Self-Esteem ’ s Relations to Empathy and Parenting

The article is looking into theoretical and research relations between self-esteem, empathy and parenting. The empirical study was carried out among 199 undergraduate US students and measured empathy (Empathy Quotient and Interpersonal Reactivity Index), self-esteem (Rosenberg Self-Esteem Scale) and perceived parenting (Parental Bonding Instrument). The results showed no direct relation between empathy and self-esteem, as measured by the instruments in this sample. The care dimension of the perceived parenting style of both parents predicted self-esteem levels. Future research is recommended to confirm findings and identify possible mediator between empathy and self-esteem.


Introduction
Self-esteem is among the most popular mental health concepts and has influenced the development of numerous scientific and public policies.Self-esteem refers to the value that people put on themselves (Baumeister, Campbell, Krueger, & Vohs, 2003) or relatively stable feelings of overall self-worth (Rosenberg, 1965).It has been named as a predictor of academic success, career achievements, lower aggression and reduced deviant behavior (Baumeister et al., 2003;Laible & Carlo, 2004;Laible, Carlo, & Roesch, 2004;Rey, Extremera, & Pena, 2011).Some studies however report modest or no correlation with emotional intelligence (Baumeister et al., 2003), academic achievement (Shopov, Atanasova, Ivanova, Nenova, & Sofronieva, 2015) and other previously expected results of healthy self-esteem (Baumeister et al., 2003).
How does self-esteem relate to one important aspect of emotional intelligence, namely empathy?Do people with higher self-esteem tend be more empathetic?Does parenting affect the self-esteem of people beyond adolescence?
These are some of the questions this text will venture to answer.

Theoretical Background
Self-esteem is a key concept in the psychological school of self-psychology, and its relations to mental health are treated in depth (Kohut, 1999).Kohut and his followers see self-esteem as closely tied to parenting, especially parental empathy, and also to the subsequent empathy of the child (Kohut, 1999;Trumpeter, Watson, O'Leary, & Weathington, 2008;Watson, Hickman, Morris, Milliron, & Whiting, 1995).
Self-psychology authors present healthy self-esteem as one of the extremes of a continuum of self-representations, with the other extreme being depleted depression, and the intermediate stages respectively adaptive and maladaptive narcissism (Trumpeter et al., 2008).All four stages have overlapping areas and these areas could have mixed characteristics.Healthy self-esteem reflects an integrated self, which has been "nourished" properly by the primary caregiver.This nourishment is parental empathy, and it is vital for psychological survival in the same way that "oxygen is needed for physical survival" (Trumpeter et al., 2008, p. 52).In the first 18 months the caregiver needs to meet most of her infant's needs, but shortly after that parenting needs to provide the "optimal frustrations" (Kohut, 1971, p. 64).The optimal frustrations are moments of empathic failure, which are prodding the child to start developing its own identity.
As soon as the self begins to form, it also starts relating to others and experiencing empathy.The initial empathic perception of the infant reflects a total overlap with the emotional state of the other; she is "flooded" by the observed emotion.In the process of maturing, the empathy becomes more differentiated and the emotional resonance reflects only a fraction of the observed emotion.When the infant cries, the self-integrated mother feels part of his or her anxiety, picks the infant up and comforts him or her.In the cases when the mother's care and empathy have remained immature, she reacts with panic to the anxiety of the child and thus creates more anxiety in the infant, triggering what Kohut calls the traumatic empathic echo (Kohut, 1971, p. 171).When this is a repeated pattern, we would probably see an individual, who is on the maladaptive side or narcissism and further from healthy selfesteem (Kohut, 1999).
In recapitulation, optimal parenting provides the person with a healthy self-esteem and an integrated self.She or he is then able to authentically connect to others in a mature empathic way and contribute to the self-integration of those in his or her care, whether children or patients.

