The interplay of parental support, parental pressure and test anxiety – Gender differences in adolescents

Journal of Adolescence 45 (2015) 67e79

Contents lists available at ScienceDirect

Journal of Adolescence journal homepage: www.elsevier.com/locate/jado

The interplay of parental support, parental pressure and test anxiety e Gender differences in adolescents Tobias Ringeisen a, 1, 2, Diana Raufelder b, *, 1 a

Merseburg University of Applied Sciences, Geusaer Str. 88, 06217 Merseburg, Germany Free University Berlin, Department of Educational Science and Psychology, AB Methods & Evaluation, Habelschwerdter Allee 45, 14195 Berlin, Germany b

a r t i c l e i n f o

a b s t r a c t

Article history: Available online xxx

This study examined gender-specific relationships between adolescents' perceptions of school-related support/pressure from their parents and test anxiety. A sample of German students (N ¼ 845; Mage ¼ 15.32; SD ¼ .49) completed questionnaires that measured their perceived parental support/pressure (for mother and father separately) as well as the four main components of test anxiety (worry, interference, lack of confidence, and emotionality). Gender-specific relations were identified using multigroup structural equation modeling: For girls, perceived maternal pressure was positively associated with emotionality and interference; for boys, perceived father pressure and father support were positively associated with interference and worry, respectively. For both genders, perceived mother pressure and support were related to lack of confidence. Our findings suggest that adolescents' perceptions of maternal attitudes are associated with students' self-confidence irrespective of the child's gender, whereas the remaining facets of test anxiety follow same-sex trajectories between perceived parental attitudes and adolescents' test anxiety. © 2015 The Foundation for Professionals in Services for Adolescents. Published by Elsevier Ltd. All rights reserved.

Keywords: Test anxiety Parental pressure Parental support Gender Adolescence

Introduction Many students suffer from heightened test anxiety (TA), which is considered to be a situation-specific form of anxiety, namely the disposition to perceive situations of assessment as threatening and thus respond with heightened state anxiety (Spielberger & Vagg, 1995; Zeidner, 2007). In the context of schooling, TA is associated with debilitating effects such as lower (intrinsic) achievement motivation, reduced engagement in academic tasks, as well as impoverished academic performance (e.g., Cassady & Johnson, 2002; Zeidner, 2007). TA is widely conceptualized as a multidimensional construct featuring three distinct cognitive facets: worry (disruptive concerns about individual performance and the consequences of failure), interference (distraction from academic tasks by intrusive, unrelated thoughts) and lack of confidence (low self-confidence and a

* Corresponding author. Tel.: þ49 30 83855748. E-mail addresses: [email protected] (T. Ringeisen), [email protected] (D. Raufelder). URL: http://www.self-project.de 1 Shared first authorship. 2 Tel.: þ49 3461 46 2428; fax: þ49 3461 46 2422. http://dx.doi.org/10.1016/j.adolescence.2015.08.018 0140-1971/© 2015 The Foundation for Professionals in Services for Adolescents. Published by Elsevier Ltd. All rights reserved.

68

T. Ringeisen, D. Raufelder / Journal of Adolescence 45 (2015) 67e79

lack of self-esteem). The facet of emotionality describes the autonomic expression of physical reactions (e.g., stomach ache, high pulse and perspiration) accompanied by intense affect (Hodapp, Rohrmann, & Ringeisen, 2011; Zeidner, 2007). Relations between parental behavior and test anxiety in adolescents Research has sought to understand how the intensity of emotions is linked to students' performance and achievement such as TA and/or the relationships between them are shaped by features of their social contexts. Existing research on this topic is condensed in socio-contextual models of achievement and its related emotions, which portray how the behavior of specific social groups might influence students' self-perceived competence, academic performance and/or related emotions (e.g., Eccles, 2007; Lowe et al., 2008; Simpkins, Fredricks, & Eccles, 2012). Besides teachers and peers, parents constitute an important social group for adolescents as their behavior may be linked to their children's scholastic adjustment, performance and achievement-related emotions, especially anxiety, during adolescence and early adulthood (cf. Bois, Lalanne, & Delforge, 2009; Putwain, Woods, & Symes, 2010; Shadach & Ganor-Miller, 2013). Among the diverse types of parenting behavior, perceived parental support and pressure have received considerable attention in contemporary research on self-perceived competence, performance and/or related emotions in students (e.g., Laible & Carlo, 2004; Ommundsen, Roberts, Lemyre, & Miller, 2006; Putwain et al., 2010; Raufelder, Hoferichter, Ringeisen, Regner, & Jacke, 2015). Parental support primarily indicates socio-academic support, which includes helping the child to prepare for a test (behavioral level) or encouraging the child to trust in his/her ability to accomplish various academic tasks (emotional level). In contrast, socio-academic pressure describes behavior such as pushing the child to work more intensely for school (behavioral level), setting high academic expectations, and/or criticizing academic performance, which may exceed the capability of the child (emotional level) (Ommundsen et al., 2006; Putwain et al., 2010; Raufelder et al., 2015; Reitzle, Metzke, & Steinhausen, 2001). Adolescence is characterized by school transitions, which require students to manage growing academic demands in increasingly formal classroom environments typified by constant evaluation and competition, which is often associated with increased TA (for an overview see Hanewald, 2013; Kennelly & Monrad, 2007). The transition to high school usually entails the loss of previous valuable relationships with peers and teachers. As a result, many students suffer from increased anxiety and socio-emotional maladjustment, which highlights their need to maintain close and supportive relationships with significant others during this tumultuous period (DeWit, Karioja, Rye, & Shain, 2011; Frey, Ruchkin, Martin, & Schwab-Stone, 2009; Martinez, Aricak, Graves, Peters-Myszak, & Nellis, 2011). Accordingly, low attachment to parents is likely associated with adolescents' impaired emotional adjustment and impoverished scholastic performance (Duchesne, Ratelle, Poitras, & Drouin, 2009; Kenny, Lomax, Brabeck, & Fife, 1998; Papini & Roggman, 1992). In contrast, parental support is crucial if students are to meet high academic demands, develop positive attitudes towards their capabilities, and regulate their achievement-related anxieties (e.g., Frey et al., 2009; Martinez et al., 2011). Research has found that adolescents with high parental support e characterized by high attachment, supportive communication, and/or encouragement of personal growth e reported better performance and lower general anxiety as well as lower test anxiety (Chapell & Overton, 1998; Papini & Roggman, 1992; Peleg-Popko & Klingman, 2002; Wolfradt, Hempel, & Miles, 2003) than adolescents with less parental support. In contrast, high parental pressure correlated positively with general anxiety and test anxiety (Besharat, 2003; Wolfradt et al., 2003), especially with the cognitive facets of worry, interference, and/or lack of confidence (Gherasim & Butnaru, 2012; Putwain et al., 2010). One study also identified a positive association between pressure and emotionality (Putwain et al., 2010). Yet no study has differentiated the stated effects by parents' gender (mother/father), and/or the gender of the adolescent child (Brand & Klimes-Dougan, 2010). Gender-specific associations between parental behavior and test anxiety Research on gender socialization and differentiation suggests that mothers and fathers set distinct examples for their children, where the same-sex parent usually constitutes the primary gender role model (for an overview see e.g., Martin, 2000; Payne, 2001; Ruble & Martin, 1998): Although the mother as primary caretaker has a significant influence on the later development of girls and boys alike, the interaction patterns or the relationship quality between each parent and their daughter/son may be associated differently with school-related attitudes, behavior and socio-emotional adjustment during childhood and adolescence (Duchesne et al., 2009; Maccoby, 2002; Martinez et al., 2011; Raufelder et al., 2015). To address these gender-specific patterns, Eccles and colleagues refined the assumptions that underpin socio-contextual models of achievement and related emotions, suggesting that parents' own gender, their gender stereotypes, and/or their child's gender can influence the interplay of parental behavior and the child's academic performance, and/or achievement emotions (Eccles, 2007; Jacobs & Eccles, 1992; Simpkins et al., 2012). However, few studies have investigated gender-specific relationships between parental behavior and adolescents' test anxiety. Studies on parental support have identified the mother as the primary source of support: her behavior is positively associated with academic self-confidence, improved performance/ competence, and better emotional adjustment such as reduced worry, and/or anxiety in both adolescent girls and boys (e.g., Chen, Liu, & Li, 2000; Duchesne et al., 2009; Gomez & McLaren, 2006; Holahan, Valentiner, & Moos, 1995; Laible & Carlo, 2004; Peleg-Popko, 2004). In contrast, the empirical evidence for father support is equivocal: some studies (e.g., Biller & Lopez-Kimptom, 1997; Gomez & McLaren, 2006; Holahan et al., 1995; Wang & Liu, 2000) have found emotional support from fathers such as caring and empathetic understanding, to predict better psychological adjustment, such as low distress, low social anxiety, and greater academic confidence, whereas others cannot attest to such associations (Flouri, 2008; Laible &

