J Child Fam Stud

DOI 10.1007/s10826-013-9818-y

ORIGINAL PAPER

Effects of Parental Warmth and Academic Pressure on Anxiety

and Depression Symptoms in Chinese Adolescents
Andrew S. Quach • Norman B. Epstein •
Pamela J. Riley • Mariana K. Falconier •
Xiaoyi Fang

Ó Springer Science+Business Media New York 2013

Abstract Due to traditional cultural values and a rapidly with adolescents’ depression and anxiety. Greater parental
expanding economy, Mainland Chinese adolescents com- warmth generally reduced the association between that
monly experience high levels of pressure from their parents parent’s academic pressure and adolescents’ psychopa-
to perform well in school, which may have negative effects thology symptoms, but one parent’s warmth had mixed
on adolescents’ psychological functioning. Data from 997 impacts on effects of the other’s pressure. Implications of
students in four Beijing high schools were used to examine the findings for fostering adolescent mental health in the
relationships of Chinese parents’ warmth and parental context of the Chinese concept of ‘‘training’’ children to be
pressure regarding students’ academic work with the ado- academically successful are discussed.
lescents’ levels of anxiety and depression symptoms. The
study also tested whether a parent’s level of warmth can Keywords Parenting behavior
Academic pressure

moderate negative effects on adolescent symptoms of the Adolescents
Depression
Anxiety
parent’s own academic pressure on the adolescent, as well
as negative effects of the other parent’s pressure. Gender
differences in associations between parenting behavior and Introduction
adolescent functioning were explored. Both female and
male adolescents rated mothers as displaying higher Academic achievement is emphasized in Mainland China
warmth and pressure than their fathers, but although both because of longstanding cultural values linking education
parents’ behaviors were associated with children’s mental with financial success and social status (Hesketh and Ding
health, overall fathers’ parenting (especially academic 2005), based on principles that have existed for thousands of
pressure) had stronger associations than mothers’ parenting years. In pre-modern China the highest respect and monetary
rewards were given to individuals who occupied limited
positions within the government, and access to such jobs was
A. S. Quach through formal education and passing state examinations (Hu
Department of family Studies and Community Development, 1984). Extensive pressure was placed on students to perform
Towson University, Towson, MD, USA
well. Adding to this pressure were traditions of collectivism
N. B. Epstein (&)
P. J. Riley and filial piety that emphasize interdependence among family
Department of Family Science, University of Maryland, College members and children’s responsibilities to their parents (Tri-
Park, 1142 School of Public Health, College Park, MD 20742, andis et al. 1990). Accolades that students received for per-
USA
forming well were transferred to their families (Li 2004),
e-mail: nbe@umd.edu
whereas lack of success was likely to generate criticism from
M. K. Falconier the community and shame for the child and parents (Triandis
Department of Human Development, Virginia Polytechnic et al. 1990). Today educational achievement, collectivism,
Institute and State University, Falls Church, VA, USA
and filial piety remain three bastions of Chinese culture con-
X. Fang tributing to academic pressure on children and adolescents
Beijing Normal University, Beijing, People’s Republic of China (Dello-Iacovo 2009; Tsui and Rich 2002).

123
 J Child Fam Stud

Two government mandates implemented in 1979 argu- whereas authoritarian parenting involves strict, demanding
ably increased the salience of academic achievement behavioral control. Parents who use an authoritative style
(Dello-Iacovo 2009; Hesketh and Ding 2005). The One are more likely to use control to help children develop into
Child Policy was created to reduce population growth, and responsible independent individuals, whereas those who
the Open Door Policy to enhance China’s economic status emphasize an authoritarian style tend to exert power and
(Hesketh et al. 2005; Yeung and Sung 1996). The majority restrict children’s freedom. When parenting is assessed
of parents now have one child to represent them (Tsui and with measures based on Baumrind’s model, Chinese par-
Rich 2002), exacerbating pressure on the child to succeed ents commonly score higher on the authoritarian style than
(Dello-Iacovo 2009). Additionally, rapid transition to a North American parents (Zhou et al. 2003), but their con-
market economy has intensified competition for jobs trolling behavior may be conceptualized more as caring by
(Dello-Iacovo 2009; Hesketh and Ding 2005). Chinese parents and children than by families in Western
Thus, cultural, political, and economic factors create an cultures (Chao 1994). The Chinese concept of training
environment that holds parents responsible for directing children (jiao xun) in rules of propriety is highly valued
children toward high academic performance (Shek 2005). and includes firm guidance for performing well in school
This milieu can have negative psychosocial effects on (Chao 2001; Wu and Tseng 1985). Shek (2008) identifies
students (Siu and Watkins 1997). Studies have indicated several such positive forms of Chinese parental control.
that adolescents in China experience equal or greater However, there is limited knowledge about effects that
emotional disturbance (e.g., depression, anxiety) than their Chinese parents’ forms of controlling behavior for pro-
American counterparts (Hesketh and Ding 2005; Liu et al. ducing academic achievement may have on children’s
2001). Such findings have led to increased deployment of well-being.
counselors in schools, addition of a psychological compo- There is ample precedent for focusing on parental
nent to the health education curriculum in schools, and pressure as a risk factor for adolescent emotional distress.
hotlines to assist adolescents (Hesketh and Ding 2005). Wolfradt et al. (2003) found that both parental control/
However, because parents are primary intermediaries demandingness (telling the child what to do), and pressure
between schools and students, there also is a need for more (expressing upset when the child fails to meet expectations)
research on the association between academic-related were positively correlated with German adolescents’ levels
parental behavior and adolescent mental health. Lee et al. of anxiety, whereas warmth (praise, emotional support)
(2006) argued that academic pressure from parents is a was negatively associated with anxiety. Bois et al. (2009)
factor in the documented increase in Chinese adolescent found that parental pressure was positively associated with
depression and anxiety disorders. pre-competition anxiety in French child and adolescent
Baumrind’s (1966) model of parenting styles, which athletes, whereas parental praise and understanding was
includes dimensions of warmth and control, has been associated with lower anxiety.
applied extensively in studies linking parenting behavior Parental warmth and support not only has a main effect
and child outcomes. Warmth has been defined as behaviors on child well-being; it also can moderate negative effects
that convey support, encouragement, acceptance, and that life stressors have on children’s distress. There is
affection (Baumrind 1966; Steinberg 2001). Studies in a substantial evidence that emotional support can buffer
variety of countries, including China, have found that negative effects of life stressors for people of all ages (e.g.,
higher parental warmth is associated with better child Sandler et al. 2003; Wolchik et al. 2000), and this pro-
school performance and psychological functioning (Fan tective function appears to operate when parents provide
and Chen 2001; Jeynes 2003). However a meta-analysis by warmth to children and adolescents who are exposed to
Wood et al. (2003) found inconsistent associations between academic stress. For example, Leung et al. (2010) found
control and adolescent functioning. The discrepancies that higher warmth and understanding from mothers and
might be due in part to some researchers focusing on fathers were associated with lower anxiety in Hong Kong
parental behavioral control (e.g., direct monitoring and school children, and higher paternal emotional support
limit setting) and others assessing psychological control or reduced the association between academic stress (the
pressure (e.g., use of guilt or appeals to one’s legitimate amount and difficulty of school work) and children’s
power in making requests of the child) (Gray and Steinberg anxiety symptoms.
1999). Nonetheless, it is not clear whether parental warmth and
Cultural differences in forms of control also may influ- support moderates negative effects of children’s stressors
ence findings regarding effects of parental control. In when the parent’s own behavior is the source of the stress.
Baumrind’s (1966) model, authoritative parenting is char- For example, Lessard et al. (2010) found that parents’ use
acterized by firm but equitable control involving reason- of pressure tactics to influence older adolescents’ eating
able demands and developmentally appropriate rules, habits was associated with higher adolescent negative

