not reveal any significant sex differences in any age group, univariate results pointed to certain specific differences in the context of social problem-solving. Among 12-year-olds, girls showed a greater negative orientation in the context of friendships with the opposite sex, while among 16-year-olds, avoidance was more characteristic of boys in friendships with the opposite sex. This is partly consistent with previous research indicating that girls respond to negative interpersonal situations with stronger emotional involvement, while boys more frequently avoid problematic situations (Zsuzsa, 2004; Kasik et al., 2026; Rose & Rudolph, 2006). Overall, the results suggest that sex differences in adolescent social problem-solving are not universal but may depend on specific friendship contexts. Thus, sex differences were not universal across all dimensions but emerged only in specific friendship contexts.

Regarding friendship functions, multivariate analysis revealed a significant difference based on age, whereas univariate results showed differences only in the case of friendships between opposite sexes. In this context, older adolescents identified higher levels of intimacy and self-validation as important functions. This is consistent with Selman’s (1981) developmental theory, which posits that as adolescence progresses, friendships increasingly rely on reciprocity and emotional support. At the same time, no age differences emerged in same-sex friendships, from which we can conclude, among other things, that these relationships already possess a relatively stable functional structure at the onset of adolescence, which continues to strengthen throughout adolescence. At the same time, there was no age difference in same-sex friendships, suggesting that the perceived functions of these relationships remain relatively stable throughout adolescence. Sex differences were more pronounced in friendship functions than in social problem-solving. Among 12-year-olds, girls scored higher in same-sex friendships on the factors of help, intimacy, and emotional security. Among 16-year-olds, these differences were even more pronounced: except for reliable alliance, girls attributed higher values to same-sex friendships across all functions. This finding supports research indicating that friendships among girls are generally more emotionally intense and support-oriented (Bukowski et al., 1994; Mendelson & Aboud, 2014).

Based on the results of latent profile analysis (LPA), distinct patterns emerged regarding the functions of friendships. In both age groups, we identified several profiles that reflect friendships with different functions, and two of these (functional, less intimate; less intimate) appeared across multiple age groups and friendship types. Among 12-year-olds, same-sex friendships can be classified into three profiles, while opposite-sex friendships into four; that is, it is possible that younger adolescents’ opposite-sex relationships are more heterogeneous. This greater heterogeneity is consistent with the developmental process in which opposite-sex friendships are even more unstable and less structured, and less stable in early adolescence (Furman & Rose, 2015; Zimmermann, 2004). Among 12-year-olds, one profile of opposite-sex friendships showed extremely high scores across all dimensions, which likely indicates idealized or even romantic relationships, particularly in their initial stages, which may reflect idealized or emotionally intense relationships. However, in the present study, we did not examine romantic relationships directly; therefore, other explanations must also be considered, such as socially desirable behavior, the highlighting of traits advantageous to the student, and selective self-presentation, but it is worth accounting for statistical imprecision resulting from the small sample size (Collins et al., 2009).

Among 16-year-olds, the structure of the profiles was more stable: in the case of same-sex friendships, a dominant, balanced profile emerged, suggesting that friendships become increasingly structured by the middle of adolescence (though this, of course, also depends on how long the friendship has lasted). Regarding social problem-solving, two basic profiles emerged in both age groups: an adaptive and a less adaptive pattern. The adaptive profile is characterized by high positive orientation and rationality, as well as low negative orientation and avoidance, while the less adaptive profile showed the exact opposite pattern. This structure fits well with the model described by D’Zurilla et al. (2002), although there are Hungarian studies (Fejes et al., 2023; Kasik et al., 2026) that have revealed more complex patterns (e.g., high positive orientation, rationality, and associated avoidance)—this undoubtedly depends largely on the composition of the sample. Although it might be assumed that more adaptive problem-solving is associated with high-functionality friendship profiles, cross-tabulation analyses revealed no significant relationship between friendship profiles and problem-solving profiles among either 12-year-olds or 16-year-olds. One possible explanation for this is that the structural pattern of friendship does not in itself determine problem-solving. Thus, the results do not support a one-to-one correspondence between friendship-functional profiles and adaptive or less adaptive problem-solving profiles. The relationship between the two constructs can be viewed as dimensional rather than categorical. Instead, friendship function profiles and social problem-solving profiles appear to reflect partially independent developmental dimensions. This is consistent with the model proposed by Mendelson and Aboud (2014), according to which friendship functions operate more as a mutually reinforcing system than as independent configurations.

