versão ISSN 1808-5687
Rev. bras.ter. cogn. vol.5 no.2 Rio de Janeiro nov. 2009
Anxiety, depression and coping strategies in adolescence: psychometric issues and proposal of a reduced version
Margarida Gaspar de Matos1; Aristides Ferreira2; Gina Tomé3; Ana Inês Borges3; Dina Manso3 Paula Ferreira4
1Clinical & Health Psychologist; Professor at FMH/ UTL (Technical University of Lisbon/ Portugal); Researcher at CMDT/ IHMT/UNL (Institute of Hygiene and Tropical Medicine) ; Aventura Social Research Team
2Organizational Psychologist; Researcher at FMH/ UTL (Technical University of Lisbon/ Portugal); Aventura Social Research Team & Minho University/ Portugal
3Psychologists, Master in Cognitive Behavioural Therapies; Researchers at FMH/ UTL (Technical University of Lisbon/ Portugal) Aventura Social Research Team
4Researcher at FPCEUL (University of Lisbon/Portugal)
This study used depression (CDI – Kovacs, 1981), anxiety (MASC – March, 1997) and coping strategy (CRI-Y – Moos, 1993) scales and studied the fitness of reduced versions. The sample consisted of 916 Portuguese pupils, 54.3% fiinine, aged 10 to 21 years old. The participants were randomly selected from public schools nation-wide. Two classes were chosen from the 5th to the 12th grade.
A set of principal component analyses was carried out in a randomly chosen sample (n = 394) and three reduced measures were found to be strongly correlated with the previous. Confirmatory factor analyses (CFA) using the other part of the sample (n = 522) revealed adjustment indexes suggesting a good fit for both the whole model and gender and age groups separately. All scales revealed a good internal consistency. According to our results, girls were more anxious and developed more coping strategies than boys. Older students tended to be less depressed, while younger adolescents present higher scores in CDI-R (depression) and CRI-R (coping) scales.
Keywords: Depression, Anxiety, Coping, Reduced version, School-based interventions.
Several studies are unanimous in considering depression as a very common pathology in childhood and adolescence. The heterogeneity of depressive symptoms is related to different periods of childhood/ adolescence - that is, when they begin co morbidity (Harrington, Rutter, & Trombone, 1996; Mash & Wolfe, 2002).
In general, studies suggest that depression symptoms appear by the age of 12 and boys present higher values of depression (McGee, Feehan, Williams, & Anderson, 1992). As age increases, there is a substantial increase of the number of depression symptoms. According to some researchers, the differences can be verified at a deeper level between the ages of 13 and 15 years, which drastically increases between the ages of 15 and 18, where there’s a bigger prevalence among the fiinine gender (Duggal, Carlson, Srouf, & Egeland, 2001; Schraedly, Gotlib, & Hayward, 1999).
Anxiety appears as a common, functional and transitory experience and its nature and intensity can vary to a large extent, depending on the individuals’ period of development, allowing in this sense, children and adolescents to engage in new, unexpected or dangerous situations (Rosen & Schulkin, 1998). However, the intensity of anxiety can increase and often become chronic and dysfunctional from a social and iotional point of view (Fonseca, 1998).
From a sample of adolescents between the ages of 12 and 17, Essau, Conradt and Petermann (2000) found that fiales presented more anxiety symptoms than males and that these symptoms tend to increase with age - generally between the ages of 12 and 15.
Some studies have dionstrated the existence of a relation between anxiety and depression in adolescents and adults (Kovacs & Devlin, 1998; Pine, Cohen, Gurley, Brookm, & Yuju Me, 1998). In order to analyse this association, Matos, Barrett, Dadds and Short (2003) carried out a cross-sectional study with Portuguese children and adolescents, aged 10 to 17 attending 6th, 8th and 10th grade in public schools. The subjects were also evenly distributed in terms of gender. These researchers confirmed a significant association between depression and anxiety and superior frequency levels of anxiety and/or depression symptoms in fiales.
Children and adolescents use a great diversity of coping answers in different domains (in their school performance, family-life and social environment). They differ nonetheless, in stress situation evaluation and in coping strategy assessment. The effects and type of answer in any given situation that causes stress, depend on personal characteristics and abilities (Wenger, Sharrer, & Wynd, 2000).
