Behaviour Research and Therapy 51 (2013) 333e337

Contents lists available at SciVerse ScienceDirect

Behaviour Research and Therapy

journal homepage: www.elsevier.com/locate/brat

Shorter communication

Reducing childrens social anxiety symptoms: Exploring a novel parent-administered cognitive bias modication training intervention
Jennifer Y.F. Lau a, *, Eleanor Pettit a, Cathy Creswell b
a b

Department of Experimental Psychology, University of Oxford, South Parks Road, Oxford OX1 3UD, UK School of Psychology and Clinical Language Sciences, University of Reading, UK

a r t i c l e i n f o
Article history: Received 5 January 2013 Received in revised form 19 March 2013 Accepted 19 March 2013 Keywords: Cognitive bias modication Interpretational bias Children Social anxiety Parenting intervention

a b s t r a c t
Social fears and worries in children are common and impairing. Yet, questions have been raised over the efcacy, suitability and accessibility of current frontline treatments. Here, we present data on the effectiveness of a novel parent-administered Cognitive Bias Modication of Interpretations (CBM-I) training tool. CBM-I capitalises on ndings demonstrating an association between anxiety symptoms and biased interpretations, the tendency to interpret ambiguous situations negatively. Through CBM-I training, participants are exposed to benign resolutions, and reinforced for selecting these. In adults and adolescents, CBM-I training is effective at reducing symptoms and mood reactivity. In the present study, we developed a novel, child-appropriate form of CBM-I training, by presenting training materials within bedtime stories, read by a parent to the child across three consecutive evenings. Compared to a test-retest control group (n 17), children receiving CBM-I (n 19) reported greater endorsement of benign interpretations of ambiguous situations post-training (compared to pre-training). These participants (but not the test-retest control group) also showed a signicant reduction in social anxiety symptoms. Pending replication and extensions to a clinical sample, these data may implicate a costeffective, mechanism-driven and developmentally-appropriate resource for targeting social anxiety problems in children. 2013 Elsevier Ltd. All rights reserved.

Social anxiety in children is common (Grant et al., 2005) and disabling - associated with marked underachievement at school and avoidance of age-typical social activities (Erath, Flanagan, & Bierman, 2007; Van Ameringen, Mancini, & Farvolden, 2003). Untreated, early social anxiety can carry risks for adult SAD (Pine, Cohen, Gurley, Brook, & Ma, 1998) and other long-term mental health difculties (Zimmermann et al., 2003), generating signicant personal and economic costs (Greenberg et al., 1999). It is therefore imperative to treat early symptoms of social anxiety. However, questions have been raised over the efcacy, suitability and accessibility of current frontline treatments. More specically, while existing psychological treatments can be effective, a signicant number of paediatric patients with SAD fail to improve (Ginsburg et al., 2011; Hudson, Rapee, Lyneham, Wuthrich, & Schniering, 2010). Concerns have also been raised over the longterm use of pharmacological treatments (Muris, 2012). Both forms of treatment are also difcult to access (long waiting lists) and costly to administer (require therapist time). Innovative, and
Abbreviations: CBM-I, cognitive bias modication of interpretations. * Corresponding author. Tel.: 44 186 527 1375; fax: 44 207 834 3998. E-mail address: jennifer.lau@psy.ox.ac.uk (J.Y.F. Lau). 0005-7967/$ e see front matter 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.brat.2013.03.008

preferably, low-intensity, yet effective treatments for early social anxiety are therefore urgently needed. Recent data implicate Cognitive Bias Modication of Interpretations (CBM-I) training in the reduction of adult anxiety symptoms (Beard, 2011; Hallion & Ruscio, 2011). CBM-I methods draw on longstanding ndings that negative interpretations of ambiguous situations characterise and may even be causally linked to anxiety symptoms (Mathews & MacLeod, 2005). Modifying these biases through training therefore presents a potential strategy for reducing anxiety symptoms. Training involves repeatedly presenting participants with incomplete ambiguous scenarios, which can only be completed satisfactorily by endorsing a benign assessment of the scenario. The idea is that over time, these positive resolutions, which are reinforced by feedback, become automatic, overriding any negative interpretational tendencies, thus reducing anxiety-vulnerability. CBM-I methods have been extended to social anxiety specically (Amir, Bomyea, & Beard, 2010; Beard & Amir, 2008). Targeting interpretations might be particularly important in social situations, because social cues are often ambiguous and thus open to distorted interpretation (Beard & Amir, 2008). Indeed adult data from socially-anxious individuals support these ideas by showing that altering biases can attenuate symptoms.