Previous Research
So, do people with high self-esteem tend to be more empathetic?It has been widely accepted that empathy influences self-esteem, as people tend to have a better self-concept when they understand others and are able to form and maintain satisfying relationships (Eisenberg, 2002;Eisenberg & Eggum, 2009;Laible, Carlo, & Roesch, 2004).The review, however, did not identify many research efforts that connect the two phenomena.Laible and colleagues look into mechanisms to develop self-esteem in late adolescence.They affirm that empathy and prosocial behavior affect self-esteem directly and are important mediators of the influence of parent and peer attachments on self-esteem (Laible, Carlo, & Roesch, 2004).Rey and colleagues focus on the relations among emotional intelligence, self-esteem and life satisfaction in adolescence and reach a similar result: self-esteem mediates the influence of emotional intelligence on life satisfaction and therefore emotional intelligence affects self-esteem directly (Rey et al., 2011).
The relationship of empathy and self-esteem is found to be bi-directional.Stotland and Dunn use an experiment to look into the relations between self-esteem and empathy (Stotland & Dunn, 1963).They expect to find higher empathy levels among low self-esteem subjects, but contrary to their expectation the subjects in the high selfesteem group show higher empathy.They suggest that self-esteem in turn affects empathy: the individuals with higher self-esteem are less occupied with their own doubts and tension and therefore are more able to focus on the other and offer help as necessary.
When a multidimensional concept of empathy is being studied in relation to self-esteem, a consistent modest (Davis, 1983;De Corte et al., 2007;Lee & Song, 2015) or strong (Passanisi, Gensabella, & Pirrone, 2015) correlation between cognitive empathy and self-esteem is reported, but not between affective empathy and self-esteem.
The strongest significant correlation between the empathy variables (affective and cognitive) and self-esteem is a negative correlation with personal distress (De Corte et al., 2007).
Compared to the relationship between self-esteem and empathy, the one between self-esteem and parenting has received much more thorough research (Garber, Robinson, & Valentiner, 1997;Manczak, DeLongis, & Chen, 2016).Gordon Parker focuses on parenting style and interprets it in terms of parental bonding (Parker, Tupling, & Brown, 1979, p. 1).He sees parental attachment, as conceptualized by Bowlby (Bowlby, 1969), as the core element of parenting style.Parker and his team carried out a series of studies, focused on how the two main dimensions of the parental style, care and overprotection, affect mental health (Parker, 1983;Parker, Tupling, & Brown, 1979).While researching the connection between parenting styles and depression, they did not find the direct correlation they expected.This prompted them to look closer into how personality characteristics, such as selfesteem, could mediate this influence.Indeed, they discover significant correlations between self-esteem and maternal care, maternal protection, and paternal protection (Parker, 1993).
Trumpeter and colleagues also look into how self-esteem affects relations between parenting and mental health (Trumpeter et al., 2008).They use the construct of the self-representations spectrum, which starts with depressive depletion and ends with a healthy self-esteem.Their work investigates the connections between self-representations from any stage of the spectrum and perceived parental empathy and love.The results show a positive correlation of self-esteem with mother's cognitive empathy and a significant, but weaker one with mother's affective empathy.
As expected, self-esteem has a strong negative correlation with depression.Trumpeter and colleagues conclude that healthy self-functioning is strongly related to perceived parental empathy.Watson and colleagues, working in the same framework, confirmed the observed correlations between self-esteem and parental nurturance (Watson et al., 1995).Manczak et al. (2016) find that parents who are empathic also have a higher self-esteem.However, they also pay a physiological price for their empathy in the form of systematic inflammation.Their adolescent children show behavioral, emotional and physiological benefits, gained from the empathic approach of their parents.
Is there a parenting style that is most beneficial for a healthy self-esteem?Baumrind and her followers assert that the authoritative parenting style, involving high warmth and high demands, is the most beneficial one (Baumrind, 1966;Lamborn, Mounts, Steinberg, & Dornbusch, 1991).There have been questions raised about whether the cultural context affects the outcomes from the parenting style (Olivari, Hertfelt Wahn, Maridaki-Kassotaki, Antonopoulou, & Confalonieri, 2015).Steinberg and colleagues find that the representatives of minority groups in the US have best outcomes when parented in an authoritative style as well (Steinberg, Mounts, Lamborn, & Dornbush, 1991).Antonopoulou and colleagues investigate specifically the effect of paternal parenting style on empathy and self-esteem among Greek children and again find the authoritative style to be the most beneficial one (Antonopoulou, Alexopoulos, & Maridaki-Kassotaki, 2012).García and Gracia, however, show that in Spain Psychological Thought 2016, Vol.9(2), 184-196 doi:10.5964/psyct.v9i2.194 Self-Esteem's Relations to Empathy and Parenting the permissive style turns out to be more beneficial regarding self-esteem and general social functioning (García & Gracia, 2009).

