Is Parents’ Education Level Associated with Adolescent Self-Compassion?

Abstract

Self-compassion has been identified as a trait that correlates with robust mental health; specifically, less anxiety, depression and stress in both adolescents and adults. However, little is known about the parental and family factors that are associated with adolescent self-compassion that may promote or enhance the development of this stress-buffering trait. In this study, 1057 adolescents in grades 7–12 from two different school settings answered questions in an online survey that related to their parents’ education level and their own self-compassion. Results indicated that fathers’ education, but not mothers’, was associated with adolescent self-compassion. Specifically, adolescents whose fathers had a college education only had the highest level of self-compassion; a significant difference was found between self-compassion of adolescents of fathers’ with a college degree and those with a doctorate/professional degree. Adolescents whose fathers had less than a college education (some college, high school graduate) or more than a college education (masters or doctorate/professional degree) reported lower self-compassion. As parent education level may be a proxy for other factors such as socioeconomic status, parenting style, or parent-adolescent relationship closeness, further research is necessary that will measure these factors and parse out that which specifically is associated with self-compassion in adolescents.

Introduction

There is growing evidence that self-compassion, a modifiable psychological characteristic linked to adaptive responses to emotional distress (Trompetter, 2017), holds protective benefits for adolescents as they navigate an emotionally challenging developmental period (Játiva & Cerezo, 2014; Marshall et al., 2015; Xavier, Pinto-Gouveia, & Cunha, 2016; Zeller, Yuval, Nitzan-Assayag, & Bernstein, 2015). Defined by the three components of mindfulness (i.e., having a balanced perspective at times of difficulty), common humanity (i.e., understanding that our emotional struggles are shared by all others) and self-kindness (i.e., taking an active role in being kind to ourselves when we are struggling) (Neff, 2003), these components interact in a dynamic system to support a healthy way of self-relating and to increase the tendency to bounce back from life’s difficulties with greater ease (Bluth, Mullarkey, & Lathren, 2018; Marsh, Chan, & MacBeth, 2017). Not surprisingly, adolescents reporting high self-compassion also report lower stress, depression, and anxiety with a large effect size (Marsh et al., 2017), and high life satisfaction (Bluth & Blanton, 2015). Given these benefits, and that self-compassion varies from person to person, it is important to elucidate the conditions that promote its development.

Supported by Bronfenbrenner’s theoretical framework which explains individual development as taking place in an interactive system of family, society, and time, we posit that the development of self-compassion is influenced by factors both within the family, such as parenting style, as well as outside the family, such as the parents’ work environment and socio-cultural expectations. Although underexplored, growing evidence suggests that self-compassion is influenced by the quality of the attachment bond, and promoted by and deeply rooted in warm, supportive early interpersonal relationships between parent and child. As such, self-compassion has been predicted by maternal support, positive family functioning, and secure attachment style, while preoccupied and fearful attachment styles are inversely associated with self-compassion (Neff & McGehee, 2010). Individuals who recall their parents being over-protective and rejecting tend to be less self-compassionate, and those who perceive their parents as warm and caring are more self-compassionate (Irons, Gilbert, Baldwin, Baccus, & Palmes, 2006). Confirming these findings, self-compassion is negatively associated with both attachment anxiety and attachment avoidance in college students and community adults (Raque-Bogdan, Ericson, Jackson, Martin, & Bryan, 2011; Wei, Liao, Ku, & Shaffer, 2011) and poor maternal attachment, high parental rejection, low parental warmth and high parental overprotection experienced in childhood were found to predict low self-compassion in undergraduates (Pepping, Davis, O’Donovan, & Pal, 2015).