334

J.Y.F. Lau et al. / Behaviour Research and Therapy 51 (2013) 333e337

Findings have been inconsistent in relation to specic crosssectional associations between social anxiety and social interpretation biases in children (Muris et al., 2000), however there is recent evidence that implicates distorted social interpretation in the development of early social anxiety symptoms (Pass, Arteche, Cooper, Creswell, & Murray, 2012; Vassilopoulos & Banerjee, 2008). Extending CBM-I to children who might be vulnerable to social anxiety could, therefore, yield therapeutic benets for this age range too, Furthermore there are several reasons why developmental extensions of CBM-I might be compelling. First, CBM-I may rely on simple reinforcement learning mechanisms, similar to those by which children and adolescents rst acquire anxious cognitions (Field, 2006). Second, training-facilitated learning during childhood may benet from higher levels of brain plasticity, prior to the more protracted brain maturation that occurs in late childhood and early adolescence (Gogtay et al., 2004). This is consistent with data showing that some cognitive styles become increasingly stable and habitual in the transition to adolescence (Hankin, 2008), raising interesting questions over whether these same cognitions may be more (or less) amenable to change before this developmental juncture. A handful of studies have begun to apply CBM-I training to social anxieties in childhood. Two of the earliest studies targeting childrens fears and worries (including social ones) contextualised training within a computer game that involved a journey through space and encounters with aliens from another planet (Muris, Huijding, Mayer, & Hameetman, 2008; Muris, Huijding, Mayer, Remmerswaal, & Vreden, 2009). Specically, children were confronted with ambiguous social situations that that they were asked to explain through selection of one of two explanations. Children who received positive training were reinforced with positive feedback whenever they chose the benign interpretation rather than the negative interpretation of the ambiguous situation. While this training task was effective in altering childrens perceived levels of threat to ambiguous situations over the course of a single training session, trainingrelated reductions of anxiety symptoms and behavioural avoidance were weak. One possible explanation for this may be the reduced ecological validity, and therefore generalizability of alien encounters to real-life situations. However, building on this training task with items that more resembled real-life social situations, Lester and colleagues again failed to nd changes in avoidant tendencies, despite some changes in threat interpretations (Lester, Field, & Muris, 2010). Another possible explanation for the weak effects on anxiety reduction is that multiple sessions of training may be required, as shown in studies of high socially anxious adults (Beard & Amir, 2008). A fourth study to administer CBM-I to high socially anxious children (aged 10e11 years) implemented a 3-session training intervention (Vassilopoulos, Banerjee, & Prantzalou, 2009). Data showed signicant alterations in interpretational style and social anxiety symptom reduction in the experimental training group compared to a group who received no training, but who simply reported their social anxiety levels three days apart. Multi-session CBM-I may therefore be an important factor for symptom reduction to occur. However, in a more recent attempt to replicate these ndings, no signicant changes in social anxiety were found (Vassilopoulos, Moberly, & Zisimatou, 2012). What other factors could enhance the therapeutic effects of CBM-I? The mode of delivery of interventions may also be crucial. Prior tasks have used self-paced computerised training tasks or ashcards administered by an unknown experimenter. Although efforts were probably made to increase accessibility and engagement of children, these procedures may also be somewhat articial, deviating from the natural processes by which interpretative