123
J Child Fam Stud

affect (e.g., anger) and behavioral resistance, but parental Given the prior research findings regarding links
warmth did not moderate that association. Despite the between forms of Chinese parenting and child functioning,
numerous studies on parenting dimensions, there has been and significant gaps in knowledge regarding joint influ-
limited research on combined effects on children and ences of maternal and paternal control and warmth, the
adolescents of parental pressure and warmth, both within present study used a family systems framework to test such
each parent and especially across parents (Meunier et al. joint parenting effects. First, we investigated parental
2012). control in the form of psychological pressure (e.g., high
Investigation of the effects that levels of mothers’ and expressed expectations) on their adolescent child to
fathers’ academic pressure and warmth have on the psy- achieve academically, testing whether greater parental
chological well-being of their offspring also raises the academic pressure is associated with adolescents’ psycho-
question of possible gender differences. Laible and Carlo logical distress in the forms of depression and anxiety
(2004) argued that warmth/support and control from symptoms. Second, we investigated whether parental
mothers and fathers may have different effects on ado- warmth may alleviate negative effects that parental aca-
lescent functioning. Because mothers tend to assume a demic pressure has on the children’s anxiety and depres-
more expressive role in the socialization process and sion symptoms. We took into account simultaneous
fathers a more instrumental role, with mothers focusing influences that both parents have on a child’s functioning
more on the socio-emotional well-being of their children, by examining not only whether each parent’s warmth
children are commonly emotionally closer to their mothers influences the effect of his or her own academic pressure
than their fathers, and mothers’ parenting behavior tends to on the adolescent’s distress, but also whether a parent’s
have a greater impact on their children’s psychological warmth moderates the association between the other par-
well-being (Meunier et al. 2012; Videon 2005). Shek ent’s pressure and the adolescent’s well-being. Given
(2000) found that both female and male Chinese adoles- limited prior research on simultaneous effects of maternal
cents perceived mothers to be higher than fathers on both and paternal parenting behavior on girls’ and boys’ func-
responsiveness and demandingness. In addition, females tioning, we primarily explored gender effects rather than
perceived their parents to be more demanding than males testing hypotheses regarding gender. Nonetheless, we
did. Shek (2005) also found that both girls and boys per- hypothesized that adolescents would perceive greater
ceived mothers as exerting more behavioral control pressure and greater warmth from mothers, given mothers’
(expectations for school, demandingness, monitoring of overall greater involvement in raising children.
school work, and disciplinary actions related to poor study Thus, the following hypotheses were tested:
habits) than fathers. Furthermore, Shek (2008) found that
1. Each parent’s academic pressure will be positively
greater Chinese family economic stress was associated
associated with the adolescent’s symptoms of anxiety
with adolescents’ more negative perceptions of the quality
and depression.
of their relationships with their fathers but not their
2. Each parent’s warmth will be negatively associated
mothers. This may be due to the more robust mother–child
with the adolescent’s symptoms of anxiety and
relationship, which is less susceptible to disruption from
depression.
external stressors. Shek (2008) stresses the importance of
3. The association between each parent’s pressure and the
examining separate effects of maternal and paternal
adolescent’s symptoms of anxiety and depression will
influences on child functioning. Given evidence of gender
be attenuated by that parent’s warmth.
differences in parenting, it is crucial that studies investi-
4. The association between each parent’s pressure and the
gate combined effects of maternal and paternal levels of
adolescent’s symptoms of anxiety and depression will
academic pressure and warmth on adolescent psychologi-
be attenuated by the other parent’s warmth.
cal functioning.
In addition, Dornbusch (1989) argued that mothers and
fathers have different relationships with children depend-
ing on the gender of the offspring, and it would be prudent Method
to explore the genders of both parents and children in order
to determine how gender influences associations between Participants
parenting and children’s functioning. Although some
studies found gender of parent and gender of adolescent A sample of 997 students attending four high schools in
differences in parenting behavior (Conrade and Ho 2001), Beijing, China (two ‘‘key’’ or college-preparatory schools
there has been minimal research simultaneously examining and two ‘‘ordinary’’ schools) was recruited. The two types
effects of mother’s and father’s parenting behavior on male of schools were chosen to enhance the generalizability of
and female adolescent psychological functioning. the findings. The students attended Grades 1 through 3

