Abstract
Research investigating the development of psychological problems among children has typically focused on the role of parenting practices. The purpose of this study was to investigate the direct effects of recalled parenting practices (emotional warmth, rejection, and overprotection) on psychological distress and indirect effects through self-criticism. Participants were 410 undergraduate students (290 females and 120 males) with a mean age of 20.10 (SD = 1.65). Participants completed the measure of My Memories of Upbringing—Egna Minnen Betraffande Uppfostra; Forms of Self Criticizing/Attacking and Self Reassuring Scale; and Depression, Anxiety, and Stress Scale as self-report measures. Results provided evidence for highly similar patterns of associations among parenting practices, self-criticism, and psychological distress for both maternal and paternal models. Findings showed that parental emotional warmth and rejection were directly related to self-criticism and were indirectly related to psychological distress through self-criticism. Only maternal rejection was directly related to psychological distress. However, maternal and paternal overprotection was not related to psychological distress either directly or indirectly. Moreover, maternal and paternal parenting practices explained 61% and 58% of the variance in psychological distress, respectively. The findings of the present study highlighted the importance of self-criticism and past memories of parenting as intervention targets to support well-being of young people.
Introduction
Depression, anxiety, and stress are common psychological problems among college students (DeRosier, Frank, Scwartz, & Leary, 2013; Greenberg, Fournier, Sisitsky, Pike, & Kessler, 2015; Kessler, 2012; Walters, Bulmer, Troiano, Obiaka, & Bonhomme, 2018), and these problems have detrimental effects on academic outcomes (Day, McGrath, & Wojitowicz, 2013). Theoretical approaches explain the causes of psychological or emotional problems in adolescence or adulthood with early poor parenting practices or insecure attachment (Bowlby, 1977). Empirical studies based on these theories provide support for the links between early poor parent–child relations and later anxiety and depression (Bifulco et al., 2006; Lazarus, McLellan, & Hudson, 2018; Moran, Bailey, & DeOliveira, 2008; Reitman & Asseff, 2010). Results of a meta-analysis study (Khaleque & Rohner, 2002) including sample from diverse ethnic groups from the United States and also from Africa, Asia, Europe, South America, and Caribbean acknowledge that recalled parenting practices transculturally predict approximately 26% of the variance in children’s psychological adjustment and 21% of the variance in adults’ psychological adjustment. Therefore, it can be said that the empirical findings underscore the etiological role of early parenting practices in the development of psychosocial functioning across cultures. However, through which mechanism parenting practices influence children’s development is not fully explored. Based on these discussions, the present study has mainly two purposes: (a) to examine the relationship between maternal and paternal parenting practices (emotional warmth, rejection, and overprotection) and psychological distress (depression, anxiety, and stress) and (b) to investigate the mediator role of self-criticism in these associations.
Parenting practices and psychological distress
The relationship between early parenting experiences and children’s psychosocial functioning is one of the main topics of interest for many theoretical and empirical studies (Baumeister & Leary, 1995; Bowlby, 1969, 1973; Gilbert, 2005; Siegel, 2001). Past research underscores the role of early parenting quality in childhood for the development of anxiety and depression in adulthood (Moran et al., 2008; Reitman & Asseff, 2010; Schimmenti & Bifulco, 2013). Individuals who have poor parent–child relations are more likely to develop cognitive framework that increase the risks for depression (Li, Jiang, Fan, & Zhang, 2018; Morley & Moran, 2011). More specifically, early poor parent–child relations is more likely to result in internal working models that may lead to a person to have a negative view of the self and others and in turn, this may increase the risk of depression, anxiety, and stress.
In this research, as parenting practices, maternal and paternal emotional warmth, rejection, and overprotection have been investigated. Parental emotional warmth has been defined as verbal and nonverbal behaviors that include acceptance, expressions of support, positive affect, and involvement. Parental rejection has been considered as negative or hostile feelings toward the child including verbal or physical behaviors such as indifference or neglect, disapproval, and criticism. Lastly, parental overprotection has been defined as parental behavior aimed at protecting the child from harm by using control behaviors (Skinner, Johnson, & Snyder, 2005).
