Learning and Instruction 19 (2009) 506e512

www.elsevier.com/locate/learninstruc

Mothers’ parenting practices and adolescents’ learning from their

mistakes in class: The mediating role of adolescent’s self-disclosure
Guy Roth*, Tali Ron, Mordechai Benita
Education Department, Ben-Gurion University of the Negev, P.O. Box 653, 84105 Beer-Sheva, Israel
Received 26 March 2008; revised 1 August 2008; accepted 12 October 2008

Abstract

This study examined 126 students’ (14e16 years of age; 66 females) perceptions of self-disclosure to their mothers with respect to their
mistakes in class activities. Specifically, we hypothesized that self-disclosure would predict adolescents’ ability to learn from mistakes they made
in classroom tasks. In addition, we hypothesized that perceived mothers’ love withdrawal would correlate negatively with adolescents’ self-
disclosure, whereas perceived autonomy support would correlate positively with self-disclosure. Further, we hypothesized that the effect of
mothers’ parenting practices on adolescents’ ability to learn from their mistakes would be mediated by adolescents’ self-disclosure of their
school experiences. Results, using SEM analyses, showed the importance of mothers’ autonomy support for adolescents’ self-disclosure and
learning from their mistakes in classroom tasks.
Ó 2008 Elsevier Ltd. All rights reserved.

Keywords: Parenting practices; Self-disclosure; Love withdrawal; Autonomy support; Self-determination theory

1. Introduction and the role of adolescents’ voluntary self-disclosure for

externalizing problem behaviors (Fletcher, Steinberg, &
The important role of the parentechild relationship in Williams-Wheeler, 2004; Kerr & Stattin, 2000; Soenens,
children’s adaptation to school has been emphasized by many Vansteenkiste, Luyckx, & Goossens, 2006; Stattin & Kerr,
researchers (Cawan, Cowan, Ablow, Johnson, & Measelle, 2000). While emphasizing the importance of children’s self-
2005; Dornbusch, Ritter, Leiderman, Roberts, & Fraleigh, disclosure and parental knowledge, researchers rejected the
1987; Grolnick & Ryan, 1989; Grolnick, Ryan, & Deci, 1991; notion of parental direct control of adolescents’ behavior as an
Parsons, Adler, & Kaczala, 1982; Thompson, 2004). Research effective proactive strategy for parents, because adolescents
has mainly focused on parental practices such as warmth, spend a large portion of their time in places where parents are
responsiveness, involvement (Grolnick & Ryan, 1989; Mac- not present; thus, children’s behavior cannot be directly
coby & Martin, 1983; Mattanah, 2001; Puustinen, Lyyra, controlled by parents (Stattin & Kerr, 2000; Wells & Rankin,
Metsapelto, & Pulkkinen, 2008), structure (Grolnick, Deci, & 1988). In line with this argument, Stattin and Kerr (2000)
Ryan, 1997), and autonomy support (Assor, Kaplan, & Roth, found that adolescents’ voluntary self-disclosure was
2002; Grolnick et al., 1991; Roth, Assor, Kanat-Maymon, & a stronger predictor of parental knowledge and of adolescents’
Kaplan, 2007; Soenens & Vansteenkiste, 2005). deviant behavior than were parents’ active attempts to super-
In the last few years, research on parenting has emphasized vise their children’s behavior. Based on this finding, one may
the role of parental knowledge of adolescents’ whereabouts claim that parents’ behavior is of a secondary importance in
explaining the links between parenting and children’s deviant
behavior.
* Corresponding author. Tel.: þ972 8 6461884; fax: þ972 77 3353030. In a recent study, Soenens et al. (2006) argued that it is
E-mail address: roth@bgu.ac.il (G. Roth). premature to conclude that parents have relatively little impact

0959-4752/$ - see front matter Ó 2008 Elsevier Ltd. All rights reserved.
doi:10.1016/j.learninstruc.2008.10.001
 G. Roth et al. / Learning and Instruction 19 (2009) 506e512 507