123
 J Child Fam Stud

Table 1 Characteristics of the four study schools This subscale has demonstrated good internal consistency
Characteristic School 1 School 2 School 3 School 4
(.89), and construct validity across a variety of cultures, test–
retest reliability averaging .62 across time periods ranging
School type Key Key Ordinary Ordinary from 3 weeks to 7 years, and evidence of good construct
Family monthly [$854 $731– \$366 \$366 validity in terms of greater parental warmth being associated
incomea $1,220 with better child psychological adjustment (see Khaleque
N of adolescents 200 305 215 277 and Rohner 2002; Rohner et al. 2003). Respondents answer
Gender % 56/44 49/51 40/60 47/53 each item twice, once to describe maternal behavior and
male/female
once to describe paternal behavior, using a 4-point scale
Mean age 16.6 16.6 17.1 16.8
ranging from 1 = almost never true to 4 = almost always
Father % college 74.5 27.1 7.3 23.7
graduate
true. Higher scores indicate greater parental warmth. The
present study used the Chinese translation provided by
Mother % college 64.3 22.4 4.6 20.2
graduate Rohner (1984). In the present study the Cronbach alphas
Parental academic pressure were .92 for mother’s warmth and .93 for father’s warmth.
Mean 39.91 42.08 52.38 48.76
SD 10.98 13.25 13.18 11.67 Parental Academic Pressure
Parental warmth
Mean 117.70 119.04 108.72 113.04 The nine-item Academic Pressure subscale from the
SD 18.15 16.58 20.91 18.53
Inventory of Parental Influence (IPI; Campbell 1994) was
Adolescent anxiety
used to assess adolescents’ perceptions of parental pressure
Mean 36.48 39.21 39.43 38.45
to achieve academically. Respondents rated the items (e.g.,
‘‘My father is never pleased with my marks.’’ ‘‘My mother
SD 5.42 5.96 5.77 5.99
pressures me too much with my homework.’’) with a
Adolescent depression
5-point scale ranging from 1 = strongly disagree to
Mean 38.91 42.88 41.99 43.02
5 = strongly agree. Higher scores indicate greater pres-
SD 6.11 7.38 7.60 8.10
sure. The IPI was demonstrated to be culturally valid in
a
Family income originally reported in Renminbi (yuan) per month. samples from five countries (Campbell 1996), and to have
Exchange rate at the time of data collection established as
an internal consistency of .83. The present study used the
$1.00 = 8.2 Renminbi
Chinese translation provided by Campbell (1994), and
Cronbach alphas in this sample were .87 for mother’s
pressure and .84 for father’s pressure.
(equivalent to Grades 10 through 12 in the US) and ranged
in age from 16 to 19. Classrooms in each school were Adolescent Depression
randomly selected for sampling, and the participants rep-
resented approximately 15 % of the total enrollment of the Depression symptoms were measured with the 27-item
four schools. Students in all schools ultimately faced Children’s Depression Inventory (CDI; Kovacs 1992),
national examinations that would determine their admis- developed for ages 8 through 17. The CDI has been used in
sion to higher education. The characteristics of the samples numerous studies worldwide. Respondents choose among
from the four schools are presented in Table 1. three options to describe their experience of each symptom;
e.g., ‘‘I am sad once in a while’’ = 0, ‘‘I am sad many
Measures times’’ = 1, ‘‘I am sad all the time’’ = 2. The CDI has
demonstrated good internal consistency and validity in
Chinese versions of the following questionnaires, which Chinese samples (e.g., Wu et al. 2010). We used the Chi-
originally were developed in English, were distributed to nese version created by the scale’s publisher, Multi-Health
the selected students from the four schools: Systems, and the Cronbach alphas in this sample were .86
for girls and for boys.
Parental Warmth
Adolescent Anxiety
Adolescents’ perceptions of parental warmth were measured
with the 19-item Warmth/Affection subscale (e.g., ‘‘My Anxiety symptoms were measured with the 37-item
mother talks to me in a warm and loving way.’’ ‘‘My father Revised Children’s Manifest Anxiety Scale (RCMAS;
makes me feel wanted and needed.’’) of the Parental Reynolds and Richmond 1985). The items have a ‘‘yes/no’’
Acceptance Rejection Questionnaire (PARQ; Rohner 1984). response format, and the total score is the number of ‘‘yes’’

123
J Child Fam Stud

responses, with higher scores indicating greater anxiety. as an exogenous variable influencing symptoms of
The scale has exhibited good reliability and validity across depression and symptoms of anxiety. Invariance of the
many cultures (e.g., Gerard and Reynolds 2004; Seligman model across schools and across gender was tested by
et al. 2004; Varela and Biggs 2006). Back translation adding cross-group constraints to the model, with results
methods were used to construct a Chinese version for the from the LM test of equality of constraints indicating
present study, and the Cronbach alphas were .79 for girls which structural paths present a statistically significant
and .82 for boys in this sample. univariate incremental in v2 values. Paths with significant
increments are those that are different across groups; in this
Procedure case, either schools or adolescent gender.