Specifically, parental rejection (Reitman & Asseff, 2010; Schimmenti & Bifulco, 2013) and parental control (Reitman & Asseff, 2010) have been linked to later anxiety. Decreased emotional warmth and increased rejection and overprotection are negatively associated with self-esteem (Arrindell et al., 2005) and self-efficacy (Niditch & Varela, 2012). In addition, parental rejection and overprotection have been linked to depression (Baker & Hoelger, 2012; Oldehinkel, Veenstra Ormel, de Winter, & Verhulst, 2006; Penelo, Viladrich, & Domenech, 2010; Richter & Eisemann, 2001) and anxiety (Bijl, De Graaf, Ravelli, Smit, & Vollebergh, 2002; Hulsenbeck, Meesters, Merkelbach, & Munis, 2000) in children and adults and in clinical and nonclinical samples. Parental rejection characterized by denying children’s ideas, low levels of emotional warmth, and high levels of judgemental and dismissive behaviors have been associated with clinical anxiety in children (Hudson & Rapee, 2001). This kind of parenting practices may result in negative self-esteem or lower self-efficacy, and children with a negative view of themselves are more likely to use self-criticism (Leary, Tate, Adams, Allen, & Hancock, 2007).
Parenting practices and self-criticism
Parenting practices and early parent–child relationships also have been considered as predictor variables for the self-criticism (Baldwin, 1992, 2005; Beck, 1967; Kopala-Sibley & Zuroff, 2014). Internal schemas that are created in relations with others during childhood, subsequently affect peoples’ sense of self and self-relating behaviors (Baldwin, 1992, 2005; Baldwin & Fergusson, 2001). Attachment theory proposes that children develop internal working models based on the responsiveness of the caregiver. These early models play crucial roles on the individuals’ cognitions and behaviors regarding relational experiences (Gillath, Shaver, & Mikulincer, 2005). When a caregiver is regarded as responsive, the child is likely to develop a sense of security and competence to deal with negative life events using self-warmth and reassurance (Gilbert & Procter, 2006). However, if the child perceives attachment figure as unresponsive, he or she is likely to develop insecure attachment relationship which may result in anxiety or avoidance (Gillath et al., 2005). Insecurely attached children may perceive others as more powerful than themselves by focusing on social rank, and in any situation that need to make attribution they are more likely to use self-criticism instead of criticizing others (Cantazaro & Wei, 2010; Gilbert & Procter, 2006). In previous research, insecure attachment and negative relational schemata were positively associated with self-criticism (Cantazaro & Wei, 2010; Kopala-Sibley, Zuroff, Leybman, & Hope, 2013). In contrast, secure attachment relationships were negatively associated with self-criticism and positively linked to self-reassurance (Irons, Gilbert, Baldwin, Baccus, & Palmer, 2006; Kopala-Sibley et al., 2013).
In addition to attachment theory, studies based on parenting theories have also shown the links between parenting behaviors and self-criticism. For example, early parenting practices (abuse, criticism, rejection, overprotection, and neglect) are related to self-critical behavior and which in turn linked to later emotional and social difficulties (Gilbert, Clarke, Hempel, Miles, & Irons, 2004). Previous findings indicated that parental psychological control is associated with self-criticism (Ahmad & Soenens, 2010; Kopala-Sibley & Zuroff, 2014; Soenens, Vansteenkiste, Luyten, Duriez, & Goossens, 2005; Soenens et al., 2008). Achievement-oriented parental control that is common parenting practice in Turkey (Sahin & Ozer, 2012) is especially associated with self-criticism (Soenens, Park, Vansteenkiste, & Mouratidis, 2012). While parental rejection and control are linked with self-criticism (Irons et al., 2006), parental warmth and affection are related to self-reassurance (Gilbert et al., 2004).
Self-criticism and psychological distress
Self-criticism is a form of self-to-self relationship that involves the feelings of self-worthlessness and self-deficiency. Self-to-self relationship has three basic features, namely inadequate self, hated self, and reassured self (Gilbert et al., 2004). The inadequate self focuses on correcting or improving the self, the hated self focuses on the will to hurt and attack to himself or herself, and reassured self focuses on positive aspects (Castilho, Pinto-Gouveia, & Duarte, 2015). Self-criticism compromises negative self-evaluation and harsh self-scrutiny (Shafran, Cooper, & Fairburn, 2002). People with high levels of self-criticism focus on excessive self-evaluation, high personal standards, and intense desires to increase self-worth. These people are more likely to be self-critical when they could not attain an achievement goal and feel a sense of failure, worthlessness, or guilt (Blatt, 2004).