on the development of problem behaviors, because parents in class activities. Hence, following Soenens et al. (2006), we
may protect their children from antisocial behavior by creating predicted that love withdrawal (one component of psycho-
a family climate that promotes children’s disclosure. This logical control) in relation to the child’s failures and lack of
argument was also raised by Fletcher et al. (2004) and by investment in school would inhibit the child’s self-disclosure,
Stattin and Kerr (2000). In a well designed study, Soenens whereas autonomy supportive parenting would promote it.
et al. (2006) demonstrated that adolescents’ self-disclosure and Assor, Roth, and Deci (2004) found that parental condi-
parental knowledge serve as intervening variables in the tional regard, which involves love withdrawal together with its
associations between parental practices and adolescents’ mirror image (i.e., provision of love contingent on children’s
problem behaviors. This research replicated Stattin and Kerr’s behavior), predicts children’s sense of disapproval by parents,
(2000) findings regarding the important role of children’s self- which, in turn, predicts children’s resentment toward parents.
disclosure in predicting parental knowledge. In a subsequent study, Roth, Assor, Niemiec, Ryan, and Deci
Up to now, research has not yet explored the associations (2007) demonstrated that resentment toward parents is pre-
between self-disclosure and school engagement, despite these dicted primarily by love withdrawal even when controlling for
concepts’ seeming relevance for adolescents’ school adapta- contingent love provision. Thus, it seems reasonable to assume
tion, because the parent is not present at school and therefore that adolescents who expect their parents to react to academic
cannot directly supervise the child’s behavior. The present failures by withdrawing their love (which was shown to be
study aimed to shed some light on this association. Specifi- associated with rejection and disapproval) might refrain from
cally, the present study focused on the mediating role of disclosure of difficulties at school.
adolescents’ self-disclosure concerning their school experi- In contrast, we hypothesized that parents’ autonomy
ences, in the link between mothers’ parenting practices and support may predict adolescents’ self-disclosure of school
adolescents’ ability to learn from the mistakes they make experience. Autonomy supportive parenting has been found to
during classroom activities. predict positive school outcomes such as academic compe-
tence, autonomous motivation, self-regulation, and achieve-
1.1. Family climate and promotion of adolescents’ ment (Grolnick & Ryan, 1989; Grolnick et al., 1991; Soenens
self-disclosure & Vansteenkiste, 2005; Vansteenkiste, Zhou, Lens, & Soe-
nens, 2005). Within self-determination theory (SDT; Deci &
In line with Kerr and Stattin’s (2000) argument that the Ryan, 2000), autonomy support is defined as an encourage-
relational part of parenting may be particularly predictive of ment of integrated internalization, self-initiation, and the
adolescents’ self-disclosure, Soenens et al. (2006) focused on promotion of volitional functioning (Roth & Deci, 2008;
the links between psychological control and responsiveness to Vansteenkiste et al., 2005) by taking the child’s perspective,
children’s self-disclosure. Soenens et al. (2006) found that validating his or her feelings and thoughts, encouraging
parental responsiveness and parental structure were positively choice, and providing meaningful rationale and relevance
associated with adolescents’ self-disclosure; likewise, through (Assor et al., 2002; Roth, 2008; Vansteenkiste et al., 2005; for
the mediation of adolescents’ self-disclosure, these parental a review, see Grolnick et al., 1997). Although autonomy
indices were associated with fewer problem behaviors. On the supportive parenting was found to be related to parental
other hand, parents’ psychological control was found to be warmth and acceptance, it also showed a unique effect above
negatively related to adolescents’ self-disclosure. Thus, the and beyond parental warmth in relation to children’s inter-
research demonstrated that adolescents’ self-disclosure can be nalization and behavior (Roth, 2008; Roth, Assor, Kanat-
predicted by parenting practices. Maymon, et al., 2007; Roth, Assor, Niemiec, et al., 2007). It
Soenens et al. (2006) argued that children’s self-disclosure seems that in relation to adolescents’ self-disclosure, the most
is promoted by warm, accepting, and empathic (i.e., respon- relevant autonomy supportive behaviors are those in which
sive) parenting, whereas intrusive parenting such as psycho- parents attempt to take the child’s perspective and validate his
logical control that refers to manipulative parenting practices or her feelings. Therefore, the present research focused on
(e.g., guilt induction, shaming, and love withdrawal; Barber, these two dimensions of autonomy support.
1996; Barber, Stolz, & Olsen, 2005) inhibits self-disclosure.
Thus, the results supported the hypothesis that children of
psychologically controlling parents, who expect their parents 1.2. Self-disclosure and ability to learn from
to respond intrusively when confronted with antisocial mistakes in class
behavior, would try to avoid such negative reactions by
refraining from voluntary self-disclosure. Current theory and research in motivation indicate that,
Unlike the former studies that focused on adolescents’ although repeated academic failures are clearly undesirable,
externalizing problem behaviors, the present study focused on a temporary failure that results in effective coping is often an
adolescents’ academic engagement. We were specifically important experience that enhances children’s motivation,
interested in adolescents’ self-disclosure concerning their learning, and emotional development (Alfi, Katz, & Assor,
school experiences and in a positive (adaptive) outcome of that 2004). The main obstacles to adaptive coping with temporary
ability to share one’s difficulties at school with one’s parents, failures are likely to be negative feelings that these situations
namely, the adolescent’s ability to learn from his/her mistakes may elicit (Alfi et al., 2004).
508 G. Roth et al. / Learning and Instruction 19 (2009) 506e512