Some researchers reported the existence of a positive relation between maladaptive coping strategies and different disorders (Endler & Parker, 1990; Holahan, Moos, & Schaefer, 1996). Selffge-Krenke (2000) stated that children and adolescents with pathological behaviours use maladaptive coping strategies, which in the future can increase their pathological behaviours, resulting in a vicious cycle.
Depressed adolescents showed a higher number of stressful life events and a higher use of maladaptive coping strategies before the onset of these events (Selffge-Krenke, 2000).
Byrne (2000) analysed the relations between anxiety, fear, self-estei and coping strategies, in a sample of 224 adolescents attending 7th, 9th and 12th grade. Results suggested that males present a significant decrease of anxiety and fear in the 12th grade and that by this time, both, males and fiales use different coping strategies in order to deal with their fear and anxiety.
These studies suggested the need to simultaneously take and analyse the three measures. However, in school-based interventions, if it was possible to focus in the most relevant features of these measures for non-clinical adolescent populations, a reduced version of those scales would avoid a time consuming assessment.
The aim of this study is thus to check for redundant (or not so relevant) itis, considering three classical scales, commonly used to evaluate depression, anxiety and coping strategies, and then try to propose a reduced version of the three scales that could be used for screening procedures, in school settings.
It is also our intent to analyse the relationship between both the total scales and the respective reduced scales, in a school population of adolescents, as well as check the effects of age, gender and family type.
The study followed strict international ethical and privacy norms, namely, the faculty’s scientific board approval, school authorizations, parental consent, anonymity guidelines and voluntary-based participation.
Researchers were chartered psychologists. Questionnaires were administered in the classroom and took an average of 50 minutes to fill in.
SPSS 16.0 software was used in order to carry out a set of principal component analyses (PCA) aiming at getting the more representative factor from the anxiety (MASC), depression (CDI) and coping (CRI-Y) scales.
The sample consisted of 916 Portuguese adolescents, 54.3% of which were fiinine, aged 10 to 21 (M = 14.4; SD = 2.62). As for family type, 72.5% of the subjects were found to have been living in a nuclear family. The participants were selected from public schools nation-wide, and two classes were randomly chosen from each of the 8 grades chosen, that is, from the 5th to the 12th grade. By using SPSS, we randomly chose two different samples. A first sample consisted of 394 individuals (217 of which were fiinine) between the ages of 10 and 21 (M = 14.4 years and SD = 2.58). The second sample included 522 subjects (289 of which were fiinine) between the ages of 10 and 21 (M = 14.3 years and SD = 2.66).
Children's Depression Inventory - CDI (Kovacs, 1981), the Multidimensional Anxiety Scale for Children - MASC (March, 1997) and the Coping Responses Inventory - Youth Form – CRI-Y (Moos, 1993) were used. Portuguese versions of these scales were available from the Psychology Laboratory of Lusofona University (Lisbon-Portugal), where international rules for independent translation, independent back translation and consensus procedure are systiatically carried out.
The CDI consists of a self-report inventory of 27 itis with a three point scale (1. Absence of symptoms; 2. Moderate symptoms; 3. Severe symptoms). The total result varies between 27 and 81 points. A higher classification corresponds to a more depressive situation. Kovacs (1981) referred satisfactory levels of internal consistency, with alpha coefficients varying between .71 and .89 for the total scale.
The originally scale included several sub-scales that were not taken in consideration for this study, once specific factorial analysis procedures were aimed.
The MASC encloses a self-report scale consisting of 39 itis with a four point scale (1- Never or almost never true; 2 - Rarely true; 3 - Sometimes true; 4 -Frequently true). The total result varies between 39 and 156. A higher classification corresponds to more anxiety. March (1997) referred satisfactory levels of internal consistency, with alpha coefficients varying between .88 and .89 for the total scale.
The originally scale included several sub-scales that were not taken in consideration for this study, once specific factorial analysis procedures were aimed.
The CRI-Y is a self-report inventory consisting of two parts. In the first section, participants are asked to describe a probli or stressful situation and answers are grouped in different categories. In the present study for example, the stressful events were recoded in five groups: physical health, family, school, friends and “others”.
The second part of the CRI includes 48 itis, which evaluate coping strategies. Participants select how they usually deal with the situation described in the first part, in a four points scale (1 - No, never; 2 - One or two times; 3 - Yes, sometimes; 4 -Frequent).