In relation to these cognitive processes, self-criticism potentially increases symptoms of psychopathology (Castilho et al., 2015; Gilbert et al., 2004; McIntyre, Smith, & Rimes, 2018). The theories that are trying to explain depression focus on the role of self-criticism. For example, making internal, stable, and global attributions for negative outcomes are considered as the important risk factor for depression in the attributional formulation of learned helplessness theory (Abramson, Seligman, & Teasdale, 1978) and hopelessness theory of depression (Alloy, Abramson, Metalsky, & Hartlage, 1988). Self-criticism or self-blame is also conceived as a leading factor for depression in the psychoanalytic tradition (Blatt & Homann, 1992) and Beck’s (1967) theory of depression. Later, empirical studies provided support for these notions. For example, self-judgment was found to be the strong predictor of depressive symptoms (Van Dam, Sheppard, Forsyth, & Earleywine, 2011). Empirical studies have shown that self-criticism is linked to psychopathology in both clinical and nonclinical samples and in cross-sectional and longitudinal studies (Blatt, 2004; Blatt & Luyten, 2009), especially with depression and anxiety (Blatt, 2004; Gilbert et al., 2004; Luyten & Blatt, 2013). On the other hand, self-reassurance has been positively related to well-being (Gilbert, Durrant, & McEwan, 2006; Gilbert et al., 2004).
Self-criticism as a mediating mechanism
Previous research has not only indicated relationship between parenting and psychosocial functioning but has also begun to explain through which mechanisms early parenting practices influence later psychosocial functioning. Given the relationship between parenting practices, self-criticism and ill-being, self-criticism may mediate the link between the recalled parenting practices and psychological distress. To date, however, the mediator role of self-criticism on ill-being was investigated in a few studies. For example, self-criticism has been shown to mediate the relationship between childhood maltreatment and adult anxiety and depression (Richter, Gilbert, & McEwan, 2009). In one of the studies (Irons et al., 2006), high self-criticism and low self-reassuring have been found to be significant mediators between negative past parental experience and depression. In another study (Castilho, Pinto-Gouveia, Amaral, & Duarte, 2014), self-criticism mediated the relationship between submissiveness experiences in childhood and later depression/anxiety. Furthermore, it has been shown that the link between paternal overprotection and depression was mediated through the perfectionism and concerns over making mistakes (Enns, Cox, & Larsen, 2000). Enns et al. (2000) also reported that self-criticism, perfectionism, and concerns over making mistakes mediated the link between lack of maternal care and depression. Lastly, the findings from a study (Lassri & Shahar, 2012) revealed the mediator role of self-criticism between emotional maltreatment in childhood and problematic romantic relationships. Therefore, addressing the mediator role of self-criticism through which early parenting practices are linked to the later psychosocial functioning can help professionals to develop effective intervention programs. It may be more effective to intervene the personality factors such as self-criticism than interpersonal schemata or attachment patterns.
Moreover, previous studies also show that the sex of participants in the relationship between parenting practices and mental health problems can make a difference. For instance, perceived parental rejection is associated with more depression in girls (Akse, Hale, Engels, Raaijmakers, & Meeus, 2004; Feinberg, Neiderhiser, Reiss, Hetherington, & Simmens, 2000). While some studies show that women are experiencing internalizing problems and men are more open to externalizing problems (Eschenbeck, Kohlman, & Lohaus, 2007; Stiles, Liu, & Kaplan, 2000), some other studies found that sex has no role in the effect of positive or negative parenting practices on internalizing problems (e.g., Barnett & Scaramella, 2013). These contradictory findings suggest that further researches are needed to examine whether the sex of the participants makes a difference on the links between recalled parenting practices, self-criticism, and psychological distress.
Cultural context for parenting
Childrearing practices differ within and across cultures and play crucial roles on children’s developmental outcomes (Kagitcibasi, 2007; Sen, Yavuz-Muren, & Yagmurlu, 2014). In traditional Turkish culture, different parenting roles have been expected from mothers and fathers. While mothers were expected to behave more nurturing toward their children, fathers were expected to behave more disciplinarian in the family system (Kagitcibasi & Sunar, 1992). However, with the effects of urbanization and globalization, Turkish family structure has shown rapid changes in recent decades (Ataca & Sunar, 1999; Yagmurlu & Sanson, 2009). Today, contemporary parental roles and childrearing practices seem to become more egalitarian, and even parental practices of both mothers and fathers seem to show general uniformity in Turkish culture (Sunar, 2002). Recent empirical research results also confirm that Turkish maternal and paternal childrearing processes do not differentially affect children’s psychosocial development (e.g., Ozdemir, Vazsonyi, & Cok, 2017). Nevertheless, studies conducted to date have focused more on the effects of either mothers or composite scores of both parents on children’s adjustment. The studies investigating the roles of maternal and paternal parenting practices on children’s adjustment separately are rather limited (Lee & Hunsley, 2006). Although these theoretical and empirical discussions provide some proof about uniformity of maternal and paternal roles, the present research aimed to examine the effects of parenting practices on children in separate models not to overlook possible parental differences.