In a well designed series of studies, McGregor and Elliot mistakes in class activities would be mediated by the adoles-
(2005) presented findings supporting the hypothesis that cents’ higher self-disclosure (Hypothesis 4).
shame is the core emotion of fear of failure. In addition,
participants who were high on fear of failure reported that they 2. Method
would be less likely to tell their mother and father about their
failures. Thus, it seems reasonable to hypothesize that child- 2.1. Participants e design
ren’s voluntary self-disclosure may predict children’s attempts
to learn from their own mistakes made during classroom Participants were 126 adolescents (66 girls) from two
activities, inasmuch as the child feels supported because there secondary schools in a midsized Israeli city. Ages ranged from
is nothing to fear about revealing failures. The ability to share 14 to 16 years, with a mean of 14.77 years (SD ¼ .44). All the
experiences of failures at school with parents is a possible participants came from intact families, and according to the
marker for the ability to contain failures and to cope with them Israeli Ministry of Education these schools serve middle class
without consistent attempts to deny or avoid them. Moreover, and lower-middle class populations.
the knowledgeable parent may be a possible resource to cope As was required by the Israeli Ministry of Education, active
emotionally and functionally with the temporary failure. informed consent was obtained from the adolescents, and
Furthermore, this open communication and sharing with passive informed consent was required from parents. The latter
parents can be transferred by children to their relationships procedure entailed parents’ receipt of a letter from the
with teachers and other socialization agents, who may also researcher providing information about the purposes of the
help the child to deal with the temporary difficulty. Thus, the study and its method; parents were asked to complete a form if
ability to disclose difficulties may be a primary resource for they did not wish their child to participate in the study. Only
the ability to benefit from social support at times of need. 2% of the parents did not allow their children to participate in
the study.
The participants completed questionnaires in class in two
1.3. Hypotheses sessions separated by two weeks. In the first session, partici-
pants completed scales involving their perceptions of their
The hypotheses are presented schematically in Fig. 1. mothers’ parenting practices (Love withdrawal and autonomy
Specifically, we hypothesized that the adolescents’ perceptions support), and two weeks later they completed the self-disclosure
of their mothers as using love withdrawal in response to the and learning from mistakes scales. The first session lasted
adolescents’ academic failures would be negatively associated about 10 min, and the second session lasted about 20 min. The
with adolescents’ perceptions of voluntary self-disclosure, and questionnaires were administered when the teachers were not
with their ability to learn from mistakes made during class present in the classroom.
activities (Hypothesis 1). Moreover, we hypothesized that the
expected negative correlation between perceived maternal love 2.2. Measures
withdrawal and children’s attempts to learn from their
mistakes in class activities would be mediated by the adoles- All measures were completed by the adolescents, and all
cents’ lower self-disclosure (Hypothesis 2). items were scored on 6-point Likert-type scales, ranging from
In contrast, we hypothesized that perceptions of maternal 1 (strongly disagree) to 6 (strongly agree).
autonomy support in response to the adolescents’ academic
failures would be positively associated with adolescents’ 2.2.1. Perceptions of mothers’ love withdrawal
perceptions of voluntary self-disclosure and learning from This 5-item measure was based on Assor et al. (2004) and
mistakes (Hypothesis 3). Furthermore, we hypothesized that Roth, Assor, Niemiec, et al. (2007) studies. It is important to
the expected positive correlation between perceived maternal note that this measure differed from the three love-withdrawal
autonomy support and children’s attempts to learn from their items in Barber et al.’s (2005) measure of general psycho-
logical control because the present measure was domain-
specific for academic achievement and effort. Items included:
Love ‘‘If I do poorly in school my mother will ignore me for
withdrawal a while’’ and ‘‘I often feel that I would lose much of my
mother’s affection if (or when) I do poorly at school’’. Cron-
Self- Learning from bach’s alpha for the present sample was .90.
disclosure mistakes
2.2.2. Perceptions of mothers’ autonomy support
Autonomy This 4-item measure was adopted in part from Grolnick
support et al., (1991) and in part from Roth, Assor, Niemiec, et al.
(2007) studies. The Grolnick et al. (1991) scale is a general
Fig. 1. The hypothesized model: Self-disclosure as mediating variable for the scale, whereas the scale used in the present study specifically
relation between mothers’ parenting practices and adolescents’ learning from reflects children’s perceptions of autonomy support toward
their mistakes in classroom activities. learning at school. Given the focus of the present study on
 G. Roth et al. / Learning and Instruction 19 (2009) 506e512 509