In each school, the researchers randomly selected several

classrooms from each of the three grade levels and Results
advertised the study through letters and consent forms
distributed to students and their parents, describing the The mean for symptoms of anxiety was 38.22 (SD = 5.99)
purpose and methods. The consent form emphasized that for boys and 38.76 for girls (SD = 5.85) whereas the mean
participation was voluntary and that no school staff would for depression was 41.73 for boys (SD = 7.71) and 41.92
have access to participants’ data. Students who returned for girls (SD = 7.42). According to the Mann–Whitney
their own and their parents’ consent forms to the U test, there was no significant difference between boys
researchers were given time in the classroom to complete and girls for symptoms of anxiety or depression (see
the questionnaires anonymously. Teachers left the room Table 1). The warmth mean was 55.17 (SD = 11.59) for
while the researchers administered the forms. Non-partic- boys’ ratings of fathers, 55.28 (SD = 12.08) for girls’
ipants were free to study quietly. ratings of fathers, 61.28 (SD = 10.74) for boys’ ratings of
mothers and 61.70 (SD = 11.76) for girls’ ratings of
Analytic Strategy mothers. The Mann–Whitney U test did not yield any
significant differences between boys and girls regarding
Boys’ and girls’ data were normally distributed at the mother’s warmth, father’s warmth, mother’s pressure, or
univariate level, but not at the multivariate level; Yuan father’s pressure across schools (see Table 2). Nonetheless,
et al. (2004) normalized estimate was 84.04 for boys and the Wilcoxon signed rank test did indicate significant
51.88 for girls. Consequently, we used non-parametric overall differences between perceptions of fathers’
tests. The correlations among variables were examined (M = 55.22, SD = 11.22) and mothers’ (M = 61.50,
through Spearman correlations, differences between boys’ SD = 11.28) warmth (p = .001) and between fathers’
and girls’ data were analyzed with the Mann–Whitney (M = 22.86, SD = 6.97) and mothers’ (M = 26.45,
U test for independent samples, and differences between SD = 8.47) pressure (p = .001).
mothers and fathers were examined with the Wilcoxon Spearman correlations indicated negative associations of
signed rank test for paired samples. Due to the non-nor- either parent’s warmth with both parents’ pressure and with
malcy of the multivariate distribution, the fit of the con- the adolescent’s anxiety and depression. Coefficients ran-
ceptual model to the data was assessed through robust ged from -.25 to -.32 for girls and from -.22 to -.39 for
statistics, tested separately for boys and girls through path boys. Parents’ levels of warmth were positively correlated
model analysis using EQS 6.1 and the maximum likelihood for girls (.43) and boys (.51). Parents’ levels of pressure
estimation method. The model included all paths in the also were positively correlated for girls (.64) and boys
hypotheses. It also included covariances among predictor
variables and between symptoms of anxiety and depres- Table 2 Mann–Whitney U test for independent samples: boys’ and
sion. The model was tested with the robust Yuan–Bentler girls’ scores on study variables
scaled Chi square (v2YB), an adjusted Chi square statistic
Variable Boys Girls p
used with non-normal data (Yuan and Bentler 2000).
Model fit was also evaluated with the three fit indices Mean SD Mean SD
recommended by Hu and Bentler (1999): robust Bentler’s Anxiety 38.22 5.99 38.76 5.85 .11
comparative fit index (CFI [.96), the standardized root Depression 41.73 7.71 41.92 7.42 .70
mean square residual (SRMR \.08), and the robust root Father’s warmth 55.17 11.59 55.28 12.08 .92
mean square residual of approximation (RMSEA \.06).
Mother’s warmth 61.28 10.74 61.70 11.76 .23
The need for model re-specifications was assessed through
Father’s pressure 23.11 7.15 22.63 6.80 .37
a conceptual examination of the Lagrange multiplier (LM)
Mother’s pressure 26.93 8.33 26.02 8.58 .08
test results. Adolescent age was controlled by including it

123
 J Child Fam Stud

Table 3 Spearman correlations among main variables for girls and boys
Variables 1 2 3 4 5 6

1. Anxiety – .74** -.25** -.28** .28** .32**

2. Depression .71** – -.32** -.31** .23** .27**
3. Father’s warmth -.22** -.27** – .43** -.36** -.23**
4. Mother’s warmth -.28** -.39** .51** – -.36** -.50**
5. Father’s academic pressure .38** .34** -.38** -.33** – .64**
6. Mother’s academic pressure .39** .40** -.27** -.48** .70** –
Correlations for girls are reported above the diagonal and correlations for boys are reported below the diagonal
* p \ .05 (2-tailed); ** p \ .01 (2-tailed)

(.70). Parents’ pressure was positively associated with association with boys’ symptoms of anxiety (.74) and
symptoms of anxiety for girls and boys. Coefficients ran- depression (.55) and girls’ anxiety (.57) and depression
ged from .23 to .40 (see Table 3). (.55).
The path analysis indicated optimal fit of the model to Hypothesis 2 regarding the main effect of parents’
the girls’ data as v2AB (8) = 1.39. p = .99, robust warmth was partially supported. Even though either par-
CFI = 1.00, SRMR = .02; RMSEA: .000 (.000, .011). The ent’s warmth was negatively related to the adolescent’s
covariances among predictors were all significant, and symptoms of anxiety and or depression, such associations
coefficients ranged from .04 to .75, indicating the need to either disappeared or even became positive when effects of
control for them. The only non-significant covariance was parents’ pressure and the interaction between warmth and
between mother’s warmth and the interaction between pressure were included in the model. Fathers’ warmth had a
father’s warmth and father’s pressure (.00). The multi- negative association with boys’ anxiety (-.25) and
sample analysis with cross-group constraints indicated no depression (-.15) and girls’ symptoms of anxiety (-.05),
differences for girls across schools except a difference but no significant association with girls’ depression (-.03).
between two schools in the relation between depression Mothers’ warmth had a positive association with boys’
and the interaction between mother’s warmth and father’s anxiety (.29) and depression (.15) and on girls’ anxiety
pressure. For ‘‘key’’ school #2, the path coefficient was (.07), but no significant relationship with girls’ depression
negative (-.72) and consistent with findings for the other (.02).
schools, but for ‘‘ordinary’’ school #3 it was positive (.33). Regarding the moderating effect of a parent’s warmth on
The path model analysis also showed optimal fit to the the effects of either parent’s pressure (Hypothesis 3),
boys’ data as v2AB (8) = 1.16. p = .99, robust CFI = 1.00, results were mixed. For girls, the mother’s warmth atten-
SRMR = .02; RMSEA: .012 (.000, .056). Except for the uated the positive association between her own pressure
covariance between father’s warmth and the mother’s and the adolescent’s anxiety (-.10) and between the
warmth by father’s pressure interaction (-.02), the covar- father’s pressure and the adolescent’s anxiety (-.23) and
iances among predictors were all significant, and the depression (-.30). The interaction between mother’s
absolute value of coefficients ranged from .05 to .80, warmth and mother’s pressure on girls’ depression (-.05)
indicating the need to control for the associations in the was not significant. Similarly, the father’s warmth attenu-
model. The multi-sample analysis with cross-group con- ated the negative associations between his own pressure
straints indicated no structural differences across schools. and both anxiety (-.24) and depression (-.22) in his
Standardized results of the path model analysis are daughter. Nonetheless, father’s warmth did not attenuate
presented in Fig. 1 for girls and Fig. 2 for boys. Hypothesis the effects of mother’s pressure on either type of girls’
1 regarding the relationship between a parent’s academic symptoms. Moreover, higher father warmth exacerbated
pressure and his or her adolescent’s symptoms of anxiety the negative association between mother’s pressure and
and depression was supported. Except for the non-signifi- girls’ symptoms of anxiety (.17) and had no moderating
cant path from the mother’s pressure to the boy’s anxiety effect on their depression (.01).
(.00), all paths from a parent’s pressure to the adolescent’s Similar to girls, mothers’ warmth attenuated the asso-
symptoms were significant and positive. Mothers’ pressure ciation between their own pressure and boys’ anxiety
had a positive association with boys’ symptoms of (-.28). Unlike girls, the mother’s warmth also attenuated
depression (.23) and girls’ symptoms of anxiety (.08) and the effects of her own pressure on boys’ depression (-.68).
of depression (.12). Fathers’ pressure had a positive This difference between effects for boys and girls was