The present study
The present study examines the relationships between recalled maternal and paternal parenting practices, self-criticism, and psychological distress in a sample of emerging adults. The first hypothesis of the study was that maternal/paternal emotional warmth are directly and negatively, and maternal/paternal rejection and overprotection are directly and positively related with both self-criticism and psychological distress. The second hypothesis posits that recalled maternal/paternal parenting practices are indirectly related with psychological distress through the self-criticism. As there are contradictory findings in the literature, no a priori hypothesis was established regarding the role of sex.
Method
Participants and procedures
In the research, a total of 410 undergraduate students (290 females and 120 males) were recruited through convenience sampling from a university located in a western part of Turkey. The ages of the participants ranged between 18 and 27, with a mean age of 20.10 (SD = 1.65). For reported parental education, 62.7% of the participants’ mothers had primary school or less education, 11.5% had a secondary school degree, 15.1% had a high school degree, 9.6% had an undergraduate degree, and 1.2% had a graduate degree. For fathers, 34.4% had a primary school or less education, 19% had a secondary school degree, 24.6% had a high school degree, 20.2% had an undergraduate degree, and 1.7% had a graduate degree. These findings indicate that the vast majority of participants come from families with low educational levels. Prior to the administrations, researchers explained the purpose of the study, inclusion criteria, and got informed consent. Inclusion criteria were being a university student and currently not taking any psychotherapeutic intervention or medication. Participants voluntarily filled the battery of questionnaires in their classes and they received extra course credit for their participation.
Measures
Recalled parenting practices
In order to measure recalled parenting practices, the My Memories of Upbringing—Egna Minnen Betraffande Uppfostra (EMBU; Perris, Jacobsson, Lindstrom, von Knorring, & Perris, 1980) was used. The participants completed the EMBU separately for their mothers and fathers by focusing on their childhood and adolescence years. This scale has 23 items and three subscales, namely, emotional warmth (7 items; e.g., “My parents show that they love me”), rejection (7 items; e.g., “My parents beat me for no reason”), and overprotection (9 items; e.g., “My parents want to decide how I should be dressed or how I should look”). Items are rated on a four-point Likert-type scale. Both the original and Turkish versions of the EMBU have good validity and reliability indices (Arrindell et al., 1999; Dirik, Yorulmaz, & Karancı, 2015). In this study, for the maternal practices, Cronbach’s alpha coefficients for the emotional warmth, rejection, and overprotection subscales were calculated as .76, .66, and .78, respectively. For the paternal practices, the Cronbach’s alpha coefficients for the emotional warmth, rejection, and overprotection subscales were determined as .83, .67, and .77, respectively. As the maternal and paternal rejection subscales indicated good evidences for validity and each involved less than 10 items, Cronbach’s alpha coefficient of .60 and above was regarded as acceptable (Loewenthal, 2001).
Self-criticism
In order to measure self-judgment of the participants, we used the Forms of Self Criticizing/Attacking and Self Reassuring Scale (FSCRS; Castilho et al., 2015; Gilbert et al., 2004). The scale includes 22 items and three factors, namely, inadequate self (9 items; e.g., “I am easily disappointed with myself”), hated self (5 items; e.g., “I have become so angry with myself that I want to hurt or injure myself”), and reassured self (8 items; e.g., “I am able to remind myself of positive things about myself”). Items are rated on a five-point Likert-type scale (0 = Not at all like me, 4 = Extremely like me). Previous studies (Castilho et al., 2015; Gilbert et al., 2004) indicated that FSCRS has good validity and reliability to measure self-criticism. In this study, FSCRS was adapted into Turkish and its psychometric properties were examined. For Turkish adaptation, forward translation and back translation of the FSCRS were conducted. Research results indicated that the original three-factor structure showed good fit to the data: x2 (206, N = 406) = 599.495, x2/df = 2.91, p < .001, root mean square error of approximation (RMSEA) = .07, 90% confidence interval (CI) = (.06, .08), comparative fit index (CFI) = .90, goodness-of-fit index (GFI) = .88. Factor loadings ranged from .44 to .82 for inadequate self, .54 to .81 for hated self, and .52 to .82 for reassured self subscales, and they were all significant at p < .001. The Cronbach’s alpha coefficients were computed to be .85 for inadequate self, .80 for hated self, .88 for reassured self, and .92 for total scale.