children’s self-disclosure, the measure focused on two partial correlations, we controlled for Social Desirability Bias
autonomy supportive themes: (a) Two items referring to the while computing the correlations for Autonomy Support. The
extent to which the mother takes the respondent’s perspective table reveals that, as expected, negative correlations emerged
(e.g., ‘‘When my mother feels that I am not studying enough between Love Withdrawal and the two outcome variables
she really tries to understand why’’) and (b) two items refer- (Self-Disclosure and Learning from Mistakes), whereas posi-
ring to the extent to which the mother validates the respon- tive correlations emerged between Autonomy Support and the
dents’ perspective (e.g., ‘‘My mother understands that there two outcome variables. However, the correlation between
are other things I am interested in besides learning at school’’). Love Withdrawal and Learning from Mistakes was not
Cronbach’s alpha in the present sample was .74. significant ( p ¼ .09).
To examine the mediating role of self-disclosure in the link
2.2.3. Adolescents’ self-disclosure between mothers’ parenting practices and children’s ability to
This 5-item scale tapping adolescents’ voluntary self- learn from mistakes, a structural equation modeling (SEM)
disclosure was based on the general measure developed by analysis with latent variables was conducted. First,
Stattin and Kerr (2000) (Kerr & Stattin, 2000), although for a measurement model was tested (confirmatory factor analyses
the present study we modified the items to specifically tap self- using SEM), and then the hypothesized model was examined.
disclosure in relation to school experiences. A sample item Additionally, we compared the hypothesized model to models
read: ‘‘I spontaneously tell my mother about difficulties I that include, in addition to the indirect effects, direct effects of
encounter in class’’. Cronbach’s alpha for the present sample the parenting practices on the outcome measures. Each
was .87. comparison was conducted separately, allowing comparison of
nested models.
To test the hypothesized model as presented in Fig. 1, we
2.2.4. Learning from mistakes in class
used AMOS 5.0 (Arbuckle & Wothke, 2003) with maximum
This 3-item scale was developed for the purpose of the
likelihood estimation. Latent variables were created by using
present study and measured the child’s attempts to learn from
the items as indicators. To assess the fit of the model to the
his/her mistakes in class activities. The items were: ‘‘Some-
data, we used the ratio of chi-square to degrees of freedom,
times I go through the mistakes I made in a paper or an exam
incremental fit index (IFI; Bollen, 1989), comparative fit index
in order to understand the subject better’’, ‘‘When I am not
(CFI; Bentler, 1990), root mean square error of approximation
satisfied with the mark I got in an exam, I ask the teacher to
(RMSEA; Browne & Cudeck, 1993), and the standardized root
explain where I was wrong, so I can learn from my mistake’’,
mean square residual (SRMR). Acceptable fit would be indi-
and ‘‘When I give a wrong answer in class, I follow the lesson
cated by a ratio of c2 to df of less than 2 (Carmines & McIver,
carefully in order to understand what my mistake was’’.
1981), a RMSEA less than .08, a SRMR less than .08, and the
Cronbach’s alpha for the present sample was .71.
other fit indices of .90 or above (Browne & Cudeck, 1993;
Hoyle, 1995).
2.2.5. Adolescents’ social desirability bias First, we tested the fit of the measurement model. The
A 15-item version of Crowne and Marlowe (1964) scale results support the hypothesized factor structure. The fit
was used to control for participants’ tendency to report indices were adequate: c2(113, N ¼ 126) ¼ 167.55, p < .01;
dishonestly about the sensitive issues examined in this c2/df ¼ 1.43; and CFI, IFI, RMSEA, and SRMR of .97, .97,
research. A sample item was: ‘‘No matter who I am talking to I .04, and .04, respectively. Fig. 2 presents the results for the
am always a good listener.’’ Cronbach’s alpha for the present structural model. The results supported, in the main, the
sample was .75. hypothesized relations. Most of the path coefficients were
significant and in the hypothesized direction. The model
3. Results showed an acceptable fit to the data: c2(115,
N ¼ 126) ¼ 168.52, p < .01; c2/df ¼ 1.47; and CFI, IFI,
Table 1 presents descriptive statistics and Pearson correla- RMSEA, and SRMR, of .98, .97, .04, and .04, respectively.
tions for the study variables. Social Desirability Bias (not The correlation between Love Withdrawal and Autonomy
presented in the table) was found to be correlated only with Support was added in order to test unique effects of the
Autonomy Support, r ¼ .18, p < .05; therefore, by using parental practices.
In the next phase of data analyses, we compared the
Table 1
hypothesized model (in which the relations between mothers’
Descriptive statistics and correlations among the study variables. parental practices and learning from mistakes are indirect, that
M SD 1 2 3 4
is, mediated by self-disclosure), to two models that included
a direct path from each independent variable (the two parental
1. Love withdrawal 1.65 .66 e
2. Autonomy support 4.35 1.08 .33** e
practices, that is, Love Withdrawal and Autonomy Support) to
3. Self-disclosure 4.19 1.09 .20* .23* e Learning from Mistakes. Each direct path was added sepa-
4. Learning 4.03 1.04 .16ns .18* .28** e rately, allowing us to compare goodness of fit of nested
from mistakes models. Results showed that the direct relations did not
*p < .05; **p < .01. improve the model fit: Dc2(1) ¼ .25 for the comparison with
510 G. Roth et al. / Learning and Instruction 19 (2009) 506e512