123
J Child Fam Stud

Fig. 1 Standardized results for

Mother’s Warmth
adolescent girls (N = 552).
.07
Significant (p \ .05) 2
hypothesized path. Non- .02 R =.16
significant hypothesized path Mother’s Pressure
.08

Adolescent Girl’s
.12
Mother’s Warmth X Symptoms of
-.10
Mother’s Pressure Anxiety
-.05

.02
Mother’s Warmth X -.23
Father’s Pressure .71 Adolescent’s
-.30
Girl Age
-.05
Father’s Warmth .03
-.03

.57 Adolescent Girl’s

Father’s Pressure .55 Symptoms of
Depression
-.24

Father’s Warmth X -.22

Father’s Pressure 2
R =.17
.17

Father’s Warmth X .01

Mother’s Pressure
Significant (p < .05) Hypothesized Path
Non-significant Hypothesized path

Fig. 2 Standardized results for

Mother’s Warmth
adolescent boys (N = 569).
.29
Significant (p \ .05) 2
.15 R =.21
hypothesized path. Non-
significant hypothesized path Mother’s Pressure
.00

.23 Adolescent Boy’s

Mother’s Warmth X Symptoms of
-.28
Mother’s Pressure Anxiety
-.68
-.02
Mother’s Warmth X -.25 Adolescent’s
Father’s Pressure .71 Boy Age
.19
.00
-.25
Father’s Warmth -.15

.74 Adolescent Boy’s

Father’s Pressure .55 Symptoms of
Depression
-.38

Father’s Warmth X -.67

Father’s Pressure 2
R =.25
.62

Father’s Warmth X .80

Mother’s Pressure Significant (p < .05) Hypothesized Path
Non-significant Hypothesized path

123
 J Child Fam Stud

significant according to the test with group constraints to The results indicated no differences between female and
examine invariance of the structural model across boys and male adolescents’ ratings of parents’ warmth and academic
girls. The univariate increment yielded a v2 = 5.330 pressure, and both sexes rated mothers as displaying higher
(p = .02). However, greater mother warmth exacerbated warmth and pressure than fathers. This combination of
the positive association between fathers’ pressure and warmth and pressure, especially among mothers, is con-
boys’ depression (.19). Similar to girls, fathers’ warmth sistent with literature describing Chinese parents as firmly
attenuated the association between their own academic guiding children toward self-discipline and achievement,
pressure and boys’ symptoms of anxiety (-.38) and based on parents’ motivation to nurture children and max-
depression (-.67). Also similar to girls, father’s warmth imize their future success. Nonetheless, mothers’ appar-
exacerbated the relationship between mother’s pressure ently higher levels of psychological involvement with their
and boys’ anxiety (.62); however, unlike for girls, father’s children may influence the effects that their parenting
warmth also exacerbated the association between the behavior has on adolescents’ emotional functioning.
mother’s pressure and boys’ depression (.80). This differ- Whereas fathers’ warmth was negatively associated with
ence between boys and girls was statistically significant both girls’ and boys’ anxiety symptoms (as hypothesized),
according to the test with group constraints examining the mothers’ warmth was positively associated with their anx-
invariance of the structural model across boys and girls. iety. Furthermore, neither parent’s level of warmth was
The univariate increment yielded a v2 = 6.57 (p = .01). associated with girls’ depression, but whereas fathers’
In the path models, symptoms of anxiety and depression warmth was negatively associated with boys’ depression (as
were positively associated for boys (.71) and girls (.71). hypothesized), mothers’ warmth was positively associated
Age did not have any association with anxiety or depres- with boys’ depression. It is possible that in the context of the
sion for boys (anxiety: -.02; depression: .00) or girls emotionally involved mother–child relationships that are
(anxiety: .02; depression .03). Tests of models with path common in Chinese families (Chao 1994, 2001; Lau et al.
constraints did not indicate significant differences for boys 1990) maternal warmth raises the stakes that these pre-
or girls when the following paths were compared: (1) dominantly only children experience in living up to
mother’s warmth-anxiety versus father’s warmth-anxiety; responsibilities associated with filial piety. Given that
(2) mother’s warmth-depression versus father’s warmth- adolescents perceived their mothers as exerting greater
depression; (3) mother’s pressure-anxiety versus father’s academic pressure than their fathers do, these students may
pressure-anxiety; (4) mother’s pressure-depression versus have experienced greater psychological distress when given
father’s pressure-depression. No differences were found greater support from their mothers because they did not
either when comparing paths from the interaction variables want to disappoint them. Alternatively, because our cross-
toward depression or anxiety. sectional data do not indicate causal direction between
With the exception of two paths (father warmth 9 mother parenting and adolescent symptoms, it is also possible that
pressure - symptoms of depression; mother warmth 9 mother mothers of adolescents who exhibit more emotional distress
pressure - symptoms of depression), results from this two- may provide greater warmth for their offspring.
sample analysis supported the invariance of the structural model Overall, parental academic pressure was positively asso-
across boys and girls. The model explained 16 % of the variance ciated with adolescents’ depression and anxiety, as hypoth-
in anxiety symptoms and 17 % of depression symptoms for esized (with the exception of no significant association
girls and 21 % of the variance of anxiety symptoms and 25 % of between maternal pressure and boys’ anxiety); fathers’ path
depression symptoms for boys. coefficients tended to be large whereas mothers’ tended to be
small. The association between pressure and adolescents’
distress is consistent with prior research (e.g., Chen et al.
Discussion 2000; Lee et al. 2006), and the strong associations for fathers’
pressure are consistent with descriptions of Chinese fathers’
This study tested the relationships of Chinese parents’ roles as stern disciplinarians (Chao 1994, 2001; Lau et al.
warmth and pressure regarding students’ academic work 1990). The findings suggesting a parenting gender difference
with the adolescents’ levels of anxiety and depression also are consistent with those of Shek (2008) indicating that
symptoms, and it explored whether those associations in the context of external stressors adolescents perceive
varied by the gender of the parent or gender of the ado- quality of the father–child relationship more negatively than
lescent. There were no significant differences between the mother–child relationship. Regardless of positive con-
girls’ and boys’ levels of depression and anxiety symp- notations associated with the Chinese concept of ‘‘training’’
toms, but our findings indicate some differences in links children to be high achievers, parental pressure for academic
between parenting behaviors and emotional distress of girls performance is associated with adolescents’ psychological
versus boys. distress.

