1. Introduction
The onset of puberty marks the initiation of adolescence, a phase entailing multifaceted changes across biological, social, and psychological dimensions. The developmental significance of different stages of adolescence becomes evident when the child’s age is taken into account. These stages are categorized as early adolescence (from 10 to 13 years), middle adolescence (from 14 to 16 years), and late adolescence (from 17 to 20 years) [
1].
Sexual maturation precipitates notable physical and neurological advancements, serving as a prelude to shifts in self-perception and social conduct [
1]. Relationships with parents and friends undergo noticeable changes during this period [
2,
3]. Fundamental to the adolescent experience is the development of autonomy and independence, involving negotiations for privacy and the emancipation from parental oversight [
4,
5]. As proposed by the separation–individuation theory, the hormonal changes accompanying puberty drive adolescents toward detachment from their parents in the pursuit of self-reliance [
6]. From the viewpoint of adolescents, there is a significant decline in parental authority from early to mid-adolescence [
7,
8], fostering a more egalitarian parent–adolescent relationship [
7]. A majority of adolescents aged 12 to 20 report a harmonious parent–child relationship characterized by substantial support, minimal power dynamics, and infrequent negative interactions [
8].
Nonetheless, a higher incidence of turbulent relationships, marked by reduced support, elevated power dynamics, and negative interactions, becomes evident in early as opposed to late adolescence. Further studies underscore the challenges associated with transitioning from early to mid-adolescence, which are mirrored in communication patterns. In the initial phases of adolescence, it is commonly observed that most adolescents experience a decline in their willingness to share information, a decrease in the parental solicitation, and a heightened inclination toward maintaining secrecy [
9]. However, as adolescence progresses into the middle phase and beyond, a shift towards a more open form of communication between parents and their adolescent children becomes evident. When analyzing parent–adolescent relationships, two key aspects of communication merit consideration: closeness, fostering connections despite individual changes, and conflict, inducing psychological and physical distance [
4]. A range of theories, encompassing neo-psychoanalytic, evolutionary, and socio-cognitive perspectives, posit that the burgeoning autonomy and individuation during adolescence result in a temporary reduction in closeness and an increase in conflicts [
5]. In addition to this knowledge regarding the changes in conflicts and closeness during adolescence, there is a lack of insight into how these two aspects are interrelated.
In the subsequent sections, the state of the art concerning the development of closeness, conflict frequency, and intensity during adolescence will be delved into more extensively. It will also be shown that there is a lack of insight into how changing conflicts and closeness are interrelated over the course of adolescence and in comparison between the children’s as well as the parent’s perspectives. The current study aims to address these two research gaps in order to give more insight into the complex inner familial processes during adolescence and provide knowledge for family research and family practitioners.
1.1. Closeness during Adolescence
Closeness in social relationships describes “the extent to which two individuals are connected behaviorally and emotionally. Commonly invoked indicators include interdependence, intimacy, support, trust, and communication” [
5] (p. 18). In the context of developmental progression, the nature of closeness experienced during adolescence takes on distinct manifestations compared to the closeness observed in earlier parent–child interactions. The level of intimacy between parents and children, evident through physical affection and comprehensive shared engagements, diminishes as offspring progress in their development, while instances of dialogues facilitating the exchange of information and the articulation of emotions become more prevalent [
10].
Developmental transformations concerning closeness have been comprehensively documented, particularly during the transition from early to middle adolescence. Substantial reductions are observed in dimensions like perceived parental support and relational connectedness between parents and their offspring [
7,
11,
12,
13]. Interactions between children and parents experience notable changes, leading to a decrease in parent–child activities and an increase in peer interactions [
14], as evidenced by a decline in communal meals among older children compared to their younger counterparts [
15]. In the transition to adolescence, there is a noticeable increase in withholding personal information from parents, especially when compared to other aspects of their lives, such as schoolwork [
16]. This concealment is bidirectionally linked to a decline in parent–child relationship quality, including communication and trust, and an increase in depressive symptoms over time [
17]. The diminished closeness between parents and adolescents may be a risk factor for the development of internalizing problems [
18,
19] and initiation of delinquency, tobacco use, and polysubstance use [
20]. These changes often reflect a diminishing reliance on parents. However, they do not necessarily undermine the positive attributes or significance of these relationships [
5]. Despite indications from various studies that attachment to parents may wane during early adolescence, there is evidence of a reconciliation during later stages of adolescence [
21]. Moreover, heightened parent–child closeness during early adolescence endures over time, leading to increasingly supportive and less conflictive relationships [
22].