styles are rst acquired and shaped in this age range. Studies investigating the development of maladaptive cognitions suggest that children may adopt biases in information-processing through social learning mechanisms, for example, by modelling parental behaviour (see Field, 2006 for a review). If negative interpretative styles in children are adopted through social modelling of parents styles, it may be that benign styles can be more easily trained through CBM-I, if training packages are administered by parents. Thus shaping interpretational styles through parent-child social learning may yield stronger effects. In the present study, we developed a new CBM-I tool which, like the previous training protocols, challenged childrens threat biases but placed these rmly in the context of parent-child interactions. Specically, we asked parents to complete training items as a set of bedtime stories. As most prior data from adults and children seem to indicate that multiple sessions are important for symptom reduction, bedtime stories were read out over three consecutive evenings in a week. Pre and post-training measures of interpretational style and social anxiety symptoms were administered, and compared to a second group of children who did not receive any form of CBM-I training. Given the exploratory nature of this study, data were collected from typically-developing children varying in social fears and worries. Methods Sample and procedures Participants were 36 children aged 7e11 years (mean age: 9 years 4 months, sd: 1 year 4 months; 61% female), recruited from a local schools in Cambridgeshire and Essex to take part in a 1-week long study on childrens positive thinking and social worries. Permission from the school was sought to rst present the study to pupils and then to write to parents of interested pupils about the study. The rst nineteen children and their parents were assigned to the experimental intervention group. An additional 17 children were then recruited to the test re-test control group. Recruitment of the control group took place several weeks later, but assessments took place in the same school term as those for the experimental intervention group. There were no signicant differences in age [t(33) 0.91, p n.s.] or gender composition (c2 0.91, p n.s.) across groups. Although parents from both groups received identical information about the topic of the study (i.e. positive thinking and social worries in children), parents in the experimental intervention were told that they would have to read stories to their child on a daily basis across three evenings, while parents in the testretest group were simply told that their child would complete an assessment three days apart. In the intervention condition, each parent-child dyad attended two visits with the research psychologist. On Day 1, children received pre-training assessments of interpretational style and social anxiety symptoms. Parents were also given standardised instructions over administration of bedtime stories across three consecutive evenings. They then received booklets that contained 45 stories (15 stories per night) for use on Days 2e4. On Day 5, children and their parents returned for the post-training assessment using the same measures as the pre-training session. Parents were asked if they had carried out the exercise according to the instructions, and were then debriefed on the purpose of the study. Children and parents allocated to the control group received phone calls from the researcher, prompting them to complete a set of questionnaires on Day 1 and 5 in the given week. All parents provided informed consent, while children gave written assent. This study was approved by the local university human research ethics committee.

J.Y.F. Lau et al. / Behaviour Research and Therapy 51 (2013) 333e337

335

Materials CBM-I training task (administered Days 2e4) Forty ve stories translated from Vassilopoulos et al. (2009) were read by parents to children across 3 consecutive evenings (15 items per evening). Each story described an ambiguous social situation. For example: You have invited one of your schoolmates to come and play at your house. The phone rings and he tells you that he will not come over. At the end of each situation, children were asked why the situation happened, and were provided with both a benign and a negative account. In this example, the benign account was: Something has happened that made him change his plans, perhaps something to do with his parents. while the negative account was: He doesnt like me anymore and he would rather hang out with other children. After the child had indicated their preferred explanation they were given correct or wrong feedback by the parent. The benign solution was always the correct answer. Parents were then instructed to ask the children to take a moment to think about the correct explanation, and then to read out another sentence that paraphrased and reinforced the positive interpretation. Again in this example, the sentence was My schoolmate would have liked to play with me if he had been free. Each story was presented on a single page with an accompanying graphic to increase engagement. The answer was displayed on the following page, and could only be read upon turning the page over. Ambiguous situations interpretation (ASI) scale (administered Days 1 and 5) This was a translated version of the measure developed by Vassilopoulos and Banerjee (2008) The ASI contains 16 questions. In the current study, half were administered at pre-test and half at post-test. Each question describes an ambiguous social situation, to which two alternative interpretations are presented, one which represents a negative interpretation and the other a benign interpretation. For example: During maths class, the teacher asks children to sit in pairs to solve an exercise. However, he makes you sit by yourself. Why do you think this happened to you? ANSWERS: There were no other children who could have sat with me OR Nobody wants to sit with me. Each interpretation is rated by the child on a Likert scale of 1e5 where 1 I would not think that at all and 5 I would immediately think that. The order of presentation of benign and negative interpretations was randomised. The measure produces two scales, one for endorsement of benign interpretations, and one for negative interpretations. Cronbachs alpha (across both groups) gave, at pre-test: a 0.80 and a 0.67 for negative (8 items) and benign (8 items) subscales respectively, and at post-test: a 0.75 and a 0.79 for negative and benign items. Social Anxiety Scale for Children e Revised (SASC-R) (administered Days 1 and 5) This scale contains twenty two items describing common social worries and avoidance of common social situations e.g. I worry that other kids dont like me or Im afraid to invite others to my house because Im afraid they might say no. These are rated on a scale from 1 to 5, reecting the frequency with which they are experienced. A composite score can be computed by summing all items. Internal consistency of the scale in the present study (across both the experimental intervention and control groups) using Cronbachs alpha was 0.92 and 0.90 for the pre- and post-test assessments. Prior studies report negative associations between this scale and self-reported social acceptance and global self-worth; and positive associations with peer rejection in this age range (G. S. Ginsburg, La Greca, & Silverman, 1998; Reijntjes, Dekovic, & Telch, 2007).