Lw1 Lw2 Lw3 Lw4 Lw5

.81 .87 .85 .76 .75

Mis1 Mis2 Mis3
Love
.71
withdrawal R2 = .12 .60 .69
-.17+
-.35** Self- .32** Learning from
disclosure mistakes
.26** R2 = .11
Autonomy
support .94 .85 .67 .75 .64

Dis1 Dis2 Dis3 Dis4 Dis5

.85 .68 .59 .55

Aut1 Aut2 Aut3 Aut4

Fig. 2. Standardized path coefficients for the model predicting relations between mothers’ parenting practices, adolescents’ self-disclosure, and adolescents’
learning from their mistakes in classroom activities. þns; **p < .01. R2 ¼ Represents the value of the multiple R square of each endogenous variable.

the model that included the direct path from Love Withdrawal nonsignificant while controlling for autonomy support. Overall
to Learning from Mistakes, and Dc2(1) ¼ .43 for the the findings of both studies reveal that positive parentingdi.e.,
comparison with the model that included the direct path responsiveness in Soenens et al. (2006) study and autonomy
between Autonomy Support and Learning from Mistakes. support in the present studydis, by far, a stronger predictor of
Therefore, the hypothesized model that involved only indirect self-disclosure than psychological control or love withdrawal.
effects was preferred. The final model is presented in Fig. 2. It Following Kerr and Stattin (2000) and Soenens et al. (2006)
should be noted that in the final model the path from Love we examined disclosure and not secrecy. However, previous
Withdrawal to Self-Disclosure was not significant, p ¼ .09. research has shown that disclosure and secrecy are empirically
distinct (Finkenauer, Engles, & Meeus, 2002; Finkenauer,
4. Discussion Frijns, Engels, & Kerkhof, 2005). Kerr and Stattin (2000)
found that greater disclosure was associated with more posi-
The present study revealed three important findings. First, tive outcomes, but Finkenauer et al. (2002) found that for
and in line with Hypotheses 2 and 4, self-disclosure is adolescents, secrecy (but not disclosure) was associated with
important for learning from mistakes. Second, and as expected poorer relationships with parents, more physical complaints,
in Hypothesis 1, mothers’ love withdrawal was found to be and more depressed mood. Moreover, adolescents’ disclosure
negatively related to adolescents’ voluntary self-disclosure. and concealment each have unique associations with parenting
Interestingly, and contrary to our prediction in Hypothesis 2, (Finkenauer et al., 2005). Smetana, Metzger, Gettman, and
the significant negative correlation between love withdrawal Campione-Barr (2006) suggest that the two concepts may vary
and self-disclosure became nonsignificant (although still theoretically in the extent to which they may entail acts of
negative) while controlling for autonomy support, as was omission versus commission. Hence, it might be the case that
indicated by Fig. 2. Third, unlike love withdrawal, autonomy love withdrawal would predict more secrecy whereas
support was found to be positively and significantly related to autonomy support predicts more disclosure. Future research
learning from mistakes by enhancing self-disclosure. These may explore these hypotheses.
findings are in line with Hypotheses 3 and 4. Furthermore, the positive relation between mothers’
The present results support and extend past findings. The autonomy support and children’s ability to learn from mistakes
negative relation between love withdrawal and self-disclosure in school confirms past research that demonstrated a positive
corroborates Soenens et al.’s (2006) results. Interestingly, in relation between autonomy support and adaptation to school
Soenens et al. (2006) the relation between psychological (Grolnick & Ryan, 1989; Grolnick et al., 1991; Soenens &
control and self-disclosure remained significant while Vansteenkiste, 2005; Vansteenkiste et al., 2005). The present
controlling for parents’ responsiveness (although the relation results also extend past findings by focusing for the first time
was not very strong), whereas in the present study the relation on self-disclosure as specifically linked to the school context
between love withdrawal and self-disclosure became and by providing data on the relations of self-disclosure with
 G. Roth et al. / Learning and Instruction 19 (2009) 506e512 511