The CRI-Y scale is composed of four approach coping subscales including: Logical Analysis, Positive Reappraisal, Seeking Guidance and Support, and Probli Solving. It includes four avoidance coping subscales as well, namely, Cognitive Avoidance, Resignation or Acceptance, Seeking Alternative Rewards, and iotional Discharge. This instrument measures different types of behavioral and cognitive efforts directed at handling stressful situations and/or their aftermath. Each individual subscale yields a score ranging from 6 to 24. Additionally, there is a general summary score measuring approach strategies and one that measures avoidance strategies.
In the original version, the author considered two dimensions (active/confronting coping and passive/avoidance coping). In the present study, data did not allow the assumption of two separate factors and therefore, only one coping dimension (CRI) was considered. Accordingly, results vary between 48 and 192 for each dimension and a higher classification corresponds to a more frequent use of coping strategies.
The CRI-Y has been used with over 400 adolescents (masculine and fiinine) in various contexts (health, depression, and chronic illness). The author of the scale (Moos, 1993) referred satisfactory levels of internal consistency alpha coefficients, with alpha coefficients between .68 and .79 for active/confronting coping and between .59 and .72 for the passive/avoidance coping. No alpha coefficients were available for the total CRI scale.
In the present study, as said, only the scale totals of the CDI, the MASC and the CRI have been studied.
As previously mentioned the main objective of this study was to obtain three independent measures to evaluate the depression, anxiety and coping.
The PCA was developed on the total of itis, which integrate each of the scales used and a first random sample of 394 adolescents was used.
The PCA with Kaiser Criteria was then conducted to extract the higher component (in each separable scale) with eigenvalues greater than one. To interpret the factors a varimax orthogonal rotation was used. As a rule of thumb, only itis with loadings of .32 and above were interpreted (Tabacknick & Fidell, 2001). Some factors loaded only one or two itis, which were hereafter eliminated, taking into account Hakstian, Rogers and Cattell (1982) considerations. According to Ford, MacCallum and Tait (1986), we also deleted itis with loadings greater than .40 in two or more factors. This statistical procedure allowed us to develop three uni-factorial scales that measure depression, anxiety and coping strategy constructs. Those scales represent those itis that have higher loadings in the first PCA component (Table 1).
The PCA’s allowed us to analyze 12 itis (KMO = .894; χ2=3787.247, df = 351) associated with the first component, which as a consequence, became integrated components of the new reduced version of the depression scale (CDI-R). The reduced version of the anxiety scale (MASC-R) resulted from a first factor with 9 itis (KMO = .842; χ2=3603, 495 df = 741). Lastly, an instrument used to evaluate coping strategies ie rged with six itis, which were an integrated part of the first component of the PCA (KMO = .852; χ2=3748.561, df = 1128).
Table 1. Rotated first components loadings, eigenvalues and explained variance after PCA
In order to test the correlations between the three latent constructs obtained through the previous PCA, a confirmatory model was constructed (Figure 1) using the AMOS 6.0 (Arbuckle, 2005) program. With this aim, the second sample of the study was used (n= 522), which corresponds to the subjects that didn’t belong to the first sample. The CFA was conducted based on the covariance matrix while using maximum likelihood estimation. The overall fit was assessed through the following indices of fit: Chi-square, root mean square residual (RMR), root mean square error of approximation (RMSEA), comparative fit index (CFI), incriental fit index (IFI) and Akaike’s information criterion (AIC). To clarify, the AIC is used in the comparison of two or more models, with smaller values representing a better fit of the hypothesized model (Hu & Bentler, 1995). Also, RMR represents the average residual value derived from the fitting of the hypothesized model for the sample data variance-covariance matrix. In a well-fitting model, the value must be close or under .05 (Byrne, 2001). CFI and IFI values close to 1 indicate a very good fit (Bentler, 1990). Values of the RMSEA of about .05 or less would indicate a close fit of the model (Browne & Cudeck, 1993). Adjustment indexes thus suggest a good fit (c2=630.017, df =296; RMR =.032; RMSEA = .035; AIC = 792.017; CFI = .961, IFI = .961). All iti values were significant (p < .01) and had completely standardized factor loadings higher than .392 and error variances with values ranging from .225 to .770.