Data analyses We rst assessed whether benign and negative interpretations (i.e. ASI sub-scales) correlated with social anxiety symptoms (i.e. SASC-R scores) at pre-test and post-test. Next, we assessed the effects of the intervention on changes in interpretational style using a 2 (condition: intervention, test-retest) 2 (time: pre-test, posttest) 2 (interpretation-item: benign, negative) mixed ANOVA on endorsement ratings. To assess the effects of the intervention on changes in social anxiety symptoms (i.e. SASC-R scores), we conducted a 2 (condition: intervention, test-retest) 2 (time: pre-test, post-test) mixed ANOVA on symptom scores. Finally, we assessed whether changes in benign and negative interpretations (from preto post-test) correlated with changes in social anxiety symptoms (from pre- to post-test) in the whole sample or in the experimental intervention group alone. Results Correlations between negative and benign interpretations and social anxiety symptoms At pre-test, social anxiety symptoms correlated positively with endorsement ratings of negative interpretations (Pearsons r 0.51, p < 0.01) but negatively with those for benign interpretations (Pearsons r 0.50, p < 0.01). At post-test, these correlations remained signicant (rs 0.38 and 0.36, ps < 0.05 for ratings of negative and benign interpretations respectively). Assessing the effects of the intervention on changes in interpretation bias Analyses of endorsement ratings (Fig. 1) yielded a signicant 3way interaction (F(1, 32) 6.05, p < 0.05), in addition to main effects of time, interpretation-item and their interaction (all

Fig. 1. Changes in the endorsement of benign and negative interpretations of ambiguous situations in the experimental intervention group and test-retest control group, from the pre- to post-test session.

336

J.Y.F. Lau et al. / Behaviour Research and Therapy 51 (2013) 333e337

Fs > 8.59, p < 0.01). This prompted us to explore the effects of intervention versus no intervention on changes in endorsement of negative and benign statements separately. For benign interpretations, there was a signicant main effect of time (F(1, 34) 39.61, p < 0.001) but this was further modied by condition (F(1, 34) 8.18, p < 0.01). Both groups of participants showed an increase in endorsement of benign interpretations, but this was more marked in the intervention group (t(18) 6.56, p < 0.001; Cohens d 1.78) than in the test-retest group (t(16) 2.36, p < 0.05; Cohens d 0.56). Moreover, while there was no betweengroup difference in endorsement of benign interpretations before training (t(34) 0.84, p n.s.), after training, those who had received the intervention reported greater endorsement of benign items, than those who had received no intervention (t(34) 2.41, p < 0.05; Cohens d 0.80). For negative interpretations, there was a signicant main effect of time (F(1, 34) 9.33, p < 0.01) but the interaction with condition did not reach signicance (F(1, 34) 2.72, p 0.11), indicating that across all participants, there was a reduction of the endorsement of these interpretations. Given a priori hypotheses, we explored within-group changes in negative interpretations, nding that only those in the intervention group showed reduced endorsement of these interpretations (t(18) 3.37, p < 0.05; Cohens d 0.94), with no changes in the test-retest condition (t(16) 0.99, p n.s.). Assessing the effects of the intervention on changes in social anxiety symptoms Analyses of social anxiety symptom scores (Fig. 2) yielded a signicant main effect of time (F(1, 34) 14.60, p < 0.001) that was modied by a signicant condition-by-time interaction (F(1, 34) 5.98, p < 0.05). Decomposing this interaction, a main effect of time characterised the intervention group (t(18) 3.69, p < 0.05; Cohens 0.53) but not the test-retest control group (t(16) 1.49, p n.s.).