Although research is nascent on the likely complex connections between family factors and the development of adolescent self-compassion, rich and robust developmental research highlights the link between parent-child relationships and child mental health, behavior and academic outcomes, with a particular focus on the mother-child relationship. For example, Duchesne and Ratelle (2014) found that although attachment security to both mother and father predicted lower rates of depressive symptoms in pre-adolescents age 11–12, attachment security to the mother only predicted whether these symptoms would increase after age 12. Other research, however, indicates unique robust contributions of the father to emotional well-being. Amato (1994) reported that a close father-child relationship buffered against psychopathology in later adulthood, for example. Further, children with more involved fathers adjust better psychologically, do better academically on college entrance exams, have less antisocial behavior and more prosocial behavior, and have more fulfilling intimate relationships (Flouri, 2008; Flouri & Buchanan, 2002; Furr, 1998; Hwang & Lamb, 1997). Schenck and colleagues (2009) found that the amount of time that fathers spent with adolescents had a stronger association with adolescents’ perception that they mattered, and mattering was negatively associated with adolescent internalizing and externalizing behaviors. This finding is confirmed in homes where fathers are absent; in these homes, youth have been reported to have greater internalizing and externalizing behaviors (Parke, 2002).

Furthermore, extant literature suggests parental characteristics, including parents’ education level, also influence relationship quality and child outcomes, although these associations are complex and embedded in a myriad of related social and economic factors. For example, parental education is associated with greater parental investment in child’s education (Yamauchi, 2010) which is in turn associated with improved adolescent emotional functioning and mental health (Wang & Sheikh‐Khalil, 2014). Additionally, parental education is directly linked with numerous positive outcomes of children, including academic achievement (Dickson, Gregg, & Robinson, 2016) and physical health (Cochrane, Leslie, & O’Hara, 1982). Also, parental education is often used as a proxy for socioeconomic status, as the two are highly correlated (Sirin, 2005).

Acknowledging that parental factors contribute to adolescent behavior, and self-compassion is related to mental health and subsequently adolescent behavior, our research questions are: 1) What relationship does parents’ education have with self-compassion of adolescents? 2) Does mothers’ or fathers’ education differentially contribute to adolescent self-compassion? As this investigation is exploratory, and to our knowledge, no research has been conducted in this area, we are not offering a hypothesis as to the outcome of these research questions. The aim of this study is to offer preliminary insight into a potential pathway by which self-compassion is developed in adolescents, with the goal of providing motivation and direction for deeper examination in future studies. The link between parental characteristics, such as parental education, and adolescents’ self-compassion (and hence psychological and behavioral outcomes) is important for identifying preliminary ways in which self-compassion develops in adolescents. Once we can elucidate how self-compassion develops, we can find ways to support that development which will subsequently improve psychological outcomes for adolescents and contribute to establishing a lifelong positive mental health trajectory.

Method

Participants

Students who participated in this study attended two different school settings: a public middle and high school (grades 7–12), and a private all-girls’ school (grades 7–12). Both settings were in the southeast U.S. Demographics for the combined samples are included in Table 1. Inclusion criteria were attending grades 7–12 in either the public middle or high school or the all girls’ private school and being able to read English. Exclusion criteria was not being able to read English.

Table 1.

Participants’ Demographics (N = 1,057).

Frequency (%)
Gender
Male	344 (32.5)
Female	690 (65.3)
M→F; F→M	5 (0.5)
Unsure at this time	12 (1.1)
Missing	6 (0.5)
Age years
11–12	138 (13.1)
13–14	369 (34.9)
15–16	335 (31.7)
17–18	206 (19.5)
19	4 (0.4)
Missing	5 (0.5)
Highest education level of parents	Mother	Father
Less than HS/HS grad	250 (23.4)	337 (31.8)
Some college/college grad	503 (47.6)	425 (40.2)
MA/PhD degree	291 (27.6)	267 (25.2)
Missing	16 (1.5)	28 (2.6)
Race/Ethnicity
White	722 (68.3)
Black	174 (16.5)
Pacific Islander/Asian/Native Am	57 (5.4)
Hispanic/Latino	37 (3.5)
Other	62 (5.9)
Missing	5 (0.5)

Procedure

This study took place September – December, 2015. It was conducted with strict adherence to federal ethical and legal standards for conducting research with adolescents. After the university IRB and school administration approved the study, parents received a letter informing them of the purpose and procedure of the study. The letter explained that the purpose of the study was to ascertain adolescent well-being and its association with various factors. The university IRB approved passive consent; therefore, parents returned signed forms only if they did not want their child to participate. At the public middle and high school, parents of 56 students chose not to have their students participate; no reasons were provided. At the girls’ private school, all parents agreed to have their children participate. Students assented to the study via an embedded form at the beginning of the online survey and indicated their assent by proceeding to the survey questions.