123
J Child Fam Stud

The findings provided partial support for the hypothesis of students from urban families whose parents allowed
that greater parental warmth can reduce the association them to complete forms describing their parenting behavior
between parental academic pressure and adolescents’ psy- and the child’s psychological symptoms. Generalizability
chological distress, with some notable exceptions. Overall, and relevance were enhanced by sampling from both
mothers’ and fathers’ warmth helped to reduce the effects ‘‘ordinary’’ schools and ‘‘key’’ schools that focus on col-
of their own pressure on boys’ depression and anxiety, as lege preparation, and the students were at a crucial point as
well as girls’ anxiety. Thus, a parent’s emotional support they were facing college entrance examinations. Another
seems to create a positive context in which the adolescent limitation involved the reliance on adolescents’ reports for
interprets the parent’s pressure as less threatening. How- information about both their parents’ behavior and their
ever, this attenuating effect only occurred in some instan- own symptoms. In addition, it would have been easier to
ces for one parent’s warmth and the other’s pressure. interpret the associations between parenting behaviors and
Mothers’ warmth reduced the associations between fathers’ adolescent symptoms if more information had been gath-
pressure and girls’ anxiety and depression, as well as boys’ ered about meanings the adolescents attached to the pres-
anxiety. This is a typical stress-buffering effect of social sure and warmth that they perceived their parents
support. However, mothers’ warmth exacerbated the exhibiting. For example, the Chinese concept of ‘‘training’’
association between fathers’ pressure and boys’ depression. suggests that parents and children perceive parental control
Perhaps this combination of parental behaviors places boys as reflecting caring, but in the present study there was no
in a dilemma that makes them feel hopeless about being a measure of how adolescents’ perceived their parents’
good child for their mother and failing to please a actions. Finally, this study was cross-sectional, preventing
demanding father. In the only difference across schools, identification of causal processes between parenting
mothers’ warmth also exacerbated the association between behavior and adolescent functioning. Assessing parents’
fathers’ pressure and girls’ depression in school #3, an behavior and adolescents’ symptoms at multiple points
‘‘ordinary’’ school with the lowest percentages of parents during their high school years would allow a better test of a
who were college graduates and the highest adolescent developmental psychopathology model involving parent–
ratings for parental academic pressure. In that context in adolescent relations.
which parents may have been pressuring their children to
achieve a better level of academic success than their own, Implications
mothers’ warmth also may have led daughters to feel dis-
couraged about the prospect of disappointing them. Our findings support research (Meunier et al. 2012; Simons
Fathers’ warmth also exacerbated the associations between and Conger 2007) indicating the importance of assessing
mothers’ pressure and boys’ anxiety, boys’ depression, and the combined effects of maternal and paternal behaviors on
girls’ anxiety, and had no effect on the association between adolescent outcomes, examining those behaviors simulta-
mothers’ pressure and girls’ depression. The overall failure neously in the same model to determine joint effects on
of fathers’ warmth to buffer negative effects of mothers’ their children. Based on the results of this study it seems
academic pressure is consistent with evidence that Chinese important for both parents in Chinese families to be sup-
mothers play a more central role than fathers in their portive of sons and daughters when setting high academic
children’s lives, such that fathers’ support cannot reduce expectations, in order to protect their children’s psycho-
the stress of mothers’ pressure. logical health. The findings suggest either that parents may
Overall, associations between parenting behavior and need assistance in developing forms of ‘‘training’’ that
adolescent psychological distress were smaller for girls provide guidance and motivation without high levels of
than for boys. Although the One Child Policy and changing pressure, or those who use high levels of pressure need to
social norms have increased the importance that families recognize the importance of balancing pressure with
attach to a college degree for girls, the longstanding tra- warmth. However, given the Chinese emphasis on filial
dition of valuing sons more than daughters as the pathway piety, parents also need to be aware that adolescents may
to continuing the family’s lineage may enhance sons’ experience parental warmth as another form of pressure to
emotional responses to parents’ levels of expressed warmth meet parents’ expectations regarding the characteristics of
and pressure, even when parents express similar levels of a ‘‘good child,’’ which include a strong academic focus and
warmth and pressure toward daughters and sons. obedience (Shek and Chan 1999). An issue for the educa-
tion system is where and how parents can be guided in
Limitations of the Study developing styles that simultaneously encourage achieve-
ment and protect children’s psychological well-being.
The conclusions that one can draw from this study are Although we found evidence that mothers and fathers
limited to some extent by the characteristics of the sample have some different effects on children’s functioning, both