Figure 1. Global confirmatory model of the relation between depression (CDI-R), anxiety (MASC-R) and coping strategies (CRI-R)
The three factors included in the present model explain 63.3% of the variance for the latent construct depression (CDI-L), 30.5% of the variance for the latent construct anxiety (MASC-L) and 31.0% of the variance for the latent construct coping strategies (CRI-L). As for covariances between latent constructs, there is a significant negative association between depression (CDI-L) and coping strategies (CRI-L) (cov (CDI-L,CRI-L) = -.086, p<. 01). Association among the other latent constructs was not significant.
The invariance of the model according to gender and age was also tested. The model riains adjusted for girls (c2=501.125, df =321; RMR= .045; RMSEA = .044; AIC = 615.125; CFI = .937, IFI = .938) and closely fitted for boys (c2=569.606, df =321; RMR = .048; RMSEA = .058; AIC = 683.606; CFI = .896, IFI = .898). Lastly, if we consider the age factor, both the younger group (N= 202; 11 or less) (c2=440,939, df =321; RMR= .050; RMSEA = .043; AIC = 554.939; CFI = .936, IFI = .937) and the intermediate age group (N= 258; ages between 12 and 15) (c2=489.630, df =321; RMR= .036; RMSEA = .045; AIC = 603.630; CFI = .908, IFI = .910) generated adjusted models. We also verified that the model which integrates the sample including masculine subjects, presents worse fit indices, which may be acceptable if we consider the fit values mentioned in the literature. The AIC indices allowed us to confirm that of all the tested models, the younger group of adolescents presents the lowest values, which suggests a better adjustment of the model for populations with similar characteristics.
Descriptive analysis and alpha consistency of the reduced measures for the CDI, the MASC and the CRI for the two samples studied are shown on Table 2. According to Nunnaly (1978) reliabilities of .70 or greater are sufficient. In both samples, the sub-scale shows good results for the measures of the CDI-R (.92 and .91, respectively), the MASC-R (.80 and .74, respectively) and the CRI-R (.82 and .80, respectively).
Table 2. Descriptive Statistics and internal consistency for the CDI- R, MASC-R, CRI-R total scores
We tried to determine in sample 2 whether any of the riaining itis in the three reduced versions were probliatic. Correlations between the response to a particular iti and the sum of the responses to all other itis were obtained for each scale. Scale scores for each dimension were obtained for each participant by summing up the ratings of all itis in each of the three scales (CDI-R, MASC-R, and CRI-R). The iti/total correlation ranged from .482 to .877 for CDI-R, .580 to .668 for MASC-R, and .589 to .703 for CCRI-R - as all the correlation coefficients were greater than .450. An application of the Kolmogorov-Smirnov test was then developed and showed that scale scores were normally distributed for each of the three reduced scales. Reduced measures were strongly, significantly and positively correlated with the respective total scale (CDI and CDI-R, r =.888, p < .001; MASC and MASC-R, r = .697, p < .001; CRI and CRI-R, r =.704, p < .001).
Concerning gender variables, girls scored significantly higher than boys for the MASC-R [t(519) = -2.841, p = .05, r = .12] and CRI-R [t(511) = -2.782, p = .006, r = .12] subscales. Despite the significant test results, those size effect values are very lower. The Scheffé post hoc test indicated that students from the 5th and 6th grades reported higher CDI-R scores (p < .001) than all the other student grades. In accordance with this, the younger group (n = 202; 11 years old or less) showed significantly higher values in the CDI-R and CRI-R scales when compared with different age groups, such as the intermediate age group (n = 258; ages between 12 and 15 years old) and the older group (n = 62; 16 years old or more).
Reduced scales are proposed to assess depression (the CDI-R), anxiety (the MASC-R) and coping strategies (the CRI-R) in school settings, while targeting, adolescent populations. Initially, the aimed age group for this study ranged from 10 to 18 years old. However, a small number of pupils attending 12th grade were older than 18 (5% from 19 to 21 years old), and were not excluded from the present analysis, in the sake of ecological validity in the instruments in school settings. Scales were reduced by means of the PCA while using half the sample extracted randomly, and therefore, confirmed by the CFA by using the other half.