T. Ringeisen, D. Raufelder / Journal of Adolescence 45 (2015) 67e79

69

Carlo, 2004). Only two studies to our knowledge address the differential gender effects of father support for boys, identifying a positive relation between father attachment and well-being (Kenny et al., 1998), and a negative correlation between support and emotionality (Krohne, 1990). Examining the effects of maternal/paternal pressure, research suggests gender-specific pathways: For adolescent girls, pressure from the mother positively predicted both cognitive and bodily-affective manifestations of TA (Besharat, 2003; Krohne, 1990), whereas maternal control was positively associated with higher levels of anxiety and depression (Pettit, Laird, Bates, Dodge, & Criss, 2001). For adolescent boys, studies found symptoms of worry and interference and, in some cases, emotionality to correlate positively with father pressure, whereas mother pressure was positively associated with TA in general (Besharat, 2003; Krohne, 1990). Current study The research on gender-specific relations between perceived parental pressure/support and TA in adolescents has mostly focused on selected facets of test anxiety (e.g. worry or emotionality) or operationalized TA by means of a total score without considering perceived school-related parental support and/or pressure and/or differentiating between the perception of € gels & Phares, 2008; Brand & Klimes-Dougan, 2010; Zeidner, 2007). In response to this gap, mothers' and fathers' behavior (Bo the goal of the current study was to highlight gender-specific pathways between perceived mother and father support/ pressure and four components of TA (worry, interference, lack of confidence, and emotionality) in adolescent girls and boys simultaneously. The following hypotheses were founded on gender models, which state that the same-sex parent usually constitutes the primary gender role model (e.g., Martin, 2000; Payne, 2001; Ruble & Martin, 1998), and research on the gender-specific pathways between maternal/paternal behavior and the socio-emotional adjustment of their daughters/sons that highlighted the essential role of perceived parental support (e.g., Duchesne et al., 2009; Maccoby, 2002; Martinez et al., 2011; Raufelder et al., 2015). Hypothesis 1: Associations between parental support, parental pressure, and test anxiety in girls. We expected the perceived support and pressure from the mother to play a primary role in girls' TA. In detail, based on previous empirical evidence for a positive correlation between mother pressure/control and various manifestations of test and general anxiety (Besharat, 2003; Krohne, 1990; Pettit et al., 2001), we hypothesized that mother pressure would be positively associated with all four facets of TA in the current sample. Furthermore, as previous research has indicated a positive correlation between maternal support and emotional adjustment such as lowered test and general anxiety and greater academic confidence (e.g., Chen et al., 2000; Duchesne et al., 2009; Gomez & McLaren, 2006; Holahan et al., 1995; Laible & Carlo, 2004; Peleg-Popko, 2004), we hypothesized that mother support would be negatively related to all four facets of TA in girls, particularly to lack of confidence. Hypothesis 2: Associations between parental support, parental pressure, and test anxiety in boys. We expected that perceived support and pressure from the father would play a primary role for boy's TA, whereas the mother, being the primary caretaker, would also contribute to boys' TA. Based on previous studies which have shown that father pressure is positively associated with worry and interference, whereas mother pressure is positively associated with TA in general (Besharat, 2003; Krohne, 1990), we expected to find the same pattern in the current sample. Based on Krohne's (1990) research, father support was expected to negatively correlate with emotionality. Furthermore, as previous studies have shown that mother support is positively associated with reduced anxiety and/or greater academic confidence (e.g., Chen et al., 2000; Duchesne et al., 2009; Gomez & McLaren, 2006; Holahan et al., 1995; Laible & Carlo, 2004; Peleg-Popko, 2004), we hypothesized that mother support would be negatively related to all four facets of TA in boys, particularly to lack of confidence. Since the empirical evidence for further associations is sparse and/or mixed, the remaining correlation patterns among girls and boys were examined for exploratory purposes. Method Participants The quantitative data was collected in 22 secondary schools, which were randomly selected from a pool of 124 public secondary schools in the federal state of Brandenburg, Germany. Respondents (465 girls and 380 boys) were 9th grade students between the age of 13 and 17 (N ¼ 845; Mage ¼ 15.32; SD ¼ .49). Given the very small proportion of ethnic minority € residents in Brandenburg (2.6%), our sample was not refined by the ethnic background of informants (Statistische Amter des €nder, 2011 [statistical office of the federation and provinces]). Since German laws prohibit collecting socioBundes und der La economic information about an individual from a third party, this data was also omitted. Procedure All procedures were in accordance with the ethical standards of the Helsinki Declaration of 1975, revised in 2000 and 2008. In preparation for data collection, we obtained permission to conduct the study from the Department of Education, Youth and Sports of Brandenburg, as well as from participating parents and students. Schools, parents and students were

70

T. Ringeisen, D. Raufelder / Journal of Adolescence 45 (2015) 67e79

thoroughly informed about the voluntary nature of their participation, which meant that responses to questions were not obligatory and would be treated in confidentiality. The response rate was 85%. Data collection occurred in the spring term of 2013 via (anonymous) written, class-based questionnaires. At least two research assistants were present in each session to clarify any uncertainties relating to items or questions. To minimize the potential flaws of self-reported data, operationalization and data handling were employed as recommended by Chan (2009).3

Measures All measures used in this study are well-established, validated instruments for German adolescent students. The reported Cronbach's Alpha values are based on the current sample. Test anxiety The “German Test Anxiety Questionnaire” [Prüfungsangstfragebogen] (PAF) (Hodapp et al., 2011) was used to measure worry (girls a ¼ .75; boys a ¼ .71), emotionality (girls a ¼ .74; boys a ¼ .77), interference (girls a ¼ .78; boys a ¼ .76), and lack of confidence (girls a ¼ .81; boys a ¼ .82) during situations of school-related evaluation with five items each. Participants read statements and used a 4-point Likert scale (1 ¼ “strongly disagree”, 4 ¼ “strongly agree”) to assess the extent to which each statement described their feelings and thoughts during school evaluations. Item examples include: “I think about what will happen if I don't do well” (worry), “My heart pounds” (emotionality), “My concentration is interrupted by interfering thoughts” (interference), and “I think that I will succeed” (lack of confidence). The five items relating to lack of confidence were phrased positively so as to capture confidence and thus needed to be re-coded for the subsequent analyses. Parental support Following the “Zürich Questionnaire of Educational Behavior” [Züricher Kurzfragebogen zum Erziehungsverhalten] (ZKE) developed by Reitzle et al. (2001), both school-related father support (girls a ¼ .92; boys a ¼ .91) and mother support (girls a ¼ .91; boys a ¼ .88) were measured using a 10-item scale. On a 4-point Likert scale, the items ranged from 1 (totally true) to 4 (not true at all), such as “If I do not understand something, my mother/father explains it to me”, or “My mother/father is studying with me”. Support was assessed separately for the mother and father using the same set of items twice. Parental pressure Again, following the ZKE developed by Reitzle et al. (2001), both school-related father pressure (girls a ¼ .83; boys a ¼ .81) and mother pressure (girls a ¼ .80; boys a ¼ .79) were measured using a 6-item scale. The items, which ranged on a 4-point Likert scale from 1 (totally true) to 4 (not true at all), included: “My mother/father pushes me to work harder in school”, and “My mother/father expects that I do well in school”. The same set of items was used to assess mother and father pressure separately. In total, 6.9% of the sample reported living in a single-parent household with the mother (34 girls and 24 boys), and .5% of the sample reported living in a single-parent household with the father (1 girl and 3 boys).

Statistical analyses Measurement model In order to produce a measurement model, initial confirmatory factor analyses (CFA) were run in Mplus version 7.0 n & Muthe n, 1998e2012). Due to the large number of items per measure, which negatively affects the model fit (Ding, (Muthe Velicer, & Harlow, 1995; Wang & Wang, 2012), and because there are 3 advantages to using parcels instead of original items e (1) estimating large numbers of items may result in spurious correlations, (2) subsets of items from a greater item pool are likely to share sources of variance that are potentially irrelevant, and (3) solutions from parcels of items are more stable than solutions from single items (Little, Cunningham, Shahar, & Widaman, 2002) e the items were randomly assigned to two parcels for the four facets of TA (per subscale, one parcel with 2 items and 3 items, respectively), parental support (separately for mother and father, two parcels with 5 items each) and parental pressure (separately for mother and father, two parcels with 3 items each). Correlations Using Mplus we conducted bivariate correlations between all latent variables for boys and girls separately: the four components of TA (worry, emotionality, interference, lack of confidence), parental support and parental pressure.