All students at the public middle and high schools took the online survey during their first-period class. Students at the middle school took the survey on one day, and high school students took it over several days during one week (due to limited computer access). In total, 1030 students took the survey at the public middle and high schools. At the private girls’ school, 277 students took the survey at different times during the course of one day.

Measures

A number of measures were included in the online survey; however, for this research question we are including data only from the self-compassion measure and the demographic results:

Self-compassion.

The Self-Compassion Scale, short form (SCS-SF; Raes, Pommier, Neff, & Van Gucht, 2011) is comprised of 12 items, e.g., I’m intolerant and impatient towards those aspects of my personality that I don’t like (reverse scored). Responses to each item use a 5-point Likert scale ranging from 1 (Almost Never) to 5 (Almost Always). A total self-compassion score is computed by reverse scoring negatively worded items and then summing all 12 items. The potential range in values is from 12–60; higher score indicates greater self-compassion. Reported Cronbach’s alphas are good, ≥ .75 (Marshall et al., 2015; Raes et al., 2011). Correlation with the full scale is excellent; r = .97 (Raes et al., 2011). Reliability for the sample in this study is .72.

Demographics.

The online survey included questions about age, grade, race/ethnicity, level of mother and father’s education (1 = less than high school, 2 = high school graduate, 3 = some college, 4 = college graduate, 5 = master’s degree, 6 = doctorate or professional degree), and gender. For gender, the following options were provided as requested by adolescents in previous studies: male, female, male transitioning to female, female transitioning to male, unsure at this time.

Validity check.

Three questions were embedded in the survey as a validity check to determine if students carefully read the questions: 1) In the last month, how often have you eaten a meal? (5-point Likert scale from 1=never to 5=very often), 2) The President of the U.S. is Obama (5-point Likert scale from 1=agree to 5=strongly disagree), and 3) I go to school in the southeast U.S. (5-point Likert scale from 1=agree to 5= strongly disagree). If a participant answered “never” or “almost never” on question one or “strongly disagree, disagree, or neutral” on questions two or three, this was considered failing that question. If students failed two out of three of these items, these cases were removed from all analyses.

Statistical Analysis

A one-way ANOVA and Tukey’s pairwise comparison as an ad-hoc method were conducted to examine a possible mean difference of adolescents’ self-compassion by their father’s level of education. With an ANOVA result indicating a significant mean difference between different education groups, Tukey’s pairwise comparison was conducted to identify specific pairs with a significant mean difference. Cohen’s effect sizes were calculated as well. According to Cohen (1988), effect sizes of 0.20, 0.50, and 0.80 are considered as small, medium, and large, respectively.

Results

In the public middle and high school, 244 students failed the validity check and were removed from all analyses. In the all-girls’ preparatory school, six failed the validity check and were removed from all analyses. Resulting sample size was 1057.

A one-way ANOVA (Analysis of Variance) was conducted to examine the difference in the adolescents’ level of self-compassion by mothers’ and fathers’ education levels. Levene’s test showed no significant violation of the assumption of equal variance across the different groups (p= 0.13). The result showed a significant difference for fathers’ education level adjusting for sample and gender, F(5, 1023) = 2.227, p = .049, R² = 0.011. However, there was no significant difference in adolescents’ level of self-compassion by mothers’ education level, F(5, 1022) = 2.027, p = .072, R² = 0.010, so no further analyses were conducted using mothers’ education level. The results remained unchanged after controlling for the possible sample and the gender effect for both fathers’ level of education, F(5, 1004) = 2.353, p = .038, and mothers’ level of education, F(5, 1004) = 2.047, p = .070. A multi-factor ANOVA result showed that the effect of fathers’ education level on self-compassion remained unchanged when controlling for mothers’ education level (Table 2).