At the same time, the results of the structural model revealed a link between friendship and social problem-solving. For same-sex friendships, the model showed an excellent fit, and the functions of friendship were significantly and positively associated with adaptive problem-solving, while they were negatively associated with less adaptive problem-solving. For opposite-sex friendships, we found a similar but weaker relationship, suggesting that the quality of friendship does indeed play a role in adaptive problem-solving.

These results suggest parallel associations between the perceived quality of friendship and dimensions of social problem-solving, while also indicating that this effect is primarily mediated by same-sex friendships. The functions of same-sex friendships showed stronger, more unique associations with adaptive and maladaptive problem-solving than the functions of opposite-sex friendships in this sample. This finding is consistent with the friendship needs satisfaction model (Jordan, 2013) and Rose et al.’s (2016) family socialization approach, which posits that secure, reciprocal relationships promote constructive problem-solving. A particularly important finding is that when same-sex and opposite-sex friendships were considered simultaneously in the model, only same-sex friendships remained as significant predictors. This suggests that same-sex friendships may play a role in socialization during adolescence in terms of the development of problem-solving skills.

When we included both same-sex and opposite-sex friendships simultaneously in the SEM model, only same-sex friendships remained significant predictors. However, we believe this should be interpreted with caution as evidence of developmental primacy or mediation; rather, it suggests that there are stronger, unique statistical correlations in this sample. This finding is consistent with Selman’s (1981) developmental model and Zimmermann’s (2004) findings regarding emotional security. The stability across age suggests that, among 12- to 16-year-olds, the relationship between friendship and social problem-solving does not undergo a sudden transformation but is characterized by continuity. It can be assumed that this continuity also represents a form of stability during adolescence, although longitudinal studies beginning in early adolescence are needed to confirm this. The similarity of SEM relationships across age groups suggests relative continuity in the relationship between friendship functions and social problem-solving during adolescence. However, due to the cross-sectional design, no conclusions can be drawn regarding developmental trajectories or causal relationships. Longitudinal studies are needed to clarify whether supportive friendship experiences contribute to adaptive social problem-solving, whether adaptive problem-solvers form higher-quality friendships, or whether these processes mutually reinforce each other during adolescence.

Limitations

Several limitations should be considered when interpreting the results of our study. First, the study used a cross-sectional design; therefore, no causal or developmental conclusions can be drawn regarding the relationship between friendship functions and social problem-solving. The results only indicate associations between adolescents’ perceived friendship characteristics and their self-reported social problem-solving. Longitudinal studies are needed to clarify the developmental direction of these associations.

All data were based on self-report questionnaires completed in a school context. Although anonymity and voluntary participation were emphasized, and teachers received preparatory guidance before supervising the administration process, the presence of teachers may nevertheless have influenced adolescents’ responses, particularly regarding emotionally sensitive dimensions, such as impulsivity, avoidance, intimacy, or emotional insecurity. Social desirability and context effects, therefore, cannot be excluded.

Several subgroup analyses were based on relatively small sample sizes, especially in the case of opposite-sex friendships. Consequently, some latent profile solutions may be unstable and should be interpreted as exploratory patterns rather than clearly established developmental categories. This is particularly important because the profile-level associations between friendship functions and social problem-solving were mostly non-significant. Although the proportions of these subgroups provide a good illustration of the prevalence of cross-sex friendships, it would be worthwhile to conduct analyses on a larger sample in the future.

Another important limitation is that no information was collected about the duration, stability, closeness, or “best friend” status of the friendships. As previous research suggests (Mendelson & Aboud, 1999, 2014; Selman, 1981), friendship functions, such as reliable alliance, emotional security, intimacy, and self-validation, may differ substantially between newly formed and long-term friendships. Without information about friendship duration or stability, it is difficult to determine whether MFQ scores reflect mature friendship functions or more generalized positive perceptions of peer relationships.

This limitation is particularly important when interpreting friendships between people of different genders. Some highly idealized friendship profiles may reflect emotionally intense friendships, a selective self-image, or potential romantic interest. However, we did not measure romantic relationships directly; therefore, other possible explanations must also be considered in future research.

Another methodological limitation is that measurement invariance across age groups and friendship types was not examined. Consequently, it cannot be established with complete certainty whether younger and older adolescents, or adolescents reporting friendships with the opposite sex versus same-sex friends, interpreted the questionnaire dimensions in the same way. Measurement invariance should be examined in future studies.

Finally, the generalizability of the findings is limited. The sample consisted exclusively of Hungarian students from a specific educational and sociocultural context, and the schools were not selected to form a nationally representative sample. Friendship norms, emotional expression, and interpersonal problem-solving may vary across cultures and school environments (D’Zurilla et al., 2004; Mendelson & Aboud, 2014); therefore, replication in different sociocultural contexts is necessary.