3 In detail (1) only well-established scales were used, for which construct validity was supported, (2) in the case of high correlations between the variables, potential common method variance was tested, (3) the risk of social desirability was minimized through specific subtle item formulations and specific sets of items (e.g., more nonsensitive “neutral” distractors were placed between some sensitive items (for details see Glynn, 2013)), (4) additional raters were not considered, since we were particularly interested in students' perceptions.

T. Ringeisen, D. Raufelder / Journal of Adolescence 45 (2015) 67e79

71

Latent mean comparison and multigroup structural equation modeling We conducted multigroup structural equation models (MGSEM) with the MLR estimator. Following the type-is-complex approach proposed by Asparouhov (2005) for complex survey data, standard error biases owing to the nested nature of our data (845 students in 67 school classes) were corrected. Missing data values were considered using full information maximum likelihood (FIML). Model fit was estimated using five primary fit indices (cf. Hu & Bentler, 1999): Chi-Square Test of Model Fit (c2), Root Mean Square Error of Approximation (RMSEA) including the 90% CIs of the RMSEA (90% CI), Comparative Fit Index (CFI), TuckereLewis Index (TLI), and Standardized Root Mean Square Residual (SRMR). To conduct a stepwise examination of measurement invariance between boys and girls, four confirmatory factor analyses (CFA) were run. Following Brown's (2006) procedure, we tested (1) the CFA model separately for boys and girls (model 1: configurational measurement invariance); (2) for equal form (model 2: identical factor structure e all parameters free; factorial measurement invariance); (3) the equality of factor loadings (model 3: metric measurement invariance); (4) the equality of intercepts (model 4: scalar invariance) (Brown, 2006; Byrne, 2008). Using the c2-difference test, the models 2 to 4 were subsequently compared (Yuan & Bentler, 2004). Partial scalar invariance (model 4b) is the basic condition for conducting latent mean comparison (LMC) as well as MGSEM (Brown, 2006). Subsequently, two MGSEMs were conducted by applying (1) a semi-restricted multigroup model, and (2) a fully-restricted multigroup model. The semi-restricted model was applied assuming free intercepts and free regression coefficients among boys and girls, whereas the fully-restricted model was applied assuming free intercepts but equal regression coefficients across both genders. By using the c2-difference test, the two models were compared to examine whether there are essential gender differences. Results Descriptive statistics Table 1 illustrates the gender-specific mean values, standard deviations, ranges, skewness and kurtosis for the variables included in this study. According to West, Finch, and Curran (1995) skewness values below 2 and kurtosis values below 7 signify a normal distribution, implying that the variables in the current sample show a normal distribution. Bivariate correlations The gender-specific bivariate associations among the latent variables are presented in Table 2. Comparing the correlation patterns between the variables in this study revealed slight variations for girls and boys: There were no significant associations between perceived mother support and mother pressure (r ¼ .00, p > .05) as well as perceived father pressure for girls (r ¼ .02, p > .05), whereas there was a positive significant association between these variables for boys (MS/MP: r ¼ .09, p < .01; MS/FP: r ¼ .06, p < .05). Interestingly, positive associations between perceived mother support and worry (girls: r ¼ .23, p < .01; boys: r ¼ .26, p < .01) as well as between perceived father support and worry (girls: r ¼ .21, p < .01; boys: r ¼ .32, p < .001) was found for both boys and girls. Furthermore, for boys (but not for girls) perceived father pressure was positively related to worry (r ¼ .03, p < .05), emotionality (r ¼ .06, p < .001), and interference (r ¼ .07, p < .001). In addition, Table 1 Summary of means, range, standard deviations, skewness (SE) and kurtosis (SE). Measure

M

Range

SD

Skewness (SE)

Kurtosis (SE)

Girls 1. MS 2. MP 3. FS 4. FP 5. Worry 6. Emo 7. Inter 8. LoC

2.95 2.66 2.88 2.42 3.00 2.41 2.38 2.84

1e4 1e4 1e4 1e4 1e4 1e4 1e4 1e4

.65 .60 .73 .61 .70 .77 .66 .60

.55 .10 .64 .11 .44 .21 .34 .42

(.10) (.10) (.10) (.10) (.12) (.12) (.12) (.12)

.20 .40 .14 .28 .36 .62 .47 .22

(.20) (.20) (.21) (.21) (.25) (.25) (.25) (.25)

Boys 1. MS 2. MP 3. FS 4. FP 5. Worry 6. Emo 7. Inter 8. LoC

2.94 2.79 2.91 2.61 2.72 1.99 2.20 2.39

1e4 1e4 1e4 1e4 1e4 1e4 1e4 1e4

.61 .61 .69 .63 .62 .73 .63 .62

.61 .15 .67 .03 .20 .59 .13 .20

(.11) (.11) (.11) (.11) (.13) (.13) (.13) (.13)

.21 .33 .06 .20 .20 .10 .59 .29

(.22) (.22) (.23) (.23) (.27) (.27) (.27) (.27)

Note. Worry ¼ Worry, Emo ¼ Emotionality; Inter ¼ Interference, LoC ¼ Lack of confidence, MS ¼ Mother Support, FS ¼ Father Support, MP ¼ Mother Pressure, FP ¼ Father pressure.

72

T. Ringeisen, D. Raufelder / Journal of Adolescence 45 (2015) 67e79

Table 2 Gender-specific bivariate correlations between the latent variables of test anxiety, parental support and parental pressure. Measure Girls 1. MS 2. FS 3. MP 4. FP 5. Worry 6. Emo 7. Inter 8. LoC Boys 1. MS 2. FS 3. MP 4. FP 5. Worry 6. Emo 7. Inter 8. LoC

2

3

4

5

6

7

8

.55*** e

.00 .02 e

.07 .21** .54*** e

.23** .21** .03 .02 e

.06 .05 .20** .11 .61*** e

.03 .01 .19** .11 .26*** .77*** e

.24*** .17** .20** .05 .39*** .25** .31*** e

.62*** e

.23** .16* e

.23** .41*** .70*** e

.26** .32*** .17* .17* e

.05 .07 .24** .29*** .70*** e

.07 .05 .22** .35*** .58*** .76*** e

.25** .25** .23** .12 .66*** .18* .14 e

Note. Correlations are standardized coefficients; *p < .05, **p < .01, ***p < .001; Worry ¼ Worry, Emo ¼ Emotionality; Inter ¼ Interference, LoC ¼ Lack of confidence, MS ¼ Mother Support, FS ¼ Father Support, MP ¼ Mother Pressure, FP ¼ Father pressure.

maternal pressure and worry were significantly positively correlated for boys but not for girls (r ¼ .04, p < .05). In contrast, a positive association between interference and lack of confidence (r ¼ .07, p < .001) was found only for girls. For both girls and boys, positive associations between mother support and father support (girls: r ¼ .55, p < .001; boys: r ¼ .62, p < .001), father support and father pressure (girls: r ¼ .21, p < .01; boys: r ¼ .41, p < .001), as well as between father pressure and mother pressure (girls: r ¼ .54, p < .001; boys: r ¼ .70, p < .001) were identified. In addition, both mother support (girls: r ¼ .24, p < .001; boys: r ¼ .25, p < .01) and father support (girls: r ¼ .17, p < .01; boys: r ¼ .25, p < .01) were negatively related to lack of confidence for both genders. Furthermore, a positive relationship between mother pressure and emotionality (girls: r ¼ .20, p < .01; boys: r ¼ .24, p < .01), interference (girls: r ¼ .19, p < .01; boys: r ¼ .22, p < .01) and lack of confidence (girls: r ¼ .20, p < .01; boys: r ¼ .23, p < .01) was found for boys and girls. Finally, positive associations were found between worry and emotionality (girls: r ¼ .61, p < .001; boys: r ¼ .70, p < .001), worry and interference (girls: r ¼ .26, p < .001; boys: r ¼ .58, p < .001), emotionality and interference (girls: r ¼ .77, p < .001; boys: r ¼ .76, p < .001), emotionality and lack of confidence (girls: r ¼ .25, p < .01; boys: r ¼ .18, p < .05), as well as negative associations between worry and lack of confidence (girls: r ¼ .39, p < .001; boys: r ¼ .66, p < .001) (see Table 2).