a positive outcome such as the adolescent’s ability to learn method variance. Although controlling for social desirability
from mistakes, as opposed to negative outcomes such as bias may control in part for shared method variance, and
externalizing problem behaviors. although adolescents’ experiences and perceptions of their
In the context of school experiences, the importance of parents are important antecedents of the adolescents’ behavior
adolescents’ voluntary self-disclosure to parents cannot be and well-being, additional studies using multiple reporters and
overemphasized, because the parent is not present in school behavioral observations would be very helpful in confirming
and therefore direct control is impossible. Moreover, research the present results. Second, the cross-sectional data do not
has found that direct control might backfire, especially in allow causal interpretations. It is therefore important to test the
relation to child’s internalization, behavioral engagement hypotheses with prospective longitudinal research. In addition,
(Grolnick, 2003; Roth, 2008; Roth, Assor, Kanat-Maymon, future studies should investigate the consequences of the
et al., 2007; Roth, Assor, Niemiec, et al., 2007), and parental explored parental practices while controlling for parental
knowledge of child’s whereabouts (Kerr & Stattin, 2000; school involvement because parental involvement is likely to
Stattin & Kerr, 2000). Thus, antecedents of children’s self- be related to adolescents’ school experiences. Finally, the
disclosure in relation to school experiences are of particular measure that was developed to assess the concept of learning
importance. The present study revealed that a mother’s attempt from mistakes is limited; thus, future research would do well
to take her child’s perspective and to validate the child’s to use an elaborated measure.
experience may contribute to a family climate that promotes
children’s self-disclosure. Within SDT the notion of autonomy 4.1. Conclusion
support goes beyond the two dimensions that were examined
in the present study, taking the child’s perspective and vali- In conclusion, the present findings suggest that adolescents’
dating the child’s experience, and includes providing rationale, voluntary self-disclosure of their experiences at school is an
choice, self-initiation and encouraging independent thinking important determinant of positive school outcomes. Moreover,
(Assor et al., 2002; Grolnick et al., 1997). Future research may the present results indicate that perceptions of parental behavior
shed some light on the relation between other dimensions of can predict the extent to which adolescents are willing to expose
autonomy support and children’s self-disclosure. their own school experiences. Thus, parental love withdrawal as
However, it is not known if adolescents’ ability to learn a response to children’s failure or lack of investment in school
from their mistakes in classroom activities by sharing school inhibits children’s self-disclosure and, in turn, children’s ability
experiences with a teacher and benefiting from a teacher’s help to learn from mistakes in school, whereas parental validation of
and support may be related to their experiences at home. Thus, children’s experiences and parents’ taking of their children’s
in addition to the teacher’s behavior as an antecedent of perspective may predict children’s ability to benefit from
children’s disclosure with that teacher, socialization at home mistakes by promoting children’s voluntary self-disclosure.
could influence children’s inclination to share personal expe- Parents taking children’s perspective and validating their
riences with other socialization agents. Future research is experiences may be essential for sharing difficulties with
clearly needed to explore this hypothesis. parents, and as such a good foundation for children’s disclosure
Future research should also investigate the process through of their difficulties in school. However, in order to promote
which parental psychological control and other intrusive children’s learning from mistakes it seems reasonable to
parental practices might inhibit children’s self-disclosure, assume that other parenting practices would also be involved,
whereas autonomy supportive and responsive parenting may for example, explaining the sources of mistakes and how they
promote it. The lack of statistically significant relationships can be overcome. Thus, future research should investigate
suggests that love withdrawal by itself as parenting practice is which of the constituents of autonomy support promote self-
not sufficient to explain lack of self-disclosure. Based on disclosure as well as learning from mistakes.
McGregor and Elliot’s (2005) findings, we suggest that
negative feelings, and specifically the emotion of shame, may
References
mediate the relations between psychological control, self-
disclosure, and children’s ability to cope with temporary Alfi, O., Katz, I., & Assor, A. (2004). Learning to allow temporary failure:
failures in school. Thus, in line with McGregor and Elliot’s potential benefits, supportive practices and teacher concerns. Journal of
(2005) findings, it is reasonable to assume that a parental Education for Teaching, 30, 27e41.
message that a child who fails is unworthy of love and Arbuckle, J. L., & Wothke, W. (2003). Amos 5.0 users guide. Chicago: Small
Waters.
affection (i.e., love withdrawal) might predict feelings of
Assor, A., Kaplan, H., & Roth, G. (2002). Choice is good but relevance is excellent:
shame in relation to failures, which, in turn, might inhibit the autonomy affecting teacher behaviors that predict students’ engagement in
child’s willingness to share failures with the parent. Future learning. British Journal of Educational Psychology, 72, 261e278.
research should examine the current study variables alongside Assor, A., Roth, G., & Deci, E. L. (2004). The emotional costs of perceived
self-reported feelings of shame. parents’ conditional regard: a self-determination theory analysis. Journal
of Personality, 72, 47e89.
There are several limitations to the current study. First,
Barber, B. K. (1996). Parental psychological control: revisiting a neglected
the present analyses were based on correlations among construct. Child Development, 67, 3296e3319.
cross-sectional self-reports. This is problematic in that it raises Barber, B. K., Stolz, H., & Olsen, J. (2005). Parental support, psychological
the possibility that the relations are in part a function of control and behavioral control: assessing relevance across time, culture and
512 G. Roth et al. / Learning and Instruction 19 (2009) 506e512