Psychological distress
In order to measure psychological distress, Depression, Anxiety, and Stress Scale (DASS-21; Antony, Bieling, Cox, Enns, & Swinson, 1998; Henry & Crawford, 2005) was used. This scale is composed of 21 items, and all the items are rated on a four-point Likert-type scale (0 = Did not apply to me at all, 3 = Applied to me very much, or most of the time). DASS-21 includes three subscales, namely, depression (7 items; e.g., “I was unable to become enthusiastic about anything”), anxiety (7 items; e.g., “I felt scared without any good reason”), and stress (7 items; e.g., “I found it hard to wind down”). It is reported that DASS-21 has good psychometric properties in nonclinical samples (Henry & Crawford, 2005). In this study, DASS-21 was adapted into Turkish and its validity and reliability were examined. In order to test the hypothesized factor structure of the DASS-21 in Turkish culture, a second-order confirmatory factor analysis was conducted, and the findings indicated that the original factor structure of the DASS-21 demonstrated good fit indices: x2 (186, N = 395) = 596.519, x2/df = 3.062, p < .001, RMSEA = .07, 90% CI = (.07, .08), CFI = .90, GFI = .87. Factor loadings ranged from .59 to .85 for depression, .43 to .75 for anxiety, and .53 to .79 for stress subscales, and they were all significant at p < .001. In this study, Cronbach’s alpha coefficients for the depression, anxiety, stress subscales, and total scale were .87, .83, .85, and .94, respectively.
Data analysis
Before the main analyses, preliminary analyses were conducted. To begin with, missing values were screened. As Little’s missing completely at random test showed that the data were missing completely at random (x2 = 3266,024, df = 4163, p > .05), 23 cases with missing values were omitted from the data set. In addition, after deleting univariate (13 cases, z-score > 3.29) and multivariate outliers (16 cases, Mahalanobis distance larger than the critical value at .001 level), the data including 358 cases in total were used in the further analyses. Although skewness and kurtosis values indicated no threats for univariate normality assumption, multivariate kurtosis z-statistic value was higher than 5.00, suggesting the violation of multivariate normality which is a fundamental assumption in maximum likelihood estimation (Byrne, 2010). Therefore, the bootstrapping technique was used to eliminate the assumption of multivariately normal sampling distribution based on 5000 resampling iterations (Hayes, Preacher, & Myers, 2011). As a result, this study used maximum likelihood as the estimation method and bootstrapping to estimate the biases in the results. Following preliminary analyses, descriptive statistics were computed for the study variables. Later, bivariate Pearson correlations were calculated to assess the relationships among the latent variables. Next, structural equation modeling was conducted to examine the relationships among recalled parental practices, self-criticism, and psychological distress. Analyses were conducted in AMOS 21.0. Both the direct paths between parental practices, self-criticism, and psychological distress and indirect paths between parental practices and psychological distress through self-criticism were explored for paternal and maternal models separately. To determine if these relationships varied by sex of participants, a series of z-test analysis were conducted. We treated parenting practices (emotional warmth, rejection, and overprotection), self-criticism, and psychological distress as latent variables. We used items of the parenting dimensions as the observed variables. For the self-criticism and psychological distress, factors were used as the observed variables. Additionally, we drew 5000 bootstrap samples from the full data and used 95% CI to determine the significance of the indirect effects. In model evaluation, we take into account the most commonly used fit indices: x2/df ratio (x2/df), the RMSEA, the CFI, and the GFI (Mueller, 1996). An acceptable fit is indicated by an x2/df below 3, RMSEA below .08, GFI above .90, and an CFI above .95 (Schermelleh-Engel, Moosbrugger, & Muller, 2003). It is also proposed that researchers should not be so strict about cutoff values for fit indexes. It is emphasized that when evaluating overall fit of hypothesized model, model complexity and theoretical background should also be considered (Marsh, Hau, & Wen, 2004). In addition, it is posited that as the number of observed variables per latent variable increase, model fit indices tend to become lower (Kenny & McCoach, 2003). Therefore, in the present study, based on both theoretical and practical suggestions, researchers considered more lenient criteria when assessing model fits.
Results
Descriptive statistics are given in Table 1. Bivariate correlations between recalled maternal and paternal parenting practices (emotional warmth, rejection, and overprotection), self-criticism, and psychological distress are presented in Table 2. All the maternal and paternal parenting practices were significantly associated with self-criticism and psychological distress. Higher levels of maternal and paternal emotional warmth were significantly associated with lower levels of self-criticism and psychological distress. However, higher levels of maternal and paternal rejection were significantly associated with higher levels of self-criticism and psychological distress. Higher levels of maternal and paternal overprotection were significantly associated with higher levels of self-criticism and psychological distress. Self-criticism was also significantly and positively related to psychological distress. In general, most associations were quite modest in size.