The PCA with a varimax rotation allowed the extraction of one main component representative of each scale construct. This analysis allowed us to obtain three independent constructs destined to measure depression (the CDI-R), anxiety (the MASC-R) and coping strategies (the CRI-I). The internal consistencies of the reduced scales obtained for sample 1 and confirmed in sample 2 are higher for the CDI-R and the CRI-R than those found in the literature (Kovacks, 1981; Moos, 1993). However, the reduced anxiety scale (the MASC-R) revealed lower scores than those found with the original version (March, 1997). Despite this, the scale with the highest internal reliability is the CDI-R.
Those reduced scales are strongly correlated with the respective and original total scales, but besides being shorter, sei to be accurate for gender and age differentiation. Within a second different sample, the CFA revealed to be supportive of the model obtained in the PCA. By using the reduced form of the Anxiety, Coping strategies and Depression scales, the odd and unusual positive correlation between Anxiety and Coping strategies seis to fade and a new significantly negative covariance between Coping strategies and depression is highlighted. According to the literature, the negative covariance may be explained with the fact that a positive relation between maladaptive coping strategies and different disorders was found (Endler & Parker, 1990; Holahan et al., 1996), as well as a higher number of stressful life events and a higher use of maladaptive coping strategies before the onset of these events, which showed depressed adolescents (Selffge-Krenke, 2000). Also, besides the fact that significant values were attributed to the covariance, the results found are low and attest in some ways, the independence of the latent constructs that were analyzed.
We conducted other specific CFA analyses to test model invariance in relation to gender and different age groups. Although the results were adjusted to the different sub-samples studied, the AIC index highlights the highest adequacy of the data to the hypothesized model in the sub-sample, which refers to the younger group . We may conclude from this that these scales are more adequate in the evaluation of individuals with these characteristics.
Our results also revealed that girls presented significantly more anxious scores (MASC-R) and tended to develop more coping strategies (CRI-R) than boys. In accordance with our findings, Essau, Conradt and Petermann (2000) found that fiales presented more anxious symptoms than males. Also, Matos et al. (2003) confirmed a significant association between depression and anxiety, and that anxiety and/or depression symptoms were more common in fiales.
According to our results, students from earlier school grades appeared to be significantly more depressed than the ones in the grades following. Adolescents in the younger group also presented significantly more frequent coping strategies use (CRI-R). We should, nevertheless, safeguard the low values associated with the effect size obtained and which raise precautions as to the statistical analysis retrieved. This may represent that the effect it measures is not meaningful or important.
All the instruments used , CDI, MASC and CRI-Y had originally sub-scales that were not here considered, because the purpose was exactly to propose an instrument putting together 3 unidimensional scales, assessing respectively the most relevant factors of depression, anxiety and coping, in this adolescent population. Furthermore, both EFA and CFA supported the unidimensional aspect of each of the 3 reduced dimensions.
It is very interesting to verify in Table 1 that from all the definitions presented in the literature, the depressive itis retained in the PCA and associated to the CDI-R, sei related to “negative thoughts about oneself and to the environment", supporting the idea that this is the most representative factor for depression in this adolescent population. Negative life events are relevant issues (Lau, 2002; Meyer, Chrousos, & Gold, 2001; Sandberg, Rutter, Pickles, McGuinness, & Angold, 2001), and one of their most relevant aspects is the number of negative life events, because they are described as having a cumulative effect (Kaplan, 1999; Masten, Garnezy, Tellegen, Pellegrini, Larkin, Larsen, 1988). This is an important cue to keep in mind in clinical and promotional interventions in school settings – that is, the importance of building positive self-estei in adolescents and promoting an optimistic view of the world, even in the presence of negative life events.
Going further through this content analysis of itis retained after the PCA, it is quite clear that all the anxiety itis included in the MASC-R are related to “physical, iotional and biological related sensations”, supporting the idea that this is the most representative factor for anxiety in this adolescent population. This is also an important issue in clinical and promotional intervention in school settings, namely when considering the importance of helping pupils understand, manage and control their bodies and their iotions. Unfortunately, this kind of curricula is too often forgotten.