Latent mean comparison and multigroup structural equation modeling Prior to conducting MGSEM, CFA were performed in a stepwise manner to examine measurement invariance between girls and boys (see Table 3). The final model (model 4b, see Table 3) demonstrates partial scalar measurement invariance, which allows for LMC as well as for MGSEM. With girls as the reference group, LMC showed that boys reported significantly more mother pressure (b ¼ .31, p < .001) and father pressure (b ¼ .25, p < .01). Girls and boys did not differ in their perception of father support (b ¼ .03, p ¼ .659) and mother support (b ¼ .14, p ¼ .075). In terms of TA, boys reported less emotionality (b ¼ .44, p < .001), less interference (b ¼ .40, p < .001), and less lack of confidence (b ¼ .80, p < .001) than girls. Girls and boys did not exhibit significant differences in worry (b ¼ .17, p ¼ .212). Table 3 Model fit indices of the stepwise CFA procedure to proof measurement invariance. Model Model Model Model Model Model Model

1a: girls 1b: boys 2 3* 4** 4b***

df

c2

p

CFI

TLI

RMSEA

90% CI

SRMR

Dc2

Ddf

p

74 74 148 164 172 170

189.53 160.94 351.97 370.19 430.58 372.06

<.001 <.001 <.001 <.001 <.001 <.001

.96 .97 .96 .96 .95 .96

.94 .94 .94 .45 .93 .95

.06 .06 .06 .06 .06 .05

(.05e.07) (.04e.07) (.05e.07) (.05e.06) (.05e.07) (.05e.06)

.06 .05 .06 .07 .07 .07

e e e 17.91 65.82 2.85

e e e 16 8 6

e e e .33 <.001 .64

Note. Model 1 ¼ for boys and girls separately (form invariance); Model 2 ¼ all parameters free (form invariance); Model 3 ¼ equality of factor loadings (metric invariance); Model 4 ¼ equality of intercepts (scalar invariance); Model 4b ¼ partial scalar invariance; *result of the c2-difference test between Model 2 and Model 3 is presented at the end of this line; **result of the c2-difference test between Model 3 and Model 4 is presented at the end of this line; *** result of the c2-difference test between Model 3 and Model 4b is presented at the end of this line.

T. Ringeisen, D. Raufelder / Journal of Adolescence 45 (2015) 67e79

73

To test the hypotheses, two MGSEMs were constructed consisting of direct effects from each component of perceived parental support and parental pressure on the four facets of TA (see Figs. 1 and 2). Firstly, a MGSEM was conducted, in which the regression coefficients and the intercepts were free across girls and boys (semi-restricted model). Indices revealed a good fit for this semi-restricted model (c2(163) ¼ 381.00, p < .001; CFI ¼ .96, TLI ¼ .94; RMSEA ¼ .06 (.05e.06); SRMR ¼ .06). Secondly, a MGSEM was conducted in which the regression coefficients were equal and the intercepts were free across girls and boys (fully-restricted model). The indices for this fully-restricted model (c2(180) ¼ 439.82, p < .001; CFI ¼ .95, TLI ¼ .94; RMSEA ¼ .06 (.05e.07); SRMR ¼ .06) also revealed a good fit, albeit not as good as the semi-restricted model fit (see Table 4). The c2-difference test comparing the semi-restricted with the fully-restricted model did reach significance (c2(17) ¼ 60.21, p < .001), suggesting that the semi-restricted multigroup model does not fit the data significantly worse than the fullyrestricted model. This means that the interrelations among the study variables exhibited different patterns for girls and boys (Yuan & Bentler, 2004). Pattern for girls For girls, four of the hypothesized direct effects were significant. High perceived mother support was negatively related to lack of confidence (b ¼ .22, SE ¼ .08, p < .01) but to none of the other three TA facets. Perceived mother pressure was positively associated with emotionality (b ¼ .21, SE ¼ .07, p < .05), interference (b ¼ .19, SE ¼ .05, p < .05), and lack of confidence (b ¼ .25, SE ¼ .06, p < .01). Interestingly, none of the father variables were significantly related to any aspect of TA for girls. Additionally, most of the covariances among parental behavior (between father and mother support, r ¼ .23, p < .001; between father and mother pressure, r ¼ .19, p < .001; between father pressure and father support, r ¼ .07, p < .01), as well as among the four facets of TA were found to be significant: worry and emotionality (r ¼ .14, p < .001), interference and worry (r ¼ .05, p < .001), interference and emotionality (r ¼ .15, p < .001), lack of confidence and worry (r ¼ .09, p < .001), lack of confidence and emotionality (r ¼ .05, p < .01), and lack of confidence and interference (r ¼ .06, p < .001). Pattern for boys Three of the hypothesized direct effects in the model were significant. Perceived mother support was negatively related to lack of confidence (b ¼ .21, SE ¼ .07, p < .01), but to none of the other three TA facets. Perceived father pressure was positively associated with interference (b ¼ .25, SE ¼ .14, p < .01). Unexpectedly, worry was positively related to father support (b ¼ .29, SE ¼ .07, p < .01). Furthermore, lack of confidence was positively associated with boys' perception of mother pressure (b ¼ .27, SE ¼ .09, p < .01) but not with father support. For the parental variables, all covariances were found to be significant: father and mother support (r ¼ .21, p < .001), father and mother pressure (r ¼ .22, p < .001), mother support and pressure (r ¼ .09, p < .01), father support and pressure (r ¼ .12, p < .01), mother pressure and father support (r ¼ .06, p < .05), as well as father pressure and mother support (r ¼ .06, p < .01). Only four covariances between the four TA variables turned out to be

Fig. 1. Multigroup structural equation model for girls. Significant effects shown as first unstandardized coefficients (B), second standardized coefficients (b); bold pathways are significant at *p < .05, **p < .01, ***p < .001; dotted pathways are not significant; factor loadings are standardized. For the sake of clarity, the covariances between the predictor variables (MS, FS, MP, FP) and the covariances between the outcome variables (worry, emo, inter, loc) were not included in the figure, but explained within text (see Results). Worry ¼ Worry, Emo ¼ Emotionality; Inter ¼ Interference, LoC ¼ Lack of confidence, MS ¼ Mother Support, FS ¼ Father Support, MP ¼ Mother Pressure, FP ¼ Father pressure.

74

T. Ringeisen, D. Raufelder / Journal of Adolescence 45 (2015) 67e79

Fig. 2. Multigroup structural equation model for boys. Significant effects shown as first unstandardized coefficients (B), second standardized coefficients (b); bold pathways are significant at *p < .05, **p < .01, ***p < .001; dotted pathways are not significant; factor loadings are standardized. For the sake of clarity, the covariances between the predictor variables (MS, FS, MP, FP) and the covariances between the outcome variables (worry, emo, inter, loc) were not included in the figure, but explained within text (see Results). Worry ¼ Worry, Emo ¼ Emotionality; Inter ¼ Interference, LoC ¼ Lack of confidence, MS ¼ Mother Support, FS ¼ Father Support, MP ¼ Mother Pressure, FP ¼ Father pressure.

Table 4 Model fit indices for the semi-restricted model and fully-restricted model. Indices

Semi-restricted model

Fully-restricted model

c2-Test of model fit

381.00 163 <.001 .96/.94 .06 (.05e.06) .06

439.82 180 <. 001 .95/.94 .06 (.05e.07) .06

(df) p(c2) CFI/TLI RMSEA (90% CI) SRMR

Note. df ¼ degrees of freedom, CFI ¼ Comparative Fit Index; TLI ¼ TuckereLewis Index; RMSEA ¼ Root Mean Square Error of Approximation (90% confidence interval); SRMR ¼ Standardized Root Mean Square Residual.

significant: worry and emotionality (r ¼ .12, p < .001), interference and worry (r ¼ .09, p < .001), interference and emotionality (r ¼ .17, p < .001), and lack of confidence and worry (r ¼ .13, p < .001). To summarize, girls' TA facets were only associated with their perceptions of mother's behavior, whereas boys showed a mixed response to both perceived father and mother behavior. The explained variance for all four facets of TA was as follows for boys (worry: R2 ¼ .115; emotionality: R2 ¼ .088; interference: R2 ¼ .137; lack of confidence: R2 ¼ .166) and for girls (worry: R2 ¼ .060; emotionality: R2 ¼ .051; interference: R2 ¼ .037; lack of confidence: R2 ¼ .103). Discussion The current study examined the interplay of perceived parental support, parental pressure, and test anxiety. Following the socio-contextual models of achievement and related emotions (Eccles, 2007; Jacobs & Eccles, 1992; Lowe et al., 2008; Simpkins et al., 2012), our main objective was to reveal gender-specific relations by investigating the associations between perceived parental support and pressure (separately for mothers and fathers), and the four facets of TA (worry, emotionality, interference, and lack of confidence) for adolescent girls and boys simultaneously. Gender-specific correlation analyses The initial gender-specific correlations indicated slight gendered associations: For girls, perceived mother support was neither significantly related to perceived mother pressure, nor to perceived father pressure. In contrast, these associations were found to be significant for boys. For both girls and boys, perceived father support was positively associated with