Correlations between changes in interpretational style and social anxiety symptoms Changes in endorsement ratings for negative and benign interpretations did not correlate with changes in social anxiety symptoms in the whole sample (ps < 0.62) or in the experimental intervention group alone (ps < 0.79). Discussion In the present study, we explored the effectiveness of a novel procedure for training benign interpretative styles in children, and its capacity to reduce social anxiety symptoms. Unlike most other cognitive bias modication techniques which have used computerised interfaces to deliver training, here, the parents became the mode of training delivery. Thus, we capitalised on the known role of parents in the acquisition of childrens negative cognitions. These early data are supportive. First, they suggested that the intervention was successful in altering interpretative styles: although all children tended to show signicant increases in the selection of benign interpretations at a second assessment, this increase was greater in children who had received three sessions of experimental intervention (compared to those who had received no intervention). This was also somewhat true in the tendency to select negative interpretations, where post-hoc tests revealed that only those receiving the experimental intervention showed signicant reductions on this measure. Second e and perhaps more crucially e our data showed that the experimental intervention could signicantly reduce social anxiety symptoms across the course of the week, a change that did not characterise children in the comparison test retest group. While these data are encouraging, some questions remain. Most notably, why did changes in interpretational style measures not correlate with reductions in social anxiety symptoms in the experimental group? Several possibilities exist. First, our sample comprised unselected children from the community with relatively low social anxiety scores (42.97 at baseline, compared to 47 in a clinical sample of children with SAD (Epkins, 2002)). Because of potential oor effects, lower scores could inuence the extent to which symptoms change as a function of the intervention. Indeed, mean change in the present sample was 3.89, a fairly modest reduction. The size of correlations may have been reduced by a similarly restricted range of change on the bias measures (0.70 and 0.38 for benign and negative interpretations respectively). To address this, future studies should explore the effectiveness of this same intervention in children with elevated levels of social anxiety. Indeed, previous studies with adults and children are suggestive of a moderating effect of anxiety, where individuals with higher anxiety scores show greater changes associated with CBM-I interventions. In such samples, greater variability in change scores may yield the expected correlations between symptom and bias changes. With these considerations in mind, it is noteworthy that signicant experimental effects were found in this preliminary study. A second possibility is that the present parent-administered CBM-I task does not change symptoms through changes in interpretational style. It is difcult to disambiguate whether reductions in social anxiety in the present study emerged from placebo effects, from simple exposure to social situations or from intrinsically rewarding experiences of regular parent-child interactions. Placebo effects may have been present given that there were differences in information given to the experimental intervention group (i.e. that parents would have to administer a bedtime story exercise for three nights) that were absent in instructions for the test-retest group, which could result in differential expectations of symptom-change in the two groups. In addition, it is possible that face-to-face

Fig. 2. Changes in social anxiety symptoms in the experimental intervention group and test-retest control group, from the pre- to post-test session.