Method
The present study is looking into the relations of self-esteem with empathy and perceived parenting style among young adults.
The first hypothesis is that the combined empathy variable will correlate positively with self-esteem levels, i.e., higher empathy predicting higher self-esteem.The rationale is research, showing that individuals with higher selfesteem would show more empathy at the sight of another's distress, as they are less occupied with themselves and able to focus on the other (Eisenberg, 2002;Eisenberg & Eggum, 2009;Kalliopuska, 1989;Laible, Carlo, & Roesch, 2004;Stotland & Dunn, 1963).
The second hypothesis is that self-esteem would be predicted by parenting, particularly by a style, which is high in warmth and low in overprotection.The rationale here would be that parents that are warm and attentive to the needs of their young children would instill a stable feeling of self-worth, and the lower levels of overprotection would stimulate exploration and development of self-efficacy beliefs (Baumrind, 1966;Cheng & Furnham, 2004;García & Gracia, 2009;Lamborn et al., 1991).

Participants and Socio-Demographic Data
The participants were 206 undergraduate students from 6 universities in US. 7 were removed from this study for not reporting on one or both parents.The remaining 199 were 96 men, 101 women and 2 others, with an average age of 20.1 years.
This study collected data on the participants' gender, school, age, nationality, major, childhood family, current parental family, theater participation, and participation in any communication-related training.The characteristics of the sample can be found in Tables 1 and 2. Table 1 presents the distribution of the respondents according to specialty, gender and university and Table 2 presents the distribution according to childhood family, training and theater participation.
With regards to current parental family, 159 reported that their parents were still together, 30 identified their parental family as consisting of only mom, 6 -only dad and 4 -others.

Measures Empathy Quotient/The Cambridge Behaviour Scale
The Scale was published by Simon Baron-Cohen and Sally Wheelwright in 2004 and is being used with both clinical and non-clinical populations (Baron-Cohen & Wheelwright, 2004).Empathy is seen as a multi-dimensional construct, which includes affective, cognitive and mixed components.The scale consists of 60 items, of which 40 are measuring empathy and 20 are "filler" items.The respondents can answer definitely agree, slightly agree, slightly disagree, or definitely disagree and score 2, 1, or 0. The final result is one score, varying from 0 to 80.
The construct validity was evaluated by 6 external experts, who agreed on which items belong to the suggested definition of empathy.The chance of such agreement is p < 0.003.The re-test reliability of the Quotient r = 0.97 at p < .001(Baron-Cohen & Wheelwright, 2004;Lawrence, Shaw, Baker, Baron-Cohen, & David, 2004).