method. Monographs of the Society for Research in Child Development, 70, child psychology: Socialization, personality and social development (4th
1e151. ed.). (pp. 1e101) New York: Wiley.
Bentler, P. M. (1990). Comparative fit indexes in structural models. Psycho- Mattanah, J. F. (2001). Parental psychological autonomy and children’s
logical Bulletin, 107, 238e246. academic competence and behavioral adjustment in late childhood: more
Bollen, K. A. (1989). A new incremental fit index for general structural than just limit-settings and warmth. Merrill-Palmer Quarterly, 47,
equation models. Sociological Methods Research, 17, 303e316. 355e376.
Browne, M. W., & Cudeck, R. (1993). Alternative ways of assessing model fit. McGregor, H. A., & Elliot, A. J. (2005). The shame of failure: examining the
In K. A. Bollen, & J. S. Long (Eds.), Testing structural equation models link between fear of failure and shame. Personality and Social Psychology
(pp. 136e162). Thousand Oaks, CA: Sage. Bulletin, 31, 218e231.
Carmines, E. G., & McIver, J. P. (1981). Analyzing models with unobserved Parsons, J. E., Adler, T. F., & Kaczala, C. M. (1982). Socialization of
variables. In G. W. Bohrnstedt, & E. F. Borgatta (Eds.), Social measure- achievement attitudes and beliefs: parental influences. Child Development,
ment: Current issues (pp. 65e115). Beverly Hills, CA: Sage. 53, 310e321.
Cawan, A. P., Cowan, P. C., Ablow, C. J., Johnson, K. V., & Measelle, R. J. Puustinen, M., Lyyra, A. L., Metsapelto, R. L., & Pulkkinen, L. (2008).
(2005). The family context of parenting in children’s adaptation to Chidren’s help seeking: the role of parenting. Learning and Instruction, 18,
elementary school. Mahwah, NJ: Erlbaum. 160e171.
Crowne, D. P., & Marlowe, D. (1964). The approval motive. New York: Wiley. Roth, G. (2008). Perceived parental conditional regard and autonomy support
Deci, E. L., & Ryan, R. M. (2000). The ‘‘what’’ and ‘‘why’’ of goal pursuits: as predictors of young adults’ self- versus other-oriented prosocial
human needs and the self-determination of behavior. Psychological tendencies. Journal of Personality, 76, 513e534.
Inquiry, 11, 227e268. Roth, G., Assor, A., Kanat-Maymon, Y., & Kaplan, H. (2007). Perceived
Dornbusch, S. M., Ritter, P. L., Leiderman, P. H., Roberts, D. F., & autonomy in teaching: how self-determined teaching may lead to self-
Fraleigh, M. J. (1987). The relation of parenting style to adolescent school determined learning. Journal of Educational Psychology, 99, 761e774.
performance. Child Development, 58, 1244e1257. Roth, G., Assor, A., Niemiec, C. P., Ryan, R. M., & Deci, E. L. The
Finkenauer, C., Engels, R. C. M. E., & Meeus, W. (2002). Keeping secrets negative emotional and behavioral consequences of parental conditional
from parents: advantages and disadvantages of secrecy in adolescence. regard: Comparing positive conditional regard, negative conditional
Journal of Youth and Adolescence, 2, 123e136. regard, and autonomy support as parenting practices. Paper presented at