1.2. Conflict Frequency and Intensity during Adolescence
A fundamental dimension of parent–child interaction encompasses conflict, an interpersonal phenomenon often characterized by overt opposition and disputes [
23]. The dynamic of negotiating newfound freedoms, the erosion of parental control, and the concurrent rise in adolescent autonomy lays the groundwork for potential conflict [
8,
24]. Conflict, an ever-present dynamic in intimate relationships, notably becomes apparent in the context of family interactions [
5]. Considering the substantial implications of parent–adolescent conflict for the adaptive processes of adolescents [
25], a comprehensive exploration of its developmental trajectory across adolescence becomes indispensable. Throughout adolescence, a noticeable transformation becomes evident in terms of both the frequency and intensity of conflicts. Both conflict frequency [
26] as well as conflict intensity increase from early to middle adolescence [
8,
27]. Conflict frequency is relatively high from early to middle adolescence, and in the later stages of adolescence, a decline in conflict frequency between parents and children is shown [
26]. Empirical findings underscore a significant upswing in conflict intensity with parents during the progression from early to middle adolescence, succeeded by a substantial decline during the transition from middle to late adolescence [
8,
12,
27]. The initial level of conflict intensity with parents is higher for middle adolescents compared to early adolescents [
7]. Furthermore, there is a link between the perceived intensity of parental conflict and changes in parent–adolescent relationships aimed at fostering greater equality, although conflict does not act as a catalyst. Regarding the intensity of the conflict, the perceptions of the adolescents and parents diverge. In early-to-middle adolescence, discrepancies in perceptions of parents and adolescents increase, especially as adolescents recognize intensifying conflicts with their parents.
1.3. Aims of the Present Study
In the present study, our primary objective is to explore the influence of changing conflict frequency and intensity on the perceived level of closeness between parents and adolescents during the course of adolescence, putting a focus on the perspectives of both parties involved. Given the frequently observed disparities in how parents and adolescents evaluate their relationships during adolescence, it is important to differentiate between these two viewpoints [
5,
11]. Discrepancies in the perception of the parent–child relationship can arise, as adolescents seem to possess a more candid perspective than parents regarding the challenging facets of the relationship, and parents tend to underestimate the prevalence of conflicts between parents and adolescents [
4,
28,
29]. These disparities may reflect fundamental dynamics within the parent–adolescent relationship that are significantly linked to the adaptation of both adolescents and parents [
30]. Although numerous studies have examined changes in closeness and conflict during adolescence [
7,
12,
27], there is limited empirical evidence regarding the association between these two key aspects of communication. Given that closeness and conflict have significant implications for adolescent development [
31,
32], it is essential to explore the relationship between these two aspects to gain a more comprehensive understanding of parent–child relationships during adolescence. Few studies have explicitly focused on the link between conflicts and closeness, although existing evidence suggests a negative association between them [
33], as well as the notion that higher initial conflict intensity can lead to a subsequent decline in closeness between parents and adolescents over time [
34]. Our study spans from late childhood through early and middle adolescence, a period previously identified as pivotal in terms of conflicts and closeness within parent–child relationships [
8,
11]. Examining parent–child relationships over an extended timeframe, especially during adolescence, is of paramount importance as this is a critical stage during which parents face challenges in adapting their relationships with their adolescent children [
35].