T. Ringeisen, D. Raufelder / Journal of Adolescence 45 (2015) 67e79

75

perceived father pressure, whereas the perception of mother support and pressure were positively related for boys only. Overall, these results are in line with gender socialization and differentiation research (see Martin, 2000; Payne, 2001; Ruble & Martin, 1998). For instance, Hazzard, Christensen, and Margolin (1983) found that girls are more skilled in distinguishing different behavioral cues in their mother, who acts as their primary gender role model, and between both parents, yet they are less skilled in differentiating their father's behaviors. Meanwhile boys are less capable of differentiating behavioral cues from both parents and thus tend to blend their perceptions (Hazzard et al., 1983). In addition, previous research shows that boys have a tendency to perceive their parents as supportive and demanding at once, whereas girls interpret parental support as the opposite of parental pressure (Leff & Hoyle, 1995; Raufelder et al., 2015). As an alternative explanation, paternal pressure and (informational rather than emotional) support may constitute two sides of the same coin since both reflect a paternal interest in improving their daughter/son's study outcomes (Biller & Lopez-Kimptom, 1997; Brand & Klimes-Dougan, 2010; Leung, Yeung, & Wong, 2010). Turning to the associations between the four aspects of test anxiety, a positive association was found between interference and lack of confidence for girls only. Interestingly, lack of confidence was negatively associated with worry for both boys and girls. A possible explanation is that adolescent students might worry about exams due to the perceived importance of academic success for their futures (Arnold, 1995), which might go along with their thirst for new information, more structured learning approaches, and improved academic performance, eventually fostering academic confidence (Boekaerts, 1995). Perkins and Corr (2005) suggest that the tendency to worry may have intrinsic motivational power as it induces positive learning behavior and eventually increases academic confidence. Their theory is indirectly supported by research which found lack of confidence to be negatively associated with intrinsic motivation and self-oriented perfectionism, and worry to be positively associated with intrinsic motivation (Stoeber, Feast, & Hayward, 2009). This approach might also explain the identified positive association between worry and perceived father support as well as perceived mother support for both male and female adolescents. In their desire to meet the expectations of their perceivedly supportive parents, students might feel self-induced pressure to perform well and thus worry about disappointing their parents, which in turn is associated with increased anxiety during test situations. Accordingly, this finding can also be explained by research, which has shown that self-induced expectations as well as expectations from external sources lead to increased levels of test anxiety (Hill & Wigfield, 1984; Selkirk, Bouchey, & Eccles, 2011; Zohar, 1998). Latent mean comparison In accordance with previous research (cf. Raufelder et al., 2015; Zeidner, 2007), the LMC revealed that girls reported greater levels of TA than boys for all its facets except worry. Previous research offers interpretations for these findings: Gender role socialization may condition girls to report anxiety more openly than boys, who are taught to contain their emotions (Davis, Burleson, & Kruszewski, 2011; Eccles, 2007). Moreover, girls may exhibit higher levels of TA because they have greater self-awareness of bodily reactions and negative affect (Else-Quest, Higgins, Allison, & Morton, 2012), express more self-doubt (cf. Cassady & Johnson, 2002; Ringeisen, Raufelder, Schnell, & Rohrmann, 2015; Selkirk et al., 2011), and are more likely to perceive and/or feel threatened by gender stereotypes (e.g., Osbourne, 2006). In contrast, the current study found that boys perceive significantly more mother and father pressure than girls, which confirms previous research (Leff & Hoyle, 1995; Lytton & Romney, 1991; Raufelder et al., 2015). Interestingly, no significant gender differences were found for mother or father support, which also aligns with previous findings (Raufelder et al., 2015). Gender-specific perception of parental predictors of test anxiety Pattern for girls The results of the MGSEM largely confirmed the first hypothesis: Mothers (but not fathers) seem to play the primary role in determining their daughters' TA as their model behavior, feedback, and assistance are associated both with their daughters' cognitions in exam-related situations (intrusive, unrelated thoughts and reduced confidence), and with their physical manifestations of anxiety, as evidenced in previous research (e.g., Chen et al., 2000; Duchesne et al., 2009; Gomez & McLaren, 2006; Holahan et al., 1995; Laible & Carlo, 2004; Peleg-Popko, 2004; Raufelder et al., 2015). The high social significance of mothers may relate to the fact that, on the whole, (1) mothers provide more support and acceptance for their adolescent daughters, especially for the expression of negative emotions (Hunter et al., 2011; Stocker, Richmond, Rhoades, & Kiang, 2007), and (2) depending on their communicative styles, their pressure may even have positive effects on their daughters' motivation and emotional regulation (Averill & Power, 1995). Pattern for boys Hypothesis 2 was partly confirmed: Boys' perception of father pressure was positively associated with interference but not with worry, whereas lack of confidence was positively associated with maternal pressure, and negatively with mother support. These findings can be explained by previous studies which haven shown that, unlike girls, boys generally rely on both parents and experience more pressure from either one (Leff & Hoyle, 1995; Lytton & Romney, 1991; Raufelder et al., 2015). According to previous research, however, there appears to be a difference in the nature of support and pressure between mothers and fathers, and the consequences this entails for boys (see Averill & Power, 1995; Brand & Klimes-Dougan, 2010; Lytton & Romney, 1991; Stocker et al., 2007): Mothers have more realistic performance expectations, more

76

T. Ringeisen, D. Raufelder / Journal of Adolescence 45 (2015) 67e79

understanding attitudes, show greater emotional support, and tolerate more autonomous learning styles, which is related to greater enjoyment and effort in adolescents. In contrast, fathers often defend unrealistic performance expectations while being more controlling and less supportive of adolescents' emotions, which may be linked to reduced scholastic effort and socio-emotional maladjustment such as scarce enjoyment and high anxiety. To explain the strong association between perceived father pressure and interference that was found in the current study, fathers' expectations of their sons' academic performance may render boys more aware of pressure which, in turn, might be associated with their impaired concentration during exams (cf. Cassady & Johnson, 2002; Zimmermann, Salem, & Notaro, 2000). Considering the association between support and worry identified in the current study, and in conjunction with the changing gender role of fathers (Yarwood, 2011), fathers might engage in (1) affective yet constraining behavior, characterized by emotional warmth and overprotection, and/or (2) (undesired) informational support, which boys might perceive as controlling and patronizing. As previous research has shown, both kinds of behavior undermine adolescents' autonomy and render them more alert to threatening information, thereby enhancing their evaluation anxiety (e.g., Biller & LopezKimptom, 1997; Leung et al., 2010; Wang & Liu, 2000). Seeing as fathers are generally less emotionally responsive than mothers (Brand & Klimes-Dougan, 2010; Stocker et al., 2007; Zimmermann et al., 2000), even minor efforts in providing emotional attention for their sons would be beneficial. Strengths, limitations and future research The current study is subject to some methodological limitations. Firstly, the cross-sectional nature of our data means that causal relationships between variables cannot be deduced, and the stability of correlations over time remains uncertain. Hence, future longitudinal and/or experimental research that explores the development of TA from childhood to adolescence considering parental behavior is encouraged (Beesdo, Knappe, & Pine, 2009). Secondly, this study is prone to criticism against the exclusive use of self-reported data and thus common-method variance should thus be considered as a potential limitation. However, we employed self-reported measures because we were specifically interested in students' perceptions of parental behavior and the various manifestations of TA that followed, seen as previous research has found parents' and children's interpretations of parenting behavior to be contradictory: Whereas adolescents would often describe their parents as authoritarian, their parents viewed themselves as authoritative (Smetana, 1995). Moreover, research has found that many critiques of self-reported data are unfounded, as the limitations related to this kind of data equally occur with non-selfreported data (Chan, 2009; Spector, 2006). After careful consideration of the four main problems associated with selfreported data (construct validity; interpreting the correlations; social desirability of responses; value of non-self-reported data), we designed a method that would address and circumvent each issue (see Methods section). Thirdly, our findings apply to German adolescents only. Therefore, studies addressing cultural differences in the interplay of parental behavior and test anxiety are welcomed, as there are currently very few studies that examine and/or compare associations between these variables for different cultural groups (cf. McLeod, Wood, & Weisz, 2007). Fourthly, we did not consider differences between students from traditional, two-parent households and students from single-parent households because only 7.4% of the participants identified with the latter category. Therefore, future studies that consider this distinction are warranted. Fifthly, the observed effects in the present study are relatively small. Seeing as other variables may affect test anxiety, such as temper and personality traits (Chamorro-Premuzic, Ahmetoglu, & Furnham, 2008; Hoferichter, Raufelder, & Eid, 2014), we encourage future research to consider additional trait anxiety correlates (e.g., neuroticism) in revealing potential inter-individual differences in students' test anxiety patterns. Despite these limitations, the present study has its strengths. It investigated unique gender-specific patterns in adolescents' perception of parental behavior and its association with four facets of TA in a detailed manner. Firstly, by exploring the role of mothers and fathers separately, this study refines our understanding of gender-specific relations (cf. Brand & KlimesDougan, 2010; Eccles, 2007; Lytton & Romney, 1991) suggesting that fathers and mothers play distinct roles in the cognitive and emotional manifestation of TA in their adolescent children. Previous studies on test and general anxiety either jointly € gels & Phares, 2008; Brand & Klimesassessed the role of both parents or focused explicitly on peers' or teachers' support (Bo Dougan, 2010; Zimmermann et al., 2000). The current findings suggest that the perception of father's behavior (support and pressure) is associated with an increase in cognitive TA facets (worry and interference) for boys only, whereas students' perception of their mother's behavior may have supportive or debilitating effects for girls and boys alike. In line with previous findings, mother support appears to be the most important kind of parental behavior for adolescents irrespective of their gender, as it helps them construct optimistic attitudes about their capacities, and regulate general and test-related anxiety (Chen et al., 2000; Duchesne et al., 2009; Gomez & McLaren, 2006; Holahan et al., 1995; Laible & Carlo, 2004; Peleg-Popko, 2004). However, the connection between adolescents' perception of parental behavior and TA was relatively weak. Perceived pressure/support from either parent only marginally accounts for variance in the four TA facets, as only four of the potential 16 associations were significant for both boys and girls. These patterns reflect the findings of a meta-analytic study, which also found significant albeit weak associations between parenting dimensions and childhood anxiety (McLeod et al., 2007). To understand which additional factors contribute to the development of TA, future studies should examine the quality of parent-child relationships in greater detail, and consider personal characteristics of both parents and their children (e.g., consistency, differentiation and/or interactive effects of parental behavior, trait anxiety levels, gender, age), which might act €gels, Bamelis, & van der Bruggen, 2008; McLeod et al., 2007). The current as moderators and/or mutually affect each other (Bo