Finkenauer, C., Frijns, T., Engels, R. C. M. E., & Kerkhof, P. (2005). the 3rd Self-Determination Theory Conference, (2007, May), Toronto,
Perceiving concealment in relationships between parents and adolescents: Canada.
links with parental behavior. Personal Relationships, 12, 387e406. Roth, G., & Deci, E. (2008). Autonomy. In S. J. Lopez (Ed.), The encyclopedia
Fletcher, A. C., Steinberg, L., & Williams-Wheeler, M. (2004). Parental of positive psychology. London: Blackwell.
influences on adolescent problem behavior: revisiting Stattin and Kerr. Smetana, J. G., Metzger, A., Gettman, D. C., & Campione-Barr, N. (2006).
Child Development, 75, 781e796. Disclosure and secrecy in adolescent-parent relationships. Child Devel-
Grolnick, W. S. (2003). The psychology of parental control: How well-meant opment, 77, 201e217.
parenting backfires. Mahwah, NJ: Erlbaum. Soenens, B., & Vansteenkiste, M. (2005). Antecedents and outcomes of
Grolnick, W. S., Deci, E. L., & Ryan, R. M. (1997). Internalization within the self-determination in three life domains: the role of parents’ and
family: the self-determination theory perspective. In J. E. Grusec, & teachers’ autonomy support. Journal of Youth and Adolescence, 34,
L. Kuczynski (Eds.), Parenting and children’s internalization of values 589e604.
(pp. 135e161). New York: Wiley. Soenens, B., Vansteenkiste, M., Luyckx, K., & Goossens, L. (2006). Parenting
Grolnick, W. S., & Ryan, R. M. (1989). Parent styles associated with child- and adolescent problem behavior: an integrated model with adolescent
ren’s self-regulation and competence in school. Journal of Educational self-disclosure and perceived parental knowledge as intervening variables.
Psychology, 81, 143e154. Developmental Psychology, 42, 305e318.
Grolnick, W. S., Ryan, R. M., & Deci, E. L. (1991). The inner resources for Stattin, H., & Kerr, M. (2000). Parental monitoring: a reinterpretation. Child
school performance: motivational mediators of children’s perceptions of Development, 71, 1072e1085.
their parents. Journal of Educational Psychology, 83, 508e517. Thompson, T. (2004). Failure-avoidance: parenting, the achievement
Hoyle, R. H. (1995). Structural equations modeling: Concepts, issues and environment of the home and strategies for reduction. Learning and
applications. Thousand Oaks, CA: Sage. Instruction, 14, 3e26.
Kerr, M., & Stattin, H. (2000). What parents know. How they know it, and Vansteenkiste, M., Zhou, M., Lens, W., & Soenens, B. (2005). Experiences of
several forms of adolescent adjustment: further support for a reinterpreta- autonomy and control among Chinese learners: vitalizing or immobiliz-
tion of monitoring. Developmetal Psychology, 36, 366e380. ing? Journal of Educational Psychology, 97, 468e483.
Maccoby, E. E., & Martin, J. A. (1983). Socialization in the context of the Wells, E., & Rankin, J. (1988). Direct parental controls and delinquency.
family: parent-child interaction. In E. M. Hetherington (Ed.), Handbook of Criminology, 26, 263e285.