Table 2.

Multifactor ANOVA results of fathers’ education level on self-compassion controlling for mothers’ education

	df	F	η2	p
Fathers’ education level	5 (0.92)	2.35	0.011	*
Mothers’ education level	5 (0.52)	1.32	0.006	0.25
Gender (0 = male, 1 = female)	1 (10.05)	25.69	0.025	***
School (0 = school A, 1 = school B)	1 (2.82)	7.22	0.007	**
Residual	999 (0.39)

Note. Numbers in parentheses are mean squared errors. Fathers’ and mothers’ education level: 1 = less than high school, 2 = high school graduate, 3 = some college, 4 = college graduate, 5 = master’s degree, 6 = doctorate or professional degree.

Tukey pairwise comparisons on the adolescents’ level of self-compassion between different fathers’ education levels (Table 3) showed that a significant difference of adolescents’ level of self-compassion appeared between the adolescents of fathers with a college degree (N = 273, M = 3.09, SD = 0.66) and those of fathers with a doctorate/professional degree (N = 109, M = 2.88, SD = 0.72), p = .034. Adolescents of fathers with a doctorate/professional degree showed significantly lower levels of self-compassion compared to adolescents of fathers with a college degree. The Cohen’s effect size for the mean difference is 0.319, indicating a small to medium effect size. Adolescents of fathers with a doctorate/professional degree also showed marginally lower levels of self-compassion compared to adolescents of fathers who were high school graduates (N = 234, M = 3.08, SD = 0.57), p = 0.07 (Figure 1). The Cohen’s effect size for this mean difference is 0.324, also a small to medium effect (Cohen, 1988).

Table 3.

Tukey pairwise comparisons on the adolescents’ level of self-compassion by fathers’ education levels

Father’s education level	mean difference	lower bound	upper bound	adjusted p-value
1 vs. 6	0.18	−0.07	0.43	0.32
1 vs. 5	0.06	−0.17	0.29	0.97
1 vs. 4	0.04	−0.17	0.25	1.00
1 vs. 3	0.02	−0.22	0.25	1.00
1 vs. 2	0.02	−0.19	0.24	1.00
2 vs. 6	0.20	−0.01	0.41	0.07†
2 vs. 5	0.09	−0.10	0.27	0.78
2 vs. 4	0.01	−0.15	0.18	1.00
2 vs. 3	0.01	−0.18	0.20	1.00
3 vs. 6	0.19	−0.03	0.42	0.15
3 vs. 5	0.08	−0.13	0.29	0.88
3 vs. 4	0.02	−0.16	0.21	1.00
4 vs. 6	0.22	0.01	0.42	0.03*
4 vs. 5	0.10	−0.08	0.28	0.61
5 vs. 6	0.11	−0.11	0.34	0.70

Note. Fathers’ education level: 1 = less than high school (adolescents’ level of self-compassion, n = 103, M = 3.06, SD = 0.6), 2 = high school graduate (n = 234, M = 3.08, SD = 0.57), 3 = some college (n = 152, M = 3.07, SD = 0.61), 4 = college graduate (n = 273, M = 3.09, SD = 0.66), 5 = master’s degree (n = 158, M = 2.99, SD = 0.68), 6 = doctorate or professional degree (n = 109, M = 2.88, SD = 0.72).

Figure 1.

The boxplot with mean self-compassion scores of adolescents by fathers’ education level. Fathers’ education level 1 = less than high school, 2 = high school graduate, 3 = some college, 4 = college graduate, 5 = master’s degree, 6 = doctorate or professional degree

Discussion

Although significant empirical evidence supports parenting and family experiences as playing a critical role in shaping children’s skills for managing distress (Morris, 2007), the specific factors that lead to optimal child outcomes are complex. In an effort to elucidate the family characteristics that may promote healthy adolescent self-compassion development, this study is a preliminary investigation of the relationship of parents’ education to adolescents’ self-compassion. As self-compassion is associated with robust mental health, determining whether a parental attribute, such as parents’ education, is associated with self-compassion can begin to elucidate the factors that contribute to, foster, and support adolescent self-compassion and subsequently adolescent mental health.