Descriptive statistics of recalled maternal and paternal parenting practices, self-criticism, psychological distress (N = 358).
Bivariate Pearson correlations between recalled maternal and paternal parenting practices, self-criticism, and psychological distress.
*p < .05; **p < .01.
Measurement model
Before testing the structural model, measurement models were tested. The measurement model for recalled maternal parenting practices involved five latent constructs, namely, maternal emotional warmth, maternal rejection, maternal overprotection, self-criticism, and psychological distress. In latent constructs of parenting practices, scale items were used as observed variables. Self-criticism was represented by three observed variables: inadequate self, hated self, and reassured self. Psychological distress was indicated by three observed variables, namely, depression, anxiety, and, stress. The confirmatory factor analysis showed that the measurement model for recalled maternal parenting practices fits the data well: x2 (367, N = 358) = 670.010, x2/df = 1.83, p < .001, RMSEA = .05, 90% CI = (.04, .05), CFI = .90, GFI = .88. The measurement model for paternal parenting practices was constructed in parallel with measurement model for maternal practices. The measurement model for paternal parenting practices also demonstrated adequate fit to the data: x2(367, N = 358) = 697.287, x2/df = 1.90, p < .001, RMSEA = .05, 90% CI = (.05, .06), CFI = .90, GFI = .88. In both measurement models, standardized factor loadings were significant at .01 level, and correlations between latent constructs were in the expected direction as well.
Structural model
We tested latent structural models separately for maternal and paternal parenting practices. These models examine the direct effects of the recalled parenting practices both on self-criticism and psychological distress, and the indirect effects of recalled parenting on psychological distress through self-criticism. For maternal parenting, the result indicated that the hypothesized model fits the data adequately, x2(367, N = 358) = 670.010, x2/df = 1.826, p < .001, RMSEA = .05, 90% CI = (.04, .05), CFI = .90, GFI = .88. Hypothesized model and standardized direct effects are presented in Figure 1.

Hypothesized model for recalled maternal parenting practices. The values are standardized estimates. *p <.05; **p <.01; ***p <.001.
Maternal emotional warmth had a significant direct effect on self-criticism (β = −.22, p < .05) but not on psychological distress (β = −.09, p > .05). The indirect effect of maternal emotional warmth on psychological distress via self-criticism was also significant (β = −.16 (95% bootstrapped CI = −.31, −.01, p < .05)). Maternal rejection had a significant direct effect on self-criticism (β = .32, p < .01) as well as psychological distress (β = .21, p < .05). The indirect effect of maternal rejection on psychological distress through self-criticism was also significant (β = .22 (95% bootstrapped CI = .04, .41, p < .05)). The direct effect of maternal overprotection on self-criticism (β = .02, p > .05) and on psychological distress (β = −.01, p > .05) was not significant. In addition, the indirect effect of maternal overprotection on psychological distress via self-criticism was also not significant (β = .01 (95% bootstrapped CI = −.12, .15, p > .05)). Lastly, self-criticism significantly predicted psychological distress (β = .70, p < .001). Furthermore, the findings indicated that recalled maternal emotional warmth, rejection, and overprotection together explained almost 23% of the variance in self-criticism, and maternal parenting practices and self-criticism altogether accounted for almost 61% of the variance in psychological distress.
For paternal parenting, the results showed that the hypothesized model fits the data adequately: x2(367, N = 358) = 697.287, x2/df = 1.900, p < .001, RMSEA = .05, 90% CI = (.05, .06), CFI = .90, GFI = .88. Hypotesized model for paternal parenting practices and standardized direct effects are presented in Figure 2.

Hypothesized model for recalled paternal parenting practices. The values are standardized estimates. *p <.05; **p <.01; ***p <.001.