J.Y.F. Lau et al. / Behaviour Research and Therapy 51 (2013) 333e337

337

interactions with research psychologists as part of a research study may have contributed to demand characteristics in the experimental intervention group that was again absent in the test-retest group. Future studies could disentangle these possibilities by the inclusion of different comparison conditions, such as instructing parents to read stories about social situations without clear emotionally-valenced outcome, or to teach their children other skill sets. Given the similarity between our test of interpretation bias to the training task, in terms of format, our results could arise from demand effects. This possibility calls for the use of interpretation bias measures that are procedurally different to preclude possible demand effects and to explore generalizability of training to new measures. More covert methods for assessing interpretational style, including the use of ecologically-valid, in vivo measures could be useful. Notwithstanding the ambiguity in mechanisms by which our tool works, it is still notable in its capacity to reduce symptoms (beyond natural regressions to the norm), setting it apart from other packages used in children to date, where the most consistent nding is that changes in interpretational style occur in the absence of parallel mood changes. One factor may lie in the length of training. Indeed, whereas single session CBM-I can yield alterations in biases in most adult and child studies, mood improvements usually emerge in multi-session training (although this is not always true e see Vassilopoulos et al., 2012). Another factor that may enhance the effectiveness of our training tool is that it capitalises on data suggesting that children may acquire cognitive styles through social learning mechanisms (Field, 2006). Involving parents in bias modication training thus mimics natural processes by which children come to learn particular interpretational styles, and may therefore reect a more powerful strategy than learning through computerised interface. Speculating on the potential of this new training tool one step further, it would be interesting to assess the extent to which this package could also challenge and modify the parents own interpretative style, particularly in relation to events faced by their children. Using a common tool to target two sources of vulnerability (the parent and childs) may be a particularly fruitful area of intervention. Our study has shown the potential of a parentally administered interpretation bias modication programme for modifying biases associated with social anxiety, and social anxiety itself in children. It is important that future studies explore whether these benets are not simply the result of a structured parent-child activity, but instead, through the modication of biases. Much more work is clearly needed to establish clinical potential including its capacity to engage children and parents. However if effective, it could prove to be a valuable, cost-effective and developmentallyappropriate resource for preventing the development of social anxiety problems in children. References
Amir, N., Bomyea, J., & Beard, C. (2010). The effect of single-session interpretation modication on attention bias in socially anxious individuals [Comparative Study Randomized Controlled Trial Research Support, N.I.H., Extramural]. Journal of Anxiety Disorders, 24(2), 178e182. http://dx.doi.org/10.1016/ j.janxdis.2009.10.005. Beard, C. (2011). Cognitive bias modication for anxiety: current evidence and future directions. Expert Review of Neurotherapeutics, 11(2), 299e311. Beard, C., & Amir, N. (2008). A multi-session interpretation modication program: changes in interpretation and social anxiety symptoms. Behaviour Research and Therapy, 46(10), 1135e1141. Epkins, C. C. (2002). A comparison of two self-report measures of childrens social anxiety in clinic and community samples [Comparative Study]. Journal of Clinical Child & Adolescent Psychology, 31(1), 69e79. http://dx.doi.org/10.1207/ S15374424JCCP3101_09.