In the present study, we incorporate three facets highlighted in the previous section, which have been recognized as pivotal in the context of conflicts and parent–child closeness during adolescence. We take a multifaceted approach, first considering both the frequency and intensity of conflict to unveil their (possibly distinct) effects on the level of closeness between parents and children. This stands in contrast to existing research, which tends to concentrate solely on either conflict frequency or intensity. Furthermore, by incorporating the perspectives of both parents and adolescents, we aim to gain a more comprehensive understanding of the distinct perspectives in a parent–child relationship during this critical developmental stage. This approach, thus, again diverges from numerous other studies that consider either the perspective of parents or adolescents. Lastly, we factor in the initial level of closeness to investigate whether a strong parent–child bond serves as a protective factor against increased conflict frequency and intensity during adolescence.
Based on previous research examining the development of conflict frequency, conflict intensity, and parent–child closeness from late childhood to middle adolescence [
27], this study expands former research and formulates three research hypotheses that focus on two perspectives on the same relationship: those of the adolescents themselves but also their parents. The first Hypotheses (1a–1c) are grounded in prior findings of a decline in closeness and an increase in the frequency and intensity of conflicts from early to mid-adolescence, as well as the enhanced accuracy of adolescents’ assessments of their relationships with their parents [
4,
7,
8,
12,
26].
Hypotheses 1. Firstly, we hypothesize that adolescents will report a more substantial increase in both conflict frequency (Hypothesis 1a) and intensity (Hypothesis 1b) and a more significant decline in closeness (Hypothesis 1c) compared to parents.
Hypotheses 2. Since existing evidence suggests a negative association between closeness and conflicts [34], we assume, secondly, that both parents and adolescents will indicate a greater reduction in closeness as both conflict frequency (Hypothesis 2a) and intensity (Hypothesis 2b) increase. Hypotheses 3. Based on the premise that closeness in early adolescence is crucial for later parent–child relationship development [22], we postulate, thirdly, that a higher level of closeness at the onset of the survey will correspond to a lesser decline in closeness over time in the presence of heightened conflict frequency (Hypothesis 3a) and intensity (Hypothesis 3b). 2. Materials and Methods
To analyze the hypotheses derived above, we use the data from the German Panel “Analysis of Intimate Relationships and Family Dynamics” (pairfam) in waves 2 to 12 [
36]. Since one item of the closeness scale was not assessed in wave 3, we had to exclude this wave. The panel, which started in 2008, is a multidisciplinary longitudinal study that investigates the partnership and family forms in Germany. The annual survey data comprises 12,000 randomly selected German-speaking anchor persons from the population register of stratified and randomly sampled municipalities across Germany who belong to the birth cohorts 1971–1973, 1981–1983, 1991–1993, and 2001–2003 (supplementary and refreshment sample in wave 11), as well as their partners, parents, and (biological as well as adoptive, foster and step-) children. Due to the large sample size and the elaborated sampling strategy that covers the wide spectrum of the target population, the results from this study are highly reliable. Both CAPI (=Computer Assisted Personal Interview) and PAPI (=Paper and Pencil Interview) interviews were conducted and incentivized with lottery tickets or gifts worth 5 to 10 euros [
37].
As part of the survey of the anchor persons, their partners (hereafter referred to as parents if they provided information on their children), and children, the characteristics of the relationship between these three groups of persons were collected. In the following analysis, we focus on the characteristics of frequency of conflicts, conflict intensity, emotional warmth, intimacy, and admiration/esteem. These constructs were surveyed annually as part of the parenting styles and network of relationship inventory modules [
38,
39], with parents (mothers and fathers) providing information on the relationship to each child and each child providing separate information on the corresponding parents.
In our sample, we use data from 3853 parents (57.1% female, average age of 38.1 [SD = 4.87] at their first participation) providing information on one (57.0% of all parent cases over all waves) or more (two: 37.1%, three or four: 5.8%) of their children on 3.4 waves on average resulting in a total number of 17,005 parent cases. Additionally, our sample comprises data of 3064 children (48.8% female, average age of 9.7 [SD = 1.97] at their first participation) providing information on one (25.4%) or two (74.6%) of their parents on 3.2 waves on average resulting in a total number of 15,841 child cases.