T. Ringeisen, D. Raufelder / Journal of Adolescence 45 (2015) 67e79

77

study may thus offer a starting point to validate or refine existing socio-contextual models of achievement and related € gels & Phares, 2008; Brand & Klimes-Dougan, 2010; Eccles, 2007) or emotions by including gender-specific patterns (Bo developmental, gender-specific pathways between parents and their children's anxiety (Ballash, Leyfer, Buckley, & WoodruffBorden, 2006). Furthermore, we followed a fine-grained operationalization of the underlying constructs as we simultaneously examined €gels & cognitive and bodily-affective aspects of TA, as well as different kinds of parental behavior (pressure and support; Bo Phares, 2008; Leung et al., 2010). Finally, our data-set was sourced from a large sample therefore enabling some generalizability of our findings, at the very least to adolescent students in Brandenburg, Germany. Future research may explore the practical implications of establishing and integrating supportive parent-child interactions into the school context. Although parents are aware of their children's test-induced anxiety, they often feel powerless to relieve such anxiety due to insufficient resources and opportunities, and thus expect the school personnel to work through these issues (Babinski & Knotek, 2008; Figueroa, 2013). Therefore, based on the current study and future findings, school health professionals, educators and teachers may wish to introduce gender-specific family and/or parenting programs into the school curriculum in order to raise parents' awareness of the gender-specific effects that their support and pressure may have on their children, and to encourage gender-sensitive parental intervention (for an overview, see von der Embse, Barterian, & Segool, 2013): Since boys are generally more sensitive to pressure, both parents should take special care when exerting pressure on their son. Furthermore, because girls generally exhibit higher levels of TA than boys, mothers should avoid putting pressure on their daughters for school-related matters. In turn, fathers should be aware of their primary gender role function for boys and thus provide verbal and/or emotional support when necessary, while drawing clear distinctions between well-intended suggestions for improving scholastic results and conveying high expectations (cf. Biller & Lopez-Kimptom, 1997; Zimmermann et al., 2000). By raising their awareness of the gendered effects of their behavior, parents may seek guidance to reduce pressures and enhance the positive effects of their support (cf. Haddock, Zimmerman, & MacPhee, 2000; Kaslow, Broth, Smith, & Collins, 2012). Acknowledgments This work was supported by a grant (Schumpeter Fellowship, II/84 452) from the Volkswagen Foundation. The Volkswagen Foundation was involved in none of the following steps: study design, collection, analysis and interpretation of data, writing of the paper, decision to submit the article for publication. The authors would like to thank the principals, teachers, and students for their cooperation in making these studies possible.

References Arnold, K. (1995). Lives of promise: What becomes of high school valedictorians e A fourteen-year study of achievement and live choices. San Francisco, CA: Jossey-Bass. Asparouhov, T. (2005). Sampling weights in latent variable modeling. Structural Equation Modeling, 12, 37e41. http://dx.doi.org/10.1073/pnas.0910967107. Averill, P. M., & Power, T. G. (1995). Parental attitudes and children's experiences in soccer: correlates of effort and enjoyment. International Journal of Behavioral Development, 18, 263e276. http://dx.doi.org/10.1016/0163-6383(95)90015-2. Babinski, L., & Knotek, S. E. (2008). Managing student's test anxiety: parents' perceptions of an unmet need. In Paper presented at the meeting of the American Psychological Association, Boston, MA. Ballash, N., Leyfer, O., Buckley, A., & Woodruff-Borden, J. (2006). Parental control in the etiology of anxiety. Clinical Child and Family Psychology Review, 9, 113e133. http://dx.doi.org/10.1007/s10567-006-0007-z. Beesdo, K., Knappe, S., & Pine, D. S. (2009). Anxiety and anxiety disorders in children and adolescents: developmental issues and implications for DSM-V. Psychiatric Clinics of North America, 32, 483e524. http://dx.doi.org/10.1016/j.psc.2009.06.002. Besharat, M. A. (2003). Parental perfectionism and children's test anxiety. Psychological Reports, 93, 1049e1055. http://dx.doi.org/10.2466/PR0.93.8.10491055. Biller, H. B., & Lopez-Kimptom, J. (1997). The father and the school-aged child. In M. E. Lamb (Ed.), The role of the father in child development (3rd ed., pp. 143e161). New York, NY: John Wiley & Sons. Boekaerts, M. (1995). Self-regulated learning: bridging the gap between metacognitive and metamotivation theories. Educational Psychologist, 30, 195e200. http://dx.doi.org/10.1207/s15326985ep3004_4. €gels, S., Bamelis, L., & van der Bruggen, C. (2008). Parental rearing as a function of parent's own, partner's, and child's anxiety status: fathers make the Bo difference. Cognition & Emotion, 22, 522e538. http://dx.doi.org/10.1080/02699930801886706. €gels, S., & Phares, V. (2008). Fathers' role in the etiology, prevention and treatment of child anxiety: a review and new model. Clinical Psychology Review, Bo 28(4), 539e558. Bois, J. E., Lalanne, J., & Delforge, C. (2009). The influence of parenting practices on children's and adolescents' pre-competitive anxiety. Journal of Sports Sciences, 27(10), 995e1005. http://dx.doi.org/10.1080/02640410903062001. Brand, A. E., & Klimes-Dougan, B. (2010). Emotion socialization in adolescence: the roles of mothers and fathers. New Directions for Child and Adolescent Development, 128, 85e100. http://dx.doi.org/10.1002/cd.270. Brown, T. A. (2006). Confirmatory factor analysis. New York, NY: Guilford. Byrne, B. M. (2008). Testing for multigroup equivalence of a measuring instrument: a walk through the process. Psicothema, 20(4), 872e882. Cassady, J. C., & Johnson, R. E. (2002). Cognitive test anxiety and academic performance. Contemporary Educational Psychology, 27, 270e295. http://dx.doi. org/10.1006/ceps.2001.1094. Chamorro-Premuzic, T., Ahmetoglu, G., & Furnham, A. (2008). Little more than personality: dispositional determinants of test anxiety (the big five, core selfevaluations, and self-assessed intelligence). Learning and Individual Differences, 18, 258e263. http://dx.doi.org/10.1016/j.lindif.2007.09.002. Chan, D. (2009). So why ask me? Are self-report data really that bad? In C. E. Lance, & R. J. Vandenberg (Eds.), Statistical and methodological myths and urban legends (pp. 309e336). New York, NY: Routledge. Chapell, M. S., & Overton, W. F. (1998). Development of logical reasoning in the context of parental style and test anxiety. Merrill-Palmer Quarterly, 44(2), 141e156.