The paternal emotional warmth had a significant direct effect on self-criticism (β = −.19, p < .05) but not on psychological distress (β = .02, p > .05). The indirect effect of paternal emotional warmth on psychological distress via self-criticism was also not significant (β = −.13 (95% bootstrapped CI = −.29, .04, p > .05)). The direct effect of paternal rejection on self-criticism was significant (β = .40, p < .01) but was not significant on psychological distress (β = .10, p > .05). The indirect effect of paternal rejection on psychological distress through self-criticism was significant (β = .28 (95% bootstrapped CI = .10, .51, p < .01)). The direct effects of paternal overprotection on self-criticism (β = −.04, p > .05) and on psychological distress (β = .00, p > .05) were not significant. In addition, the indirect effect of paternal overprotection on psychological distress through self-criticism was not significant (β = −.030 (95% bootstrapped CI = −.17, .09, p > .05)). Lastly, self-criticism significantly predicted psychological distress (β = .72, p < .001). We also found that recalled paternal emotional warmth, rejection, and overprotection together accounted for almost 26% of the variance in self-criticism, and paternal parenting practices and self-criticism altogether explained almost 58% of the variance in psychological distress.
To examine the effect of sex on the associations between recalled maternal and paternal parenting practices, self-criticism, and psychological distress, we performed a series of z-tests (Paternoster, Brame, Mazerolle, & Piquero, 1998). In this test, the difference between each of the paths in the females’ and males’ models should be greater than |1.96| (absolute value of 1.96) to be able to argue that there is a statistically significant difference between paths at p < .05 (two-tailed test). The results showed that there were no significant differences between path coefficients of females’ and males’ in the maternal model. For the paternal model, the only difference was found between paternal rejection and self-criticism link. The z-value for the paternal rejection and self-criticism was −2.069. This finding shows that the paternal rejection and self-criticism path (.42) in the females’ model were significantly different from that of the males’ model (.07). In other words, this finding suggests that father rejection is associated with more self-criticism in females.
Discussion
Some important findings were obtained from this study. First, the findings of the present study showed that the hypothesized model examining the importance of recalled parenting practices in predicting self-criticism and psychological distress fit the data well. Second, consistent with the theoretical and empirical expectations, we got evidence of both direct and indirect effects of recalled parenting practices on psychological distress. As expected, recalled maternal and paternal emotional warmth was related to lower levels of self-criticism which in turn was related to lower rates of psychological distress. On the other hand, recalled maternal and paternal rejection was directly related to higher levels of self-criticism and psychological distress and indirectly related with greater psychological distress through self-criticism. However, we did not find any direct effect of maternal and paternal overprotection both on self-criticism and psychological distress and any indirect effect on psychological distress as well.
When parenting practices are addressed separately, direct effects of both maternal and paternal emotional warmth on psychological distress were not significant, but the indirect effects appeared to be significant. In other words, maternal and paternal emotional warmth supports a healthy sense of self (low levels of self-criticism), thus in turn reduce the likelihood of experiencing mental health problems. This finding is consistent with the previous studies that reported parental support and warmth as the salient factors for self-reliance and social competence, and lower levels of psychological distress (Gray & Steinberg, 1999; Steinberg, 2001).
The findings of this study revealed that recalled maternal and paternal rejection is also a significant factor in the mental health of young people. While both direct and indirect effects were significant in the maternal model, only the indirect effect was found significant in the paternal model. In terms of the direct effect, the findings are mostly consistent with previous research indicating that parental rejection is related to higher levels of depression (Baker & Hoelger, 2012; Oldehinkel et al., 2006; Penelo et al., 2010; Richter & Eisemann, 2001) and anxiety (Bijl et al., 2002; Hulsenbeck et al., 2000). In addition, these findings are also consistent with theoretical explanations suggesting that early poor parent–child relations is likely to increase negative views of the self and in turn lead to psychological distress (Lazarus et al., 2018; Li et al., 2018; Morley & Moran, 2011). For paternal model, there was no direct effect of paternal rejection on psychological distress. The possible explanation for this finding would be that in Turkish culture, parenting roles are highly carried out by mothers, and compared to fathers, children give more importance to mothers’ behaviors toward themselves. Besides, since fathers are higly accepted as authority figure in the family system, paternal rejection is more likely to influence children’s sense of self in a negative way, and which in turn leads to later mental health problems.
Previous studies have shown that parental control and overprotection are related to higher depression and anxiety (Cassidy, Lichtenstein-Phelps, Sibrava, Thomas, & Borkovec, 2009; Oldehinkel et al., 2006; Penelo et al., 2010; Reitman & Asseff, 2010), lower self-esteem (Arrindell et al., 2005), and self-efficacy (Niditch & Varela, 2012). In this study, we did not find a significant relationship between overprotection and self-criticism, and also no relationship between overprotection and psychological distress. It has been discussed that perceptions regarding parental control differ across cultures (Ballash, Leyfer, Buckley, & Woodruff-Borden, 2006; Crockett, Veed, & Russell, 2010; Kagitcibasi, 1970). It means that in eastern cultures, parental control is not perceived negatively as in Western cultures. The finding that parental overprotection is not related to self-criticism and psychological distress can be explained by normative parenting strategies emphasizing overprotection on children in Turkish culture (Kagitcibasi, 1970). As a result, we can conclude that this parenting style is expected in Turkish culture, and it is not a factor that influences mental health negatively.