123
 J Child Fam Stud

parents’ behaviors were associated with children’s mental 2: Instruments for children and adolescents (3rd ed., pp. 63–79).
health. The notion of parental responsibility can be useful Mahwah, NJ: Erlbaum.
Gray, M. R., & Steinberg, L. (1999). Unpacking authoritative
when considering therapeutic interventions with Chinese parenting: Reassessing a multidimensional construct. Journal
families. Parents who bring a child to therapy due to of Marriage and the Family, 6, 574–587. doi:10.2307/353561.
school-related stresses may want the therapist to focus on Hesketh, T., & Ding, Q. J. (2005). Anxiety and depression in
the child’s behavior, but the therapist may need to engage adolescents in urban and rural China. Psychological Reports, 96,
435–444. doi:10.2466/pr0.96.2.435-444.
parents in modifying their approaches to motivating the Hesketh, T., Lu, L., & Xing, Z. W. (2005). The effect of China’s one-
child. Because parents who adhere to traditional values child family policy after 25 years. The New England Journal of
may perceive requests to change their behavior as threat- Medicine, 353, 1171–1176. doi:10.1056/NEJMhpr051833.
ening a loss of face, therapists can collaborate with them, Hu, C. T. (1984). The historical background: Examinations and
control in pre-modern China. Comparative Education, 20, 7–26.
asking what they think could be done to help their child doi:10.1080/0305006840200103.
deal with depression, and anxiety. One can work within the Hu, L. T., & Bentler, P. M. (1999). Cutoff criteria for fit indexes in
Chinese value system by reframing a change in parenting covariance structure analysis: Conventional criteria versus new
as a demonstration of concern for the child’s well-being. alternatives. Structural Equation Modeling, 6, 1–55.
Jeynes, W. H. (2003). A meta-analysis: The effects of parental
involvement on minority children’s academic achievement.
Acknowledgments Funding for this project was from Grant # Education and Urban Society, 35, 202–218. doi:10.1177/
11JZD037 awarded to Dr. Xiaoyi Fang by the Key Projects of Phi- 0013124502239392.
losophy and Social Sciences Research, China Ministry of Education. Khaleque, A., & Rohner, R. P. (2002). Reliability of measures assessing
the pancultural association between perceived parental accep-
tance-rejection and psychological adjustment: A meta-analysis of
cross-cultural and intracultural studies. Journal of Cross-Cultural
References Psychology, 33, 87–99. doi:10.1177/0022022102033001006.
Kovacs, M. (1992). Children’s Depression Inventory manual. North
Baumrind, D. (1966). Effects of authoritative parental control on child Tonawanda, NY: Multi-Health Systems.
behavior. Child Development, 37, 887–907. Laible, D. J., & Carlo, G. (2004). The differential relations of
Bois, J. E., Lalanne, J., & Delforge, C. (2009). The influence of maternal and paternal support and control to adolescent social
parenting practices and parental presence on children’s and competence, self-worth, and sympathy. Journal of Adolescent
adolescents’ pre-competitive anxiety. Journal of Sports Sciences, Research, 19, 759–782. doi:10.1177/0743558403260094.
27, 995–1005. doi:10.1080/02640410903062001. Lau, S., Lew, W. J. F., Hau, K. T., Cheung, P. C., & Berndt, T. J. (1990).
Campbell, J. R. (1994). Differential socialization in mathematics Relations among perceived control, warmth, indulgence, and
achievement: Cross-national and cross-cultural perspectives. family harmony of Chinese in mainland China. Developmental
International Journal of Education Research, 21, 685–696. Psychology, 26, 674–677. doi:10.1037/0012-1649.26.4.674.
Campbell, J. R. (1996). Cross-national retrospective studies of Lee, M. T. Y., Wong, B. P., Chow, B. W. Y., & McBride-Chang, C.
mathematics Olympians. International Journal of Educational (2006). Predictors of suicide ideation and depression in Hong
Research, 25, 485–496. Kong adolescents: Perceptions of academic and family climates.
Chao, R. K. (1994). Beyond parental control and authoritarian Suicide and Life-Threatening Behavior, 36, 82–96. doi:10.1521/
parenting style: Understanding Chinese parenting through the suli.2006.36.1.82.
cultural notion of training. Child Development, 65, 1111–1119. Lessard, J., Greenberger, E., & Chen, C. (2010). Adolescents’
doi:10.1111/j.1467-8624.1994.tb00806.x. response to parental efforts to influence eating habits: When
Chao, R. K. (2001). Extending research on the consequences of parenting parental warmth matters. Journal of Youth and Adolescence, 39,
style for Chinese Americans and European Americans. Child 73–83. doi:10.1007/s10964-008-9376-6.
Development, 72, 1832–1843. doi:10.1111/1467-8624.00381. Leung, G. S. M., Yeung, K. C., & Wong, D. F. K. (2010). Academic
Chen, X., Liu, M., & Li, D. (2000). Parental warmth, control, and stressors and anxiety in children: The role of paternal support.
indulgence and their relations to adjustment in Chinese children. Journal of Child and Family Studies, 19, 90–100. doi:10.1007/
A longitudinal study. Journal of Family Psychology, 14, s10826-009-9288-4.
401–419. doi:10.1037/0893-3200.14.3.401. Li, L. (2004). Parental expectations of Chinese immigrants: A folk
Conrade, G., & Ho, R. H. (2001). Differential parenting styles for theory about children’s school achievement. Race, Ethnicity, and
fathers and mothers. Australian Journal of Psychology, 53(1), Education, 20, 167–183. doi:10.1080/1361332042000234286.
29–35. doi:10.1080/00049530108255119. Liu, X., Kurita, H., Uchiyama, M., Okawa, M., Liu, L., & Ma, D.
Dello-Iacovo, B. (2009). Curriculum reform and ‘quality education’ (2001). Life events, locus of control, and behavioral problems
in China: An overview. International Journal of Educational among Chinese adolescents. Journal of Clinical Psychology, 56,
Development, 29, 241–249. doi:10.1016/j.bbr.2011.03.031. 1565–1577. doi:10.1002/1097-4679.
Dornbusch, S. M. (1989). The sociology of adolescence. Annual Meunier, J. C., Bisceglia, R., & Jenkins, J. M. (2012). Differential
Review of Sociology, 15, 233–259. doi:10.1146/annurev.so.15. parenting and children’s behavioral problems: Curvilinear asso-
080189.001313. ciations and mother–father combined effects. Developmental
Fan, X., & Chen, M. (2001). Parental involvement and students’ Psychology, 48, 987–1002. doi:10.1037/a0026321.
academic achievement: A meta-analysis. Educational Psychol- Reynolds, C. R., & Richmond, B. O. (1985). Revised Children’s
ogy Review, 13, 1–22. doi:10.1023/A:1009048817385. Manifest Anxiety Scale. Los Angeles: Western Psychological
Gerard, A. B., & Reynolds, C. R. (2004). Characteristics and Services.
applications of the Revised Children’s Manifest Anxiety Scale Rohner, R. P. (1984). Handbook for the study of parental acceptance
(RCMAS). In M. E. Maruish (Ed.), The use of psychological and rejection. Unpublished manscript, Center for the Study of
testing for treatment planning and outcomes assessment: Volume Parental Acceptance and Rejection, University of Connecticut.