As for the coping strategy itis included in the reduced version of the CRI-R, it is also clear that all of the itis are related to “probli analyses and resolution” - that is, to “active coping”, supporting the idea that “ active coping” is the most representative factor for coping in this adolescent population. This feature highlights and reinforces the importance of checking the adaptative/active strategies that adolescents use in order to deal with stress, considering that coping can be a key issue in understanding the success of performances under stressful events (Hussong & Chassin, 2004). Being able to manage one’s own life and becoming autonomous, and simultaneously, using autonomous and positive coping skills successfully, is often related to the feeling of well-being – competencies which should be contiplated when considering school curricula.
Although the sample used in this study does not have national significance, it is a broad sample covering all school grades evenly (from 5th to 12th). Furthermore, classes were randomly chosen and selected from the schools, chosen to assure a broad national distribution. It is also important to underline that the CFA model, which corresponds to the sub-sample of the masculine gender, presented some values close to adjustment, which allows us to suggest a new replica of the scales in future studies.
Besides the limitations found, this study managed to propose valid, reliable and sensitive reduced form enabling to assess in school settings with non-clinical populations, three important features of adolescents conditions: depression, anxiety and active coping, including their most relevant features, which were in the present study, the “negative thoughts about oneself and to the environment”, the "physical, iotional and biological related sensations” and the “probli analyses and resolution”.
Arbuckle, J. L. (2005). AMOS 6.0 User’s Guide. Chicago, IL: SPSS Inc.
Bentler, P.M. (1990) Comparative fit indexes in structural models. Psychological Bulletin, 107(2), 238–246.
Browne, M. W., & Cudeck, R. (1993). Alternative ways of assessing model fit. In K. A. Bollen & J. S. Long (Eds.). Testing Structural Equation Models. Newbury Park, CA: Sage.
Byrne, B. (2000). Relationships between anxiety, fear, self-estei and coping strategies. Adolescence, 26, 316-322.
Byrne, B. (2001). Structural Equation Modeling with AMOS. Mahwah, NJ: Lawrence Erlbaum Associates Publishers.
Duggal, S., Carlson, E., Sroufe, A., & Egeland, B. (2001). Depressive symptomatology in childhood and adolescence. Development and Psychopathology, 13, 143-164.
Endler, N. S., & Parker, J.A. (1990). Multidimensional Assessment of Coping: A critical evaluation. Journal of Personality and Social Psychology, 49, 739-747.
Essau, C. A., Conradt, J., & Petermann, F. (2000). Frequency, Comorbidity, and Psychosocial Impairment of Anxiety Disorders in German Adolescents. Journal of Anxiety Disorders, 14, 263-279.
Fonseca, A. C. (1998). Problias de ansiedade i crianças e adolescentes: Elientos para a sua caracterização e para o seu diagnóstico. Revista Portuguesa de Pedagogia, 18(2), 7-34.
Gore, S., & Eckenrode, J. (1996). Context and process in research on risk and resilience. In R. J. Haggerty, L. R. Sherrod, N. Garmezy & M. Rutter (Eds.), Stress, risk, and resilience in children and adolescents: Processes, mechanisms, and interventions (pp. 19-63). New York: Cambridge University Press.
Ford, J. K., MacCallum, R. C., & Tait, M. (1986). The application of exploratory factor analysis in applied psychology: Critical review and analysis. Personnel Psychologist, 39, 292-314.
Hakstian, A. R., Rogers, W. T., & Cattell, R. B. (1982). The behaviour of number-of-factors rules with simulated data. Journal of Marketing, 62, 30-45.
Harrington, R., Rutter, M., & Fombonne, E. (1996). Development pathways in
depression: Multiple meanings, antecedents, and end-point. Development and Psychopathology, 8, 601-616.
Holahan, C., Moos, R., & Schaefer, S. (1996). Coping, stress and resistence and growth: Conceptualizing adaptative functioning. In M . Zeidner & N. S. Endler (Eds). Handbook of coping: Theory, research, applications (pp. 24-45). New York: Wiley.
Hu, L.T., & Bentler, P. (1995). Evaluating model fit. In R. H. Hoyle (Ed.), Structural Equation Modeling. Concepts, Issues, and Applications (pp. 76-99). London: Sage.
Hussong, A. M., & Chassin, L. (2004). Stress and coping among children of alcoholic parents through the young adult transition. Development and Psychopathology, 16, 985-1006.
Kaplan, H. B. (1999). Toward an understanding of resilience: A critical review of definitions and models. In M. D. Glantz & J. L. Johnson (Eds.), Resilience and development: Positive life adaptations (pp. 17-83). New York: Klumer Acadiic / Plenum Publishers.