One of the important findings of the present study is the strong relationship between self-criticism and psychological distress. Studies conducted in Western cultures have shown that self-criticism is a risk factor for mental health problems (Blatt, 2004; Luyten & Blatt, 2013; McIntyre et al., 2018). Theoretical views also acknowledge that internal schemas created related to important others in childhood subsequently influence individuals’ judgements about self (Baldwin, 1992, 2005; Baldwin & Fergusson, 2001). Moreover, this finding particularly supports learned helplessness theory (Abramson et al., 1978), hopelessness theory of depression (Alloy et al., 1988), and Beck’s (1967) theory of depression emphasizing the role of self-criticism on depression.
In this study, it was also investigated whether the links between latent variables differ according to the sex of the participants. The results showed that the link between rejection and self-criticism in the paternal model was different between females’ and males’ model, and females who perceived more rejection from their fathers are more likely to have self-criticizing tendencies than males. When this finding is evaluated in terms of the gender of the participants, it is known from existing studies that females are likely to show more internalizing behavior in the face of negative parental behaviors (Eschenbeck et al., 2007). It is also known that females give more importance to interpersonal relationships compared to males and that they have more problems with internalization when problems arise in their close relationships (Akse et al., 2004; Feinberg et al., 2000). In other words, females are more likely to have maladaptive psychological or emotional problems by doing more internalization to cope with their stress. The finding that females are more likely to show self-criticizing behaviors when they perceive paternal rejection (but not maternal) can be explained by the differentiation of gender roles for mother and father in Turkish culture. In Turkish culture, the father is expected to be disciplinarian, while the mother is expected to be more affectionate (Kagitcibasi, 1982). In such a case, while daughters are more likely to talk with their mothers about negative parenting behaviors, they are less likely to share those experiences with their fathers and thus in turn may show more self-related problems.
Although this study is important to show the links between recalled maternal and paternal parenting practices, self-criticism, and psychological distress, the findings of the study should be considered in the context of some limitations. Firstly, this is a cross-sectional study. For this reason, it is difficult to establish causal relationships between variables based on these assumptions. Secondly, this study is based on the individual’s self report and past memories about parenting practices. Although some studies suggest that childhood memories may be distorted (Hertel & Brozovich, 2010; Loftus, 1993; Schacter, 2001), some other researchers claimed that unreliability of past memories has been exaggerated (Brewin, Andrews, & Gotlib, 1993). According to Brewin et al. (1993), the basic characteristics that adults remember from their childhood are true and consistent over time. Even to them, little is known about the fact that childhood memories are disguised by mental health problems such as depression. Nevertheless, in future studies, the quality of past parent–child relationships can be assessed by the parental report, and this may lead researchers to have more unbiased data. Moreover, participants of this study were from nonclinical sample and consisted only of undergraduate students. In subsequent studies, the same model can be tested on out-of-school youth and also on clinical groups. Furthermore, research sample mostly coming from low educational levels of families might limit the generalizability of findings. Lastly, in the present study, although all the scales have acceptable reliability values, the fact that maternal and paternal rejection subscales have relatively lower reliability indices is a limitation of this study.
Despite these limitations, as emphasized in other studies (Johnson, Gliga, Jones, & Charman, 2015), the findings of the present study underscore that when people have problematic parent–child relationships in early ages, its effect is likely to be seen later in life. Thus, intervention studies focusing on treating psychological distress in emerging adult sample groups should consider the role of recalled parenting and self-criticism. Acting on the grounds that self-esteem and self-reliance are the main factors for the well-being (Gilbert, 2005) and supporting the development of these skills in individuals is of great importance. Therapists can teach young people alternative ways of self-acceptance as alternative methods of criticizing themselves. In this study, it was shown that emotional warmth in a parent–child relationship is a protective factor in terms of mental health, whereas parental rejection is a risk factor, especially for daughters. However, overprotection seems to have no effect on the self-criticism and mental health of young people. In future studies, different parental control dimensions (e.g., behavioral and psychological control) may be included in studies, and the consistency of these results can be examined in different age and sample groups. As a conclusion, this study further contributes to our understanding of how recalled maternal and paternal parenting practices directly and indirectly affect psychogical distress in emerging adults and provides crucial findings from a non-Western culture.