78

T. Ringeisen, D. Raufelder / Journal of Adolescence 45 (2015) 67e79

Chen, X., Liu, M., & Li, D. (2000). Parental warmth, control, and indulgence and their relations to adjustment in Chinese children: a longitudinal study. Journal of Family Psychology, 14, 401e419. http://dx.doi.org/10.1037//0893-3200.14.3.401. Davis, M. C., Burleson, M. H., & Kruszewski, D. M. (2011). Gender: its relationship to stressor exposure, cognitive appraisal, coping processes, stress responses, and health outcomes. In R. J. Contrada, & A. Baum (Eds.), The handbook of stress science: Biology, psychology, and health (pp. 247e267). New York, NY: Springer. DeWit, D. J., Karioja, K., Rye, B. J., & Shain, M. (2011). Perceptions of declining classmate and teacher support following the transition to high school: correlates of increasing student mental health difficulties. Psychology in the Schools, 43, 556e572. http://dx.doi.org/10.1002/pits.20576. Ding, L., Velicer, W. F., & Harlow, L. L. (1995). Effects of estimation methods, number of indicators per factor, and improper solutions on structural equation modeling fit indices. Structural Equation Modeling Journal, 2, 119e144. http://dx.doi.org/10.1080/10705519509540000. Duchesne, S., Ratelle, C. F., Poitras, S. C., & Drouin, E. (2009). Early adolescent attachment to parents, emotional problems, and teacher e academic worries about the middle school transition. Journal of Early Adolescence, 29, 743e766. http://dx.doi.org/10.1177/0272431608325502. Eccles, J. S. (2007). Families, schools, and developing achievement-related motivations and engagement. In J. E. Grusec, & P. D. Hastings (Eds.), Handbook of socialization: Theory and research (pp. 665e691). New York, NY: Guilford Press. Else-Quest, N. M., Higgins, A., Allison, C., & Morton, L. C. (2012). Gender differences in self-conscious emotional experience: a meta-analysis. Psychological Bulletin, 138, 947e981. http://dx.doi.org/10.1037/a0027930. von der Embse, N., Barterian, J., & Segool, N. (2013). Test anxiety interventions for children and adolescents: a systematic review of treatment studies from 2000e2010. Psychology in the Schools, 50, 57e71. http://dx.doi.org/10.1002/pits.21660. Figueroa, L. (2013). Teachers' awareness and skills in addressing students with anxiety symptoms. PCOM Psychology Dissertations. Paper 281. Retrieved from: http://digitalcommons.pcom.edu/cgi/viewcontent.cgi?article¼1277&context¼psychology_dissertations. Flouri, E. (2008). Fathering and adolescents' psychological adjustment: the role of fathers' involvement, residence and biology status. Child Care, Health and Development, 34, 152e161. http://dx.doi.org/10.1111/j.1365-2214.2007.00752.x. Frey, A., Ruchkin, V., Martin, A., & Schwab-Stone, M. (2009). Adolescents in transitions: school and family characteristics in the development of violent behaviors entering high school. Child Psychiatry and Human Development, 40, 1e13. http://dx.doi.org/10.1007/s10578-008-0105-x. Gherasim, L. R., & Butnaru, S. (2012). The effort attribution, test anxiety and achievement in sciences: the moderating effect of parental behavior. The International Journal of Learning, 18(10), 283e294. Glynn, A. (2013). What can we learn with statistical truth serum? Design and analysis of the list experiment. Public Opinion Quarterly, 77(S1), 159e172. http://dx.doi.org/10.1093/poq/nfs070. Gomez, R., & McLaren, S. (2006). The association of avoidance coping style, and perceived mother and father support with anxiety/depression among late adolescents: applicability of resiliency models. Personality and Individual Differences, 40, 1165e1176. http://dx.doi.org/10.1016/j.paid.2005.11.009. Haddock, S. A., Zimmerman, T. S., & MacPhee, D. (2000). The power equity guide: attending to gender in family therapy. Journal of Marital and Family Therapy, 26, 153e170. http://dx.doi.org/10.1111/j.1752-0606.2000.tb00286.x. Hanewald, R. (2013). Transition between primary and secondary school: why it is important and how it can be supported. Australian Journal of Teacher Education, 38, 62e74. http://dx.doi.org/10.14221/ajte.2013v38n1.7. Hazzard, A., Christensen, A., & Margolin, G. (1983). Children's perceptions of parental behaviors. Journal of Abnormal Child Psychology, 11(1), 49e59. Hill, K. T., & Wigfield, A. (1984). Test anxiety: a major educational problem and what can be done about it. The Elementary School Journal, 85(1), 105e126. € ttingen, Germany: Hogrefe. Hodapp, V., Rohrmann, S., & Ringeisen, T. (2011). Prüfungsangstfragebogen (PAF) [German test anxiety questionnaire]. Go Hoferichter, F., Raufelder, D., & Eid, M. (2014). The mediating role of socio-motivational relationships in the interplay of perceived stress, neuroticism, and test anxiety among adolescent students. Psychology in the Schools. http://dx.doi.org/10.1002/pits.21778 (Advance online publication). Holahan, C. J., Valentiner, D. P., & Moos, R. H. (1995). Parental support, coping strategies, and psychological adjustment: an integrative model with late adolescents. Journal of Youth and Adolescence, 24, 633e648. http://dx.doi.org/10.1007/BF01536948. Hu, L. T., & Bentler, P. M. (1999). Cutoff criteria for fit indexes in covariance structural analysis: conventional criteria versus new alternatives. Structural Equation Modeling: A Multidisciplinary Journal, 6, 1e55. http://dx.doi.org/10.1080/10705519909540118. Hunter, E. C., Fainsilber Katz, L., Wu Shortt, J., Davis, B., Leve, C., Allen, N. B., & Sheeber, L. B. (2011). How do I feel about feelings? Emotion socialization in families of depressed and healthy adolescents. Journal of Youth and Adolescence., 40, 428e441. http://dx.doi.org/10.1007/s10964-010-9545-2. Jacobs, J. E., & Eccles, J. S. (1992). The impact of mothers' gender-role stereotypic beliefs on mothers' and children's ability perceptions. Journal of Personality and Social Psychology, 63, 932e944. Kaslow, N. J., Broth, M. R., Smith, C. O., & Collins, M. H. (2012). Family-based interventions for child and adolescent disorders. Journal of Marital and Family Therapy, 38, 82e100. http://dx.doi.org/10.1111/j.1752-0606.2011.00257.x. Kennelly, L., & Monrad, M. (Eds.). (2007). Easing the transition to high school: Research and best practices designed to support high school learning. Washington, DC: National High School Center at the American Institutes for Research. Kenny, M. E., Lomax, R., Brabeck, M., & Fife, J. (1998). Longitudinal pathways linking adolescent reports of maternal and paternal attachments to psychological well-being. Journal of Early Adolescence, 18, 221e243. http://dx.doi.org/10.1177/0272431698018003001. €sel (Eds.), Health hazards in adolescence (pp. 115e130). New Krohne, H. W. (1990). Parental childrearing and anxiety development. In K. Hurrelmann, & F. Lo York, NY: De Gruyter. Laible, D. J., & Carlo, G. (2004). The differential relations of maternal and paternal support and control to adolescent social competence, self-worth, and sympathy. Journal of Adolescent Research, 19, 759e782. http://dx.doi.org/10.1177/0743558403260094. Leff, S. S., & Hoyle, R. H. (1995). Young athlete's perceptions of parental support and pressure. Journal of Youth and Adolescence, 24, 187e203. http://dx.doi. org/10.1007/BF01537149. Leung, G. S. M., Yeung, K. C., & Wong, D. F. K. (2010). Academic stressors and anxiety in children: the role of paternal support. Journal of Child and Family Studies, 19, 90e100. http://dx.doi.org/10.1007/s10826-009-9288-4. Little, T. D., Cunningham, W. A., Shahar, G., & Widaman, K. F. (2002). To parcel or not to parcel: exploring the question, weighing the merits. Structural Equation Modeling, 9(2), 151e173. http://dx.doi.org/10.1207/S15328007SEM0902_1. Lowe, P. A., Lee, S. W., Witteborg, K. M., Prichard, K. W., Luhr, M. E., Cullinan, C. M., et al. (2008). The test anxiety inventory for children and adolescents (TAICA). Journal of Psychoeducational Assessment, 26, 215e230. http://dx.doi.org/10.1177/0734282907303760. Lytton, H., & Romney, D. M. (1991). Parents' differential socialization of boys and girls: a meta-analysis. Psychological Bulletin, 109, 267e296. http://dx.doi. org/10.1037/0033-2909.109.2.267. Maccoby, E. E. (2002). Parenting effects: issues and controversies. In J. G. Borkowski, & S. L. Ramey (Eds.), Monographs in parentingParenting and the child's world: Influences on academic, intellectual, and socialeemotional development (pp. 35e46). Mahwah, NJ: Erlbaum. Martin, C. L. (2000). Cognitive theories of gender development. In T. Eckes, & H. M. Trautner (Eds.), The developmental social psychology of gender (pp. 91e121). Mahwah, NJ: Lawrence Erlbaum. Martinez, R. S., Aricak, O. T., Graves, M. N., Peters-Myszak, J., & Nellis, L. (2011). Changes in perceived social support and socioemotional adjustment across the elementary to junior high school transition. Journal of Youth and Adolescence, 40(5), 519e530. http://dx.doi.org/10.1007/s10964-010-9572-z. McLeod, B. D., Wood, J. J., & Weisz, J. R. (2007). Examining the association between parenting and childhood anxiety: a meta-analysis. Clinical Psychology Review, 27, 155e172. http://dx.doi.org/10.1016/j.cpr.2006.09.002. n, L. K., & Muthe n, B. O. (1998e2012). Mplus user's guide (7th ed.). Los Angeles, CA: Muthe n & Muthe n. Muthe Ommundsen, Y., Roberts, G. C., Lemyre, P. N., & Miller, B. W. (2006). Parental and coach support or pressure on psychosocial outcomes of pediatric athletes in soccer. Clinical Journal of Sport Medicine, 16(6), 522e526. Osbourne, J. W. (2006). Gender, stereotype threat and anxiety: psychophysiological and cognitive evidence. Journal of Research in Educational Psychology, 8(4), 109e138.