Results indicated that adolescent self-compassion differed relative to the level of fathers’ education, but not to mothers’ education. In particular, self-compassion of adolescents was greatest when fathers’ education was at the college level (i.e., having graduated college), and significantly greater than that of adolescents whose fathers had a doctorate/professional degree. There were no significant differences in self-compassion between other levels of fathers’ education (i.e., below high school, high school, some college, master’s degree).

We frame these findings within several theories regarding the role of parents and family environment in shaping adolescents’ emotional skills and mental health outcomes, particularly in the context of challenging circumstances. First, attachment theory provides a robust general framework for understanding how the quality of parenting behavior and resulting attachment style plays a life-long role in how one interprets and manages distressing experiences (e.g., see Mikulincer, Shaver & Pereg, 2003). Indeed, research on attachment bonds in adolescence suggests close and supportive parental relationships continue to facilitate adolescents’ positive sense of self, emotional well-being and capacity to cope constructively with difficulties (Moretti & Peled, 2004). Meanwhile, Belsky’s model of parental functioning posits that parent characteristics, child characteristics, and contextual sources of stress and support (e.g. social networks, martial and occupation factors) within the family environment all play an important role in shaping a parent’s ability to cultivate these nurturing and supportive parent-child bonds (Belsky, 1994).

Drawing on both attachment theory and Belsky’s model, it is plausible that highly educated fathers are employed in positions that require a high level of personal commitment, including long hours, travel, or other demands that negatively impact parental functioning and deter the development of supportive father-adolescent relationships. Youth from these families may frequently spend time either alone during after school hours or involved in extracurricular activities; in either case, these practices do not engender family intimacy (Luthar, 2003; Luthar & D’Avanzo, 1999; Rosenfeld & Wise, 2000; Shafran, 1992). As a result, adolescents in these contexts may feel isolated and disconnected from their parents (Luthar & D’Avanzo, 1999). These adolescents may lack the opportunity to internalize support they would otherwise receive, and thus become more self-critical and less self-compassionate. However, as parent behavior and parent-child relationship quality were not measured in this study, these explanations are speculative and require further investigation.

Furthermore, researchers also concede the importance of parent characteristics in influencing parental behavior and, ultimately, child outcomes (Belsky,1994; Morris 2007). Considering this, well-educated fathers may have traits and behavioral tendencies that inhibit self-compassion development in their child. For example, as some evidence suggests educational attainment and perfectionism are associated (Perrone, 2007), it is possible that father’s with professional degrees are more prone to self-criticism and low self-compassion. Given parents serve as important models of emotional skills (e.g., emotional expression, acceptability and regulation) in their day-to-day interactions with their children (Morris, Silk, Steinberg, Myers & Robinson, 2007), and influence both children’s strategies for managing negative emotions (Brand & Klimes-Dougan, 2010) and ways of appraising and coping with stressors (including self-critical responses) (Power, 2004), it is plausible that parents with low self-compassion may both explicitly and implicitly teach their children to be less compassionate with themselves when faced with emotional pain. In line with this hypothesis, preliminary work by our group suggests parent and pre-adolescent self-compassion may be strongly positively correlated (unpublished). Studies that examine both parent and adolescent self-compassion, as well as emotion socialization practices and responses to negative experiences that may promote or hinder self-compassion development in children, are needed in order to explore these possibilities.

Similarly, father’s education may be a proxy for socioeconomic status (SES) (Douglas-Hall & Chau, 2007; Sirin, 2005), which may also play a role in parenting behavior and parent-child relationships. Evidence suggests affluent families may hold high academic standards for their adolescent children, causing increased stress and mental health symptomatology (Luthar & D’Avanzo, 1999). Among high-income families, maladaptive perfectionism (i.e., concern over mistakes, doubts about actions, parental expectations, parental criticism) has been linked with adolescent distress, delinquency, and substance use (Luthar & Becker, 2002), which has then been associated with anxiety and depression, suggesting that these adolescents may use substances to self-medicate and relieve distress (Luthar, 2003). Indeed, in one study, adolescents from high SES families were reported to have greater anxiety and depression than both their peers from low SES inner-city populations and national norms (Luthar, 2003; Luthar & D’Avanzo, 1999). Depression and anxiety are inversely related to self-compassion in both adults and adolescents (see meta-analyses: MacBeth & Gumley, 2012; Marsh et al., 2017).