Interpersonal Reactivity Index
The Index was created in 1980 by Mark Davis and also treats empathy as a multi-dimensional construct (Davis, 1980).The end results are four separate scores for each subscale.
The Index consists of 28 items, to which the respondents have to answer on a 5-point scale from Does not describe me well to Describes me very well.There are four subscales, each one containing 7 items.The first subscale is Fantasy, and it measures respondents' tendencies to transpose themselves imaginatively into the feelings and actions of fictitious characters in books, movies, and plays (Cheetham, Hänggi, & Jancke, 2014).The second one is Perspective Taking, and it taps the ability to understand and discriminate among similar emotions and motivations of an observed individual.It is conceived as the cognitive component of empathy.The third one is Empathic Concern, and it represents the affective component of empathy.It is the ability to share the emotion of an observed individual and experience other-oriented care, especially in cases of observed suffering.The last one is Personal Distress, and it measures self-oriented anxiety and discomfort at the sight of observed suffering (Davis, 1983).
Davis provides the analysis of the construct validity of the Index by correlating it to two other empathy measures: Hogan scale and Emotional Empathy Scale by Mehrabian and Epstein (in Davis, 1980).The first one is meant to measure the cognitive aspect of empathy and the second scale -the emotional aspect (Davis, 1983).He finds correlations with Hogan's scale r = 0.43 for men and r = 0.37 for women, with p < .05,while the correlations with the Mehrabian and Epstein Scale are respectively r = 0.63 for men and r = 0.56 for women, with p < .05(Davis, 1980).Test-retest reliability is from r = 0.61 to r = 0.79 for men and from r = 0.62 to r = 0.81 for women, with р < .001(Davis, 1980).

Parental Bonding Instrument
The PBI was created by Gordon Parker, Hillary Tupling and L. B. Brown in 1979 and conceptualizes parental style as parental bonding (Parker, Tupling, & Brown, 1979).The scale is a retrospective instrument, containing two identical questionnaires about the behaviors and attitudes of the mother and the father of the respondents, as Overprotection.The correlation between the two scales is r = -0.238(p < .001),which means the two scales are not independent of each other (Parker et al., 1979).

Data Collection
The data collection began in September, 2015 and ended in December, 2015.
The study was reviewed and approved by the internal review board for compliance with ethical standards.An informed consent form was collected from each participant, with their full name, signature and date.The students were recruited in or out of class, and a snack was offered as a compensation for their participation.The instruments were offered in paper-and-pencil format, stapled in one bundle.After hearing a short introduction by the researcher, they read a general instruction and each instrument was prefaced by its original instruction.The survey took between 15-25 minutes.The respondents placed the filled out questionnaire bundles and informed consent forms into two separate boxes in order to preserve confidentiality.

Data Analysis
The raw data was inserted into Excel tables and was afterwards converted into files, used by the statistical software products JASP and SPSS.T-tests, ANOVA analyses with post-hoc tests and linear regression analysis were performed in order to control for influence of socio-demographic variables, as gathered by the socio-demographic component of the study.The connections among self-esteem, combined empathy variable (empathy quotient), the affective empathy variable (empathic concern), the cognitive empathy variable (perspective taking), fantasy, and personal distress were analyzed with Pearson's correlations.The relationship of parenting dimensions (maternal care, maternal overprotection, paternal care, paternal overprotection) and self-esteem was analyzed with linear regression analysis.Through direct logistic regression analysis it was also examined whether the same factors affect the probability of a respondent to fall into the below the norm category, as the ones, affecting the probability of him/her to fall into the in the norm category.