123
J Child Fam Stud

Rohner, R. P., Khaleque, A., & Cournoyer, D. E. (2003). Cross- Tsui, M., & Rich, L. (2002). The only child and educational opportunity
national perspectives on parental acceptance-rejection theory. for girls in urban China. Gender and Society, 16, 74–92.
Marriage and Family Review, 35, 85–105. doi:10.1300/ Varela, R. E., & Biggs, B. K. (2006). Reliability and validity of the
J002v35n03_06. Revised Children’s Manifest Anxiety Scale (RCMAS) across
Sandler, I. N., Ayers, T. A., Wolchik, S. A., Tein, J.-Y., Kwok, O.-M., samples of Mexican, Mexican American, and European Amer-
Haine, R. A., et al. (2003). The family bereavement program: ican children: A preliminary investigation. Anxiety, Stress &
Efficacy evaluation of a theory-based prevention program for Coping: An International Journal, 19, 67–80. doi:10.1080/
parentally bereaved children and adolescents. Journal of Con- 10615800500499727.
sulting and Clinical Psychology, 7, 587–600. doi:10.1037/0022- Videon, T. M. (2005). Parent–child relations and children’s psycho-
006X.71.3.587. logical well-being: Do dads matter? Journal of Family Issues,
Seligman, L. D., Ollendick, T. H., Langley, A. K., & Baldacci, H. B. 26, 55–78. doi:10.1177/0192513X04270262.
(2004). The utility of measures of child and adolescent anxiety: Wolchik, S. A., Wilcox, K. L., Tein, J.-Y., & Sandler, I. N. (2000).
A meta-analytic review of the Revised Children’s Anxiety Scale, Maternal acceptance and consistency of discipline as buffers of
the State-Trait Anxiety Inventory for Children, and the Child divorce stressors on children’s psychological adjustment prob-
Behavior Checklist. Journal of Clinical Child and Adolescent lems. Journal of Abnormal Child Psychology, 28(1), 87–102.
Psychology, 33, 557–565. doi:10.1207/s15374424jccp3303_13. doi:10.1023/A:1005178203702.
Shek, D. T. L. (2000). Differences between fathers and mothers in the Wolfradt, U., Hempel, S., & Miles, J. N. V. (2003). Perceived
treatment of, and relationship with, their teenage children: parenting styles, depersonalisation, anxiety and coping behav-
Perceptions of Chinese adolescents. Adolescence, 35, 135–145. iour in adolescents. Personality and Individual Differences, 34,
Shek, D. T. L. (2005). Perceived parental control and parent–child 521–532. doi:10.1016/S0191-8869(02)00092-2.
relational qualities in Chinese adolescents in Hong Kong. Sex Wood, J. J., McLeod, B. D., Sigman, M., Hwang, W. C., & Chu, B. C.
Roles, 53, 635–646. doi:10.1007/s11199-005-7730-7. (2003). Parenting and childhood anxiety: Theory, empirical
Shek, D. T. L. (2008). Economic disadvantage, perceived family life findings, and future directions. Journal of Child Psychology and
quality, and emotional well-being in Chinese adolescents: A Psychiatry, 44, 134–151. doi:10.1111/1469-7610.00106.
longitudinal study. Social Indicators Research, 85, 169–189. Wu, W. F., Lu, Y. B., Tan, F. R., & Yao, S. Q. (2010). Reliability and
Shek, D. T. L., & Chan, L. K. (1999). Hong Kong Chinese parents’ validity of the Chinese version of Children’s Depression
perceptions of the ideal child. The Journal of Psychology, 133, Inventory. Chinese Mental Health Journal, 24, 775–779.
291–302. doi:10.1080/00223989909599742. Wu, D. Y. H., & Tseng, W. (1985). Introduction: The characteristics of
Simons, L. G., & Conger, R. D. (2007). Linking mother–father Chinese culture. In W. Tseng & D. Y. H. Wu (Eds.), Chinese
differences in parenting to a typology of family parenting styles culture and mental health (pp. 3–13). New York: Academic Press.
and adolescent outcomes. Journal of Family Issues, 28, 212–241. Yeung, Y. M., & Sung, Y. (Eds.). (1996). Shanghai: Transformation
doi:10.1177/0192513X06294593. and modernization under China’s open door policy. Hong Kong:
Siu, A., & Watkins, D. (1997). Coping with stress in Hong Kong: An Chinese University Press.
investigation of the influence of gender, age, and the self- Yuan, K.-H., & Bentler, P. M. (2000). Inferences on correlation
concept. Psychologia: An International Journal of Psychology in coefficients in some classes of nonnormal distributions. Journal
the Orient, 40, 59–66. of Multivariate Analysis, 72, 230–248.
Steinberg, L. (2001). We know some things: Parent–adolescent Yuan, K.-H., Lambert, P. L., & Fouladi, R. T. (2004). Mardia’s
relationships in retrospect and prospect. Journal of Research on multivariate kurtosis with missing data. Multivariate Behavioral
Adolescence, 11, 1–19. doi:10.1111/1532-7795.00001. Research, 39, 413–437.
Triandis, H. C., McCusker, C., & Hui, C. H. (1990). Multimethod Zhou, Z., Peverly, S. T., Xin, T., Huang, A. S., & Wang, W. (2003).
probes of individualism and collectivism. Journal of Personality School adjustment of first-generation Chinese-American adoles-
and Social Psychology, 59, 1006–1020. doi:10.1037/0022-3514. cents. Psychology in the Schools, 40, 71–84. doi:10.1002/pits.
59.5.1006. 10070.

123