T. Ringeisen, D. Raufelder / Journal of Adolescence 45 (2015) 67e79

79

Papini, D. R., & Roggman, L. A. (1992). Adolescent perceived attachment to parents in relation to competence, depression, and anxiety e a longitudinal study. Journal of Early Adolescence, 12, 420e440. http://dx.doi.org/10.1177/0272431692012004005. Payne, K. E. (2001). Different but equal: Communication between the sexes. Westport, CT: Praeger. Peleg-Popko, O. (2004). Differentiation and test anxiety in adolescents. Journal of Adolescence, 27, 645e662. http://dx.doi.org/10.1016/j.adolescence.2004.06. 002. Peleg-Popko, O., & Klingman, A. (2002). Family environment, discrepancies between perceived actual and desirable environment, and children's test and trait anxiety. British Journal of Guidance & Counselling, 30, 451e466. http://dx.doi.org/10.1080/0306988021000025646. Perkins, A. M., & Corr, P. J. (2005). Can worriers be winners? The association between worrying and job performance. Personality and Individual Differences, 38, 25e31. http://dx.doi.org/10.1016/j.paid.2004.03.008. Pettit, G. S., Laird, R. D., Bates, J. E., Dodge, K. A., & Criss, M. M. (2001). Antecedents and behavior-problem outcomes of parental monitoring and psychological control in early adolescence. Child Development, 72, 583e598. http://dx.doi.org/10.1111/1467-8624.00298. Putwain, D. W., Woods, K. A., & Symes, W. (2010). Personal and situational predictors of test anxiety of students in post compulsory education. British Journal of Educational Psychology, 80, 137e160. http://dx.doi.org/10.1348/000709909X466082. Raufelder, D., Hoferichter, F., Ringeisen, T., Regner, N., & Jacke, C. (2015). The perceived role of parental support and pressure in the interplay of test anxiety and school engagement among adolescents: evidence for gender-specific relations. Journal of Child and Family Studies. http://dx.doi.org/10.1007/ s10826-015-0182-y (Advance online publication). Reitzle, M., Metzke, C. W., & Steinhausen, H.-C. (2001). Eltern und Kinder: der Zürcher Kurzfragebogen zum Erziehungsverhalten (ZKE) [Parents and children: the Zürich Questionnaire of educational behavior (ZKE)]. Diagnostica, 47, 196e207. http://dx.doi.org/10.1026//0012-1924.47.4.196. Ringeisen, T., Raufelder, D., Schnell, K., & Rohrmann, S. (2015). Validating the proposed structure of the relationships among test anxiety and its predictors based on control-value theory: evidence for gender-specific patterns. Educational Psychology. http://dx.doi.org/10.1080/01443410.2015.1072134 (Advance online publication). Ruble, D. N., & Martin, C. L. (1998). Gender development. In W. Damon (Series Ed.) & N. Eisenberg (Vol. Ed.) (5th ed.,, Handbook of child psychology: Vol. 3. Social, emotional, and personality development (pp. 933e1016). New York, NY: Wiley. Selkirk, L., Bouchey, H., & Eccles, J. (2011). Interactions among domain-specific expectancies, values, and gender: predictors of test anxiety during early adolescence. Journal of Early Adolescence, 31, 361e389. http://dx.doi.org/10.1177/0272431610363156. Shadach, E., & Ganor-Miller, O. (2013). The role of perceived parental over-involvement in student test anxiety. European Journal of Psychology of Education, 28, 585e596. http://dx.doi.org/10.1007/s10212-012-0131-8. Simpkins, S. D., Fredricks, J. A., & Eccles, J. S. (2012). Charting the Eccles' expectancy-value model from mothers' beliefs in childhood to youths' activities in adolescence. Developmental Psychology, 48, 1019e1032. http://dx.doi.org/10.1037/a0027468. Smetana, J. G. (1995). Morality in context: abstractions, ambiguities, and applications. In R. Vasta (Ed.), Annals of child development (Vol. 10, pp. 83e130). London, England: Jessica Kingsley Publishers. Spector, P. E. (2006). Method variance in organizational research: truth or urban legend? Organizational Research Methods, 9, 221e232. http://dx.doi.org/10. 1177/1094428105284955. Spielberger, C. D., & Vagg, P. R. (1995). Test anxiety: a transactional process model. In C. D. Spielberger, & P. R. Vagg (Eds.), Test anxiety: Theory, assessment, and treatment (pp. 1e14). Washington, DC: Taylor & Francis. Stocker, C. M., Richmond, M. K., Rhoades, G. K., & Kiang, L. (2007). Family emotional processes and adolescents' adjustment. Social Development, 16, 310e325. http://dx.doi.org/10.1111/j.1467-9507.2007.00386.x. Stoeber, J., Feast, A. R., & Hayward, J. A. (2009). Self-oriented and socially prescribed perfectionism: differential relationships with intrinsic and extrinsic motivation and test anxiety. Personality and Individual Differences, 47, 423e428. http://dx.doi.org/10.1016/j.paid.2009.04.014. Wang, J., & Liu, G. (2000). The relationship between the anxiety of senior students in secondary school and their parenting style. The Journal of Jiangxi Education and Science, 6, 30e32 (in Chinese). Wang, J., & Wang, X. (2012). Structural equation modeling: Applications using Mplus. Chichester, UK: Wiley. West, S. G., Finch, J. F., & Curran, P. J. (1995). Structural equation models with nonnormal variables. Problems and remedies. In R. H. Hoyle (Ed.), Structural equation modelling: Concepts, issues and applications (pp. 56e75). Newbury Park, CA: Sage. Wolfradt, A. U., Hempel, S., & Miles, J. N. V. (2003). Perceived parenting styles, depersonalisation, anxiety and coping behaviour in adolescents. Personality and Individual Differences, 34, 521e532. http://dx.doi.org/10.1016/S0191-8869(02)00092-2. Yarwood, A. G. (2011). The pick and mix of fathering identities. Fathering: A Journal of Theory, Research and Practice About Men as Fathers, 9, 150e168. http:// dx.doi.org/10.3149/fth.0902.150. Yuan, K.-H., & Bentler, P. M. (2004). On chi-square difference and z-tests in mean and covariance structure analysis when the base model is misspecified. Educational and Psychological Measurement, 64, 737e757. http://dx.doi.org/10.1177/0013164404264853. Zeidner, M. (2007). Test anxiety in educational contexts: concepts, findings, and future directions. In P. A. Schutz, & R. Pekrun (Eds.), Emotion in education (pp. 165e184). San Diego, CA: Academic Press. Zimmermann, M. A., Salem, D. A., & Notaro, P. C. (2000). Make room for Daddy II: the positive effects of fathers' role in adolescent development. In R. D. Taylor, & M. C. Wang (Eds.), Resilience across contexts (pp. 233e253). Mahwah, NJ: Lawrence Erlbaum. Zohar, D. (1998). An additive model of test anxiety: role of exam-specific expectations. Journal of Educational Psychology, 90, 330e340. http://dx.doi.org/10. 1037/0022-0663.90.2.330.