Kovacs, M. K. (1981). Children´s Depression Inventory. Toronto: Multi-Health, Inc.
Kovacs, M. K., & Devlin, B. (1998). Internalizing disorders in childood. Journal of Child Psychology and Psychiatry, 19(4), 47-63.
Lau, B. W. (2002). Does the stress in childhood and adolescence matter? A psychological perspective. Journal of the Royal Society of Health, 122(4), 238-244.
March, J. (1997). Multidimensional Anxiety Scale for children. Technical manual. Canada: MHS, Inc.
Mash, E. J., & Wolfe, D. A. (2002). Abnormal Child Psychology (2nd Edition). EUA: Wodsworth Thomson Learning.
Masten, A. S., Garmezy, N., Tellegen, A., Pellegrini, D. S., Larkin, K., & Larsen, A. (1988). Competence and stress in school children: The moderating effects of individual and family qualities. Journal of Child Psychology and Psychiatry, 29(6), 745-764.
Matos, M. G., Barrett, P., Dadds, M., & Shortt, A. (2003). Anxiety, depression, and peer relationships during adolescence: Results from the Portuguese national health behaviour in school-aged children survey. European Journal of Psychology of Education, 18(1), 3-14.
McGee, R.; Feehan, M.; Williams, S. & Anderson, J. (1992). DSM-III disorders from age 11 to 15 years. Journal of the American Acadiy of Child and Adolescent Psychiatry, 31, 50-59.
Meyer, S. E., Chrousos, G. P., & Gold, P. W. (2001). Major depression and the stress systi: A life span perspective. Development and Psychopathology, 13, 565-580.
Moos, R. H. (1993). Coping Responses Inventory. Odessa: Psychological Assessment Resources, Inc.
Nunnally, J.C. (1978). Psychometric theory. New York: McGraw-Hill.
Pine, D. S., Cohen, P., Gurley, D., Brook, J., & Ma, Y (1998). The risk for early-adulthood anxiety and depressive disorders in adolescents with anxiety and depressive disorders. Archives of General Psychiatry, 59, 56-64.
Rosen, J. B., & Schulkin, J. (1998). From normal fear to pathologiacl anxiety. Psychological Review, 105(2), 325-350.
Sandberg, S., Rutter, M., Pickles, A., McGuinness, D., & Angold, A. (2001). Do high-threat life events really provoke the onset of psychiatric disorder in children? Journal of Child Psychology and Psychiatry, 42(4), 523-532.
Sanders, M. (2000). Family intervention and prevention of behaviour disorders in children and adolescents. In M. G. Matos, C. Simões & S. F. Carvalhosa (Eds.), Desenvolvimento de competências de vida na prevenção do desajustamento social (pp. 35-55). Lisboa: IRS /MJ.
Schraedly, P. K., Gotlib, I. H., & Hayward, C. (1999). Gender differences in correlates o depressive symptoms in adolescents. Journal of Adolescent Health, 25, 98-108.
Selffge-Krenke, I. (2000). Causal links between stressful events, coping style and adolescent symptomatology. Journal of Adolescence, 23, 675-691.
Tabachnick, B.G., & Fidell, L.S. (2001). Multivariate statistics (4th ed.). Boston, MA: Allyn & Bacon.
Toumbourou, J. W. (2001). Working with families to promote health adolescent development. Family Matters. Retrieved Noviber 15, 2001, from http://www.findarticles.com
Vigil, JM, & Geary, DC. (2008). A preliminary investigation of family coping styles and psychological well-being among adolescent survivors of Hurricane Katrina. Journal of Family Psychology, 22(1), 176-180.
Wenger, N., Sharrer, V., & Wynd, C. (2000). Stress, coping, and health in children. In Rice, V. (Eds.), Handbook of stress, coping, and health implications for nursingresearch, theory, and practice (pp. 265-289). California: Sage Publications, Inc.
Werner, E. E., & Smith, R. S. (2001). Journeys from childhood to midlife: Risk, resilience and recovery. New York: Cornell University Press.
Margarida Gaspar de Matos
FMH/UTL - Estrada da Costa Cruz Quebrada
1499 Lisboa CODEX - Portugal
Tel +351 21 4149152
Word count: 4943