2.1. Closeness
The dependent variable in our analysis is closeness [
5]. We construct it from a total of seven indicators of the constructs: emotional warmth [
40], intimacy, and admiration/esteem [
38,
39]. For example, emotional warmth is represented by the item “Show your child that you like him/her”. Similarly, children responded to the item, “Anchor shows you that she/he likes you”. As the wording for the items was adapted for both parents and children, in the following, we only report the parents’ version. The exact wording of the questions for both versions can be found in
Appendix A Table A1. An example of intimacy is the item on the frequency of “Your child tells you what he/she is thinking”. An example of the construct admiration/esteem is the item on the frequency of “You express recognition for what your child does”. For all questions, a 5-point response scale (1—never, 2—rarely, 3—sometimes, 4—often, 5—very often/always) was used. We construct the scale of closeness by adding the values of all seven items for both parents and children (over all cases, parents and children: Cronbach’s alpha = 0.857). We center the resulting sum index around the mean and standardize it (z-score). Since we are interested not only in the development of closeness over time but also in the initial level of this characteristic as an independent variable, we have calculated the first measurement value of closeness for the corresponding person as a separate variable for each case.
2.2. Conflict Frequency
The first independent variable we employed to explain the development of closeness over time is the frequency of conflicts [
38,
39] between children and their parents. We use two items that refer to the frequency of “You and your child disagree and quarrel” and “You and your child are annoyed/angry with each other”. The same 5-point response scale as described above was used for these questions. As in the case of closeness, we added the values of the items (Cronbach’s alpha = 0.767) and formed the z-score of the resulting sum index.
2.3. Conflict Intensity
As a second independent variable, we construct a sum index for conflict intensity [
41] between children and their parents. The index is based on three items: “You criticize your child”, “You scream at your child when he/she did something wrong,” and “You scold your child when you are angry at him/her”, using the same 5-point response scale again. After adding up the values of the three items (Cronbach’s alpha = 0.663), we also calculated the z-score for this sum index.
2.4. Controls
In the following fixed effects regression analysis, we controlled for general differences according to the age of the children and cohort effects in addition to the independent variables explained above. While the age of the children is entered as a metric variable both linearly and quadratically into the model, we control for cohort effects based on aggregated survey waves [
42]. For this purpose, we combined waves 4, 11, and 12 and tested them against the remaining waves since closeness is slightly lower on average in these three waves.
Table 1 shows descriptive statistics for all characteristics used over the sample of parents and their children. The summary statistics combine the cases of parents and children.
2.5. Analysis
In the first step, we look at how closeness, frequency of conflicts, and conflict intensity change from the perspectives of both children and parents over time as children grow older. For this aim, we employ longitudinal fixed-effects models [
43,
44], in which we specifically consider the time-varying variance components of the three constructs. In cross-sectional statistical analyses, inferences are based on the comparison of characteristics considered relevant between the units of analysis (e.g., persons) at a fixed point in time (between-estimator). This implies the assumption that the units of analysis differ only in the observed characteristics (no unobserved heterogeneity); that is, differences in the dependent variable can only be attributed to differences in the independent variables. Furthermore, it has been pointed out that this assumption is usually violated, and it is suggested that panel data be used to examine causal effects [
45]. Since panel data captures the course of the characteristics of interest within a unit of analysis over time, the desired statistical effects can be estimated based on intra-individual changes (within-estimator). As a prerequisite, both dependent and independent characteristics must vary within a unit of analysis over time. Time-constant characteristics cannot be included in the estimation. Thus, the problem of unobserved heterogeneity is reduced, as time-constant unobserved heterogeneity is excluded from the analysis.
In the second step, we combine the time courses of the three constructs in a fixed effects model by explaining the time-related change in closeness as the dependent variable by the time-related changes in the frequency of conflicts and conflict intensity. Using interaction effects, we also test whether the associations between closeness and frequency of conflicts, as well as conflict intensity, are the same for children as for parents. In addition to controlling for the age of the children and the survey year, we take into account that children and parents from the same household are not independent in their answers to the survey questions and have the standard errors of the regressors clustered by household affiliation. Cases with missing values were deleted case-wise. All analyses were performed with STATA 17.0.