Results
T-tests, a linear regression analysis and ANOVA analyses with post-hoc tests did not show any significant influence on self-esteem by the characteristics, included in the socio-demographic component.The statistics on the sociodemographic characteristics are as follows: gender: t(195) = 1.62, p = .12;university: F( 5 The next step was to look into the correlations and see if there was a significant relation between empathy and/or any of its components and self-esteem, as expected.The results are presented in Table 3. Psychological Thought 2016, Vol.9(2), 184-196 doi:10.5964/psyct.v9i2.194 Self-Esteem's Relations to Empathy and Parenting Surprisingly, self-esteem did not correlate with any of the empathy variables: the combined empathy variable (empathy quotient), the affective or the cognitive variables (empathic concern and perspective taking, respectively).
The only significant, but modest correlation with a phenomenon from the empathy "family" (p < 0.01) was a negative correlation with personal distress.The higher the self-esteem was, the less the person was prone to personal distress at the sight of other people's suffering.
The correlation analysis between self-esteem and the four parenting variables confirmed a moderate relationship with two of the parenting variables: maternal care (r = 0.317, p < .001)and paternal care (r = 0.263, p < .001).In order to look further into this relationship, linear regression analysis was performed.Self-esteem was the dependent variable and all parenting variables were plugged in as factors.A significant regression model was found, F(4, 194) = 9.458, p < .001,and more than 16% of the results on self-esteem in the current survey were predicted by the perceived parenting styles of the parents, R 2 = .163,RMSE = 6.006.The linear regression analysis confirmed that the care dimension of both parents affected self-esteem, but the overprotection one did not exert statistically significant influence (see Table 4).Note.Cl = confidence interval; LL = lower limit; UL = upper limit.The next step was to compare if the same factors influenced the chances of a respondent from the sample to report self-esteem below, in, or above the norm.A direct logistic regression was performed, as is presented in Table 5.
The independent variables were the four parenting variables, gender, and the empathy quotient.The dependent variable was self-esteem.The full model was statistically significant, χ Cone esteem in the sample.The self-esteem norm for the sample was designed by adding one standard deviation to the mean in both directions of the variance.In order to record any differences in the level of self-esteem, three subgroups were formed: below the norm, in the norm, and above the norm.Above the norm was selected as the reference category.Table 5 presents the resulting two models.The first one is the likelihood ratio of a person's dropping below the norm, and the second one is the likelihood of a person's being in the in the norm category, compared to above the norm.
The reference category is: above the norm.b This parameter is set to zero because it is redundant.
In the first model, the likelihood of a person reporting self-esteem below the norm, only two of the variables contributed significantly: again mother care and father care, as expected.The higher their value was, the lower was the likelihood of a person falling in the below the norm category.
In the second model, the likelihood of a person being in in the norm category, compared to the above the norm one, was not affected by any of the included factors.

Discussion
The study investigated the influence of empathy and parenting style on self-esteem.In accordance with the majority of the previous studies and literature, it was expected that both empathy and parenting style affect self-esteem.
Psychological Thought 2016, Vol.9(2), 184-196 doi:10.5964/psyct.v9i2.194 Self-Esteem's Relations to Empathy and Parenting The first hypothesis was that the combined empathy variable, the empathy quotient, would be positively correlated to self-esteem.Such correlation was not confirmed by the data from the sample.The cognitive component of empathy, measured by the Interpersonal Reactivity Index, did not show any relation to self-esteem, either, in contrast to earlier studies (Davis, 1983;De Corte et al., 2007).The data from this survey does not let us conclude that there is any significant relation between empathy and self-esteem, as measured by our instruments.This finding is surprising, especially given the backdrop of theory and research agreeing on a bi-directional nature of the relation between self-esteem and empathy.We cannot assert at this point whether the results mean that empathy and self-esteem, as measured by the instruments, indeed do not correlate for this sample, or that the phenomena are related, but through a mediator.Further research would be beneficial, especially including a wider spread of quantitative and qualitative tools and a bigger number of relevant variables.More specifically, relevant mediating variables from the current literature review could include guilt and shame (Eisenberg, 2000), overall emotional intelligence (Rey et al., 2011), and relationships with peers (Laible, Carlo, & Roesch, 2004).
In agreement with a small group of studies (Davis, 1983;De Corte et al., 2007), the current work detected a significant correlation between self-esteem and personal distress.It is a negative one, showing that people with high self-esteem tend to show lower personal distress.The concept of personal distress was developed in order to distinguish between our empathetic, other-focused reaction at the sight of someone's distress and a self-oriented reaction (Batson, 2009).Personal distress, however, as measured by the Interpersonal Reactivity Index, is not entirely empathic in nature.It refers to a more general type of distress in critical situations.For example, some of the items in the scale are "In emergency situations, I feel apprehensive and ill-at-ease," and "I am usually pretty effective in dealing with emergencies" (Davis, 1980, p. 11).Therefore, we could conclude that people with higher self-esteem tend to panic less and react more adequately in cases of emergency.No conclusions can be made about their empathic understanding or sensitivity towards others.
The second hypothesis about parenting style affecting self-esteem in young adults was partially confirmed.Parental warmth, or care does affect self-esteem, as expected and corroborated by previous studies.The second conceptualized element of parenting style, however, parental overprotection, was not found to be a significant factor for self-esteem development.It is interesting that the current study does not support Parker's findings with regards to parental overprotection and self-esteem (Parker, 1993).Both studies use the same instruments to measure the two constructs, and the number of respondents is comparable (n = 123 in Parker's study, n = 199 in current study).The most probable reason is that Parker's sample was clinical (depressed individuals), and the current respondents present a non-clinical sample.It has been validated that depressed people tend to perceive their parents as more controlling in the first 16 years of their life (Parker, 1983).This finding hints at the dynamic relationship between self-esteem and parenting.The logistic regression showed that the influence of the parental care dimension on self-esteem is visible in the lower self-esteem group, but is absent in the normal self-esteem group.
In conclusion, our study showed a lack of direct correlation between self-esteem and empathy but confirmed a moderate influence of parental care on self-esteem.Parental overprotection was not found to affect self-esteem either, which does not allow us to make any conclusions about the most beneficial parenting style with regards to self-esteem, based on this study.It was suggested that future research, exploring possible mediating variables and employing other measurements, needs to be carried out in order to gain a better insight into relations that seem more complex than expected.