Erath, S. A., Flanagan, K. S., & Bierman, K. L. (2007). Social anxiety and peer relations in early adolescence: behavioral and cognitive factors. Journal of Abnormal Child Psychology, 35(3), 405e416. Field, A. P. (2006). Is conditioning a useful framework for understanding the development and treatment of phobias? [Review]. Clinical Psychology Review, 26(7), 857e875. http://dx.doi.org/10.1016/j.cpr.2005.05.010. Ginsburg, G. S., Kendall, P. C., Sakolsky, D., Compton, S. N., Piacentini, J., Albano, A. M., et al. (2011). Remission after acute treatment in children and adolescents with anxiety disorders: ndings from the CAMS [Peer Reviewed]. Journal of Consulting and Clinical Psychology, 79(6). http://dx.doi.org/10.1037/ a0025933 22122292. Ginsburg, G. S., La Greca, A. M., & Silverman, W. K. (1998). Social anxiety in children with anxiety disorders: relation with social and emotional functioning [Research Support, U.S. Govt, P.H.S.]. Journal of Abnormal Child Psychology, 26(3), 175e185. Gogtay, N., Giedd, J. N., Lusk, L., Hayashi, K. M., Greenstein, D., Vaituzis, A. C., et al. (2004). Dynamic mapping of human cortical development during childhood through early adulthood. Proceedings of the National Academy of Sciences of the United States of America, 101(21), 8174e8179. http://dx.doi.org/10.1073/ pnas.04026801010402680101 [pii]. Grant, B. F., Hasin, D. S., Blanco, C., Stinson, F. S., Chou, S. P., Goldstein, R. B., et al. (2005). The epidemiology of social anxiety disorder in the United States: results from the National Epidemiologic Survey on Alcohol and Related Conditions. Journal of Clinical Psychiatry, 66(11), 1351e1361. Greenberg, P. E., Sisitsky, T., Kessler, R. C., Finkelstein, S. N., Berndt, E. R., Davidson, J. R., et al. (1999). The economic burden of anxiety disorders in the 1990s. Journal of Clinical Psychiatry, 60(7), 427e435. Hallion, L. S., & Ruscio, A. M. (2011). A meta-analysis of the effect of cognitive bias modication on anxiety and depression [Meta-Analysis]. Psychological Bulletin, 137(6), 940e958. http://dx.doi.org/10.1037/a0024355. Hankin, B. L. (2008). Stability of cognitive vulnerabilities to depression: a shortterm prospective multiwave study [Peer Reviewed]. Journal of Abnormal Psychology, 117(2). http://dx.doi.org/10.1037/0021e843X.117.2.324 18489208. Hudson, J. L., Rapee, R. M., Lyneham, H., Wuthrich, V., Schniering, C. A. (2010, June). Treatment outcome for children with social phobia. Paper presented at the World Congress of Behavioural and Cognitive Therapies, Boston, MA. Lester, K. J., Field, A. P., & Muris, P. (2010). Experimental modication of interpretation bias regarding social and animal fear in children. Journal of Anxiety Disorders, 25(5), 697e705. Mathews, A., & MacLeod, C. (2005). Cognitive vulnerability to emotional disorders. Annual Review of Clinical Psychology, 1, 167e195. Muris, P. (2012). Treatment of childhood anxiety disorders: what is the place for antidepressants? Expert Opinion on Pharmacotherapy, 13(1), 43e64 http:// dx.doi.org/10.1517/14656566.2012.642864. Muris, P., Huijding, J., Mayer, B., & Hameetman, M. (2008). A space odyssey: experimental manipulation of threat perception and anxiety-related interpretation bias in children. Child Psychiatry & Human Development, 39(4), 469e480. Muris, P., Huijding, J., Mayer, B., Remmerswaal, D., & Vreden, S. (2009). Ground control to Major Tom: experimental manipulation of anxiety-related interpretation bias by means of the space odyssey paradigm and effects on avoidance tendencies in children. Journal of Anxiety Disorders, 23(3), 333e340. Muris, P., Kindt, M., Bogels, S., Merckelbach, H., Gadet, B., & Moulaert, V. (2000). Anxiety and threat perception abnormalities in normal children. Journal of Psychopathology and Behavioral Assessment, 22(2), 183e199. Pass, L., Arteche, A., Cooper, P., Creswell, C., & Murray, L. (2012). Doll Play narratives about starting school in children of socially anxious mothers, and their relation to subsequent child school-based anxiety [Peer Reviewed]. Journal of Abnormal Child Psychology, 40(8). http://dx.doi.org/10.1007/s10802-012-9645-4 22588362. Pine, D. S., Cohen, P., Gurley, D., Brook, J., & Ma, Y. (1998). The risk for earlyadulthood anxiety and depressive disorders in adolescents with anxiety and depression disorders. Archives of General Psychiatry, 55, 56e64. Reijntjes, A., Dekovic, M., & Telch, M. J. (2007). Support for the predictive validity of the SASC-R: linkages with reactions to an in vivo peer evaluation manipulation. Journal of Anxiety Disorders, 21(7), 903e917. Van Ameringen, M., Mancini, C., & Farvolden, P. (2003). The impact of anxiety disorders on educational achievement. Journal of Anxiety Disorders, 17(5), 561e 571. doi: S0887618502002281 [pii]. Vassilopoulos, S. P., & Banerjee, R. (2008). Interpretations and judgments regarding positive and negative social scenarios in childhood social anxiety [Peer Reviewed]. Behaviour Research and Therapy, 46(7). http://dx.doi.org/10.1016/ j.brat.2008.03.008 18462705. Vassilopoulos, S. P., Banerjee, R., & Prantzalou, C. (2009). Experimental modication of interpretation bias in socially anxious children: changes in interpretation, anticipated interpersonal anxiety, and social anxiety symptoms. Behaviour Research and Therapy, 47(12), 1085e1089. Vassilopoulos, S. P., Moberly, N. J., & Zisimatou, G. (2012). Experimentally modifying interpretations for positive and negative social scenarios in children: a preliminary investigation. Behavioural and Cognitive Psychotherapy, 1e14. http:// dx.doi.org/10.1017/S1352465812000537. Zimmermann, P., Wittchen, H. U., Hoer, M., Pster, H., Kessler, R. C., & Lieb, R. (2003). Primary anxiety disorders and the development of subsequent alcohol use disorders: a 4-year community study of adolescents and young adults. Psychological Medicine, 33(7), 1211e1222.