3. Results
First, we look at how the frequency of conflicts, conflict intensity, and closeness change from both the perspectives of parents and children over time as children grow older.
Figure 1 shows these time-related changes (the regression models underlying the figure are included in
Appendix A Table A2,
Table A3 and
Table A4). In the related models, we estimate effect sizes for both parents and children and distinguish them using interaction terms (“Child” indicates the difference in the effect size between children and parents). At the beginning of adolescence, the frequency of conflicts increases as expected from both perspectives of parents and children. While this increase continues steadily for children until the age of 15, parents report the most frequent conflicts with their children around their 13th year of life, which decrease again afterward.
Hypothesis 1a can thus be confirmed, although, complementing our assumptions, it also turns out that children not only perceive a more continuous increase in the frequency of conflict than their parents but that this initial increase in the parents’ perception of conflict even declines after a certain adjustment phase.
Both parents and children note a reduction in conflict intensity in their interactions during adolescence. While this decrease is relatively continuous for parents, children observe a more pronounced decrease in conflict intensity with their parents, particularly in the initial stages. However, the children’s assessment converges again with that of the parents in middle adolescence, although it is not apparent from the right censored data used here whether the perception of conflict intensity by the children decreases parallel to the perception of the parents after the age of 14 or reaches a plateau. In total, this finding contradicts Hypothesis 1b, especially for children between the ages of 10 and 12.
At the onset of adolescence, there is a shared perception of diminished closeness reported by both parents and children. For parents, this decrease is almost linear from their children’s 8th year of life. In contrast, children experience a slight increase in closeness at the beginning of adolescence, which reverses into a stronger decrease from their 9th year of life, supporting Hypothesis 1c.
Second, we investigate whether the strong decrease in closeness from both perspectives of parents and children can be explained by changes in the frequency of conflicts
(Hypothesis 2a) and conflict intensity
(Hypothesis 2b) as well as if the level at the onset of the survey corresponds with the decline
(Hypothesis 3a,b). Table 2 shows the results of this analysis with a fixed-effects model. In this model, again, we estimate effect sizes for both parents and children and distinguish them using interaction terms. According to this finding, perceived closeness from the parents’ perspective systematically diminishes when they observe an increase in conflict frequency with their children. The surveyed children also discern a decrease in closeness to their parents when they experience an increase in conflict frequency, supporting
Hypothesis 2a. Similarly, parents perceive a decrease in closeness with an increase in conflict intensity. Compared to parents, children also experience an even stronger distancing in the relationship with their parents, with an increase in conflict intensity. These results confirm
Hypothesis 2b.
Despite the predictions of Hypothesis 3a for parents, the association between conflict frequency and closeness persists, irrespective of whether the surveyed parents rated the relationship with their child as more or less close at the first survey point. For children, moreover, an increase in conflict frequency is associated with a stronger decrease in closeness, especially in relationships that the children experienced as comparatively closer at the beginning of the survey (p = 0.023 for the combined effect: Closeness t1 × frequency of conflicts + Child × closeness t1 × frequency of conflicts). These results again contradict hypothesis 3a for the children as closer relationships between parents and their children are not affected to a lesser degree by a rising number of conflicts than relationships that are already more distanced. For children, contrary to our expectations, closer relationships are affected more negatively by an increase in conflict frequency than looser relationships. Thus, our Hypothesis 3a must be dismissed. However, the decrease in closeness with an increase in conflict intensity is weaker in relationships that parents described as closer at the first survey time point. Like parents, children also perceive less decline in closeness with increasing conflict intensity if they described the relationship with their parents as closer at the time of the first survey. These results support 3b.
Moreover, the results emphasize the course of closeness over time that is independent of the frequency of conflicts and conflict intensity: Parents and children experience a decline in closeness that is expanding over time, with children showing the already known increase at the beginning of adolescence followed by a stronger decline in closeness afterward.