a 1 =
Lincoln Christian University; 2 = University of Wisconsin-Milwaukee; 3 = Liberty University; 4 = Marquette University; 5 = Milwaukee School of Engineering; 6 = Fairfield University.bOne male participant chose not to report specialty, which resulted in him not being included in this table.
they have perceived them in the first 16 years of their life.The final PBI version, proposed by Parker et al. (1979), uses 25 items out of the original 112, 12 of which belong to the factor, named Care and 13 to the factor, named Overprotection (Parker et al., 1979).The split reliability in a non-clinical sample is Pearson r = 0.879 (p < .001)for the Care scale and Pearson r = 0.739 (p < .001)for the Overprotection scale (Parker et al., 1979).Test-retest reliability is r = 0.761 (p < .001)for the Care scale and r = Psychological Thought 2016, Vol.9(2), 184-196 doi:10.5964/psyct.v9i2.194Cone 0.63 (p < .001)for the Overprotection scale.The validity is evaluated by two experts through semi-structured interviews and agreement levels are respectively r = 0.77 (p < .001)for the Care scale and r = 0.43 (p < .001)for , 193) = 0.762, p = .578;age: b = -0.05,t(197) = 0.40, p = .69,r 2 = 0.001, with the linear model not found adequate: F(1, 197) = 0.16, p = .69;nationality: F(10, 188) = 0.725, p = .700;field of study or specialty: F(3, 194) = 0.734, p = .533;theater participation: t(197) = -0.64,p = .521,and participation in any communication-related training: t(197) = 0.88, p = .38.The variables childhood family and current parental family had four options of response: living with both parents; living with mom; living with dad; living with others.The assumptions of homogeneity of variance and normality of distribution for the ANOVA analyses of the influence of childhood family and current parental family on self-esteem were violated and the results are not included in further discussions.

Table 1
Distribution of the Respondents in the Study According to Specialty, Gender and University

Table 2
Distribution of Respondents According to Reported Childhood Family, Theater and Communication-Related Training Participation

Table 3
Pearson Correlations Between Self-Esteem and Empathy Variables

Table 5
Parameter Estimates of Direct Logistic Regression, Predicting Likelihood of Reporting Self-Esteem Below or in the Norm