Parental education may also be a proxy for parenting style, and has been associated as such in previous studies (Kiadarbandsari, Madon, Hamsan, & Mehdinezhad Nouri, 2016). Higher educated parents are more likely to use an authoritative parenting style than an authoritarian or permissive style (Dornbusch, Ritter, Leiderman, Roberts, & Fraleigh, 1987), which is both high in warmth and rule-setting, and which tends to promote positive outcomes for Western youth (Boon, 2007; Kiadarbandsari et al., 2016). Further, the inverse association between authoritative parenting style and psychological symptoms has been found in groups of both low (below college level) and high (above college level) educated parents (Lamborn, Mounts, Steinberg, & Dornbusch, 1991). Therefore, it may be that level of education per se is not what is predicting the level of self-compassion in adolescents, but other unmeasured factors, such as closeness with parents, socioeconomic status, or parenting style, or a combination thereof.

If future research finds that this is the case, lifestyle or societal level changes can be made to support the development of parental closeness and authoritative parenting. For example, policy changes at work can facilitate greater father-adolescent engagement by allowing for greater flexibility in work hours, permitting meetings to take place on online platforms rather than requiring travel, and overall, providing a more family-friendly workplace. If future research confirms that authoritative parents have children who have greater self-compassion, then parenting classes can be offered to new parents which promote skills of authoritative parenting. Thus, interventions can be created, and policies changed to promote those factors that lead to greater self-compassion in adolescents.

The finding that there was no relationship between mothers’ education and level of self-compassion was surprising, as previous studies have emphasized the salience of mothers’ education on youth outcomes (Howes, 1988; Jackson,2009). However, there is some evidence that demonstrates that during the adolescent period, the role of the father has greater influence on adolescent mental health outcomes. For example, fathers’ emotional support of adolescents predicted positive mental health of the adolescents over and above other factors (including depression severity of mother) (Collishaw et al., 2016), fathers’ warmth contributed uniquely to emotional and academic development (Suizzo et al., 2017), and father’s behavior towards their adolescent offspring contributed to the tenor of friendships that the adolescent developed (Flynn, Felmlee, Shu, & Conger, 2018). Thus, it may be that fathers’ influence and support is particularly important during the adolescent period, specifically in the development of self-compassion.

There are a number of limitations to this study. First, as previously discussed, it is unclear whether fathers’ education or unmeasured associated factors, such as family relationship quality, socioeconomic status, or parenting style, influenced the development of adolescent self-compassion. Even if fathers’ education uniquely contributes to self-compassion, the variance explained by fathers’ education was small, which suggests there are other important and unmeasured factors that warrant investigation. For example, some studies have shown that mindfulness of parents predicts self-compassion of adolescents (Moreira, Gouveia, & Canavarro, 2018). Other studies have suggested family level factors (e.g., marital strife, birth order, number of children in family) and child level factors (e.g., early life experiences, personality type) all have important implications on an individual’s development and on self-compassion specifically (Bistricky et al., 2017; Black, Grönqvist, & Öckert, 2018; Neff & McGehee, 2010; Pepping et al., 2015). Additionally, these factors need to be investigated in future studies in order to accurately ascertain that which influences the development of self-compassion in adolescents. Second, both samples came from adolescents living in the southeast U.S., it is unknown whether these results generalize to other settings or regions. Given this, we urge continued research that includes measurement of the individual, parent, family, school and environmental factors that may interact to promote self-compassionate tendencies in adolescents. By closely examining these factors, we can develop strategies and target interventions that ensure children reach adolescence armed with an emotional skillset that allows them to thrive.