- Email: [email protected]
The performance of clinically depressed children and adolescents on the modified Stroop paradigm
~ Pergamon
Person. individ. Diff. Vol. 23, No. 5, pp. 753 759, 1997 © 1997 ElsevierScienceLtd. All rights reserved Printed in Great Britain
PIh S0191-8869(97)00097-4
0191-8869/97 $17.00+0.00
THE P E R F O R M A N C E OF C L I N I C A L L Y DEPRESSED C H I L D R E N A N D ADOLESCENTS ON THE M O D I F I E D STROOP P A R A D I G M Hamid T. Neshat Doost, 1. Mohammad R. Taghavi, ~Ali R. Moradi, ~William Yule 1 and Tim Dalgleish2 ~Department of Psychology, Institute of Psychiatry, De Crespigny Park, Denmark Hill, London, England and 2MRC Applied Psychology Unit, Cambridge, England
(Received 25 February 1997) Summary--The study investigated selective processing of emotional information in childhood depression using a computerised version of the modified Stroop colour naming task. Three groups of children and adolescents--clinicallydepressed (n = 19), patients with mixed depression and anxiety (n = 19), and normal controls (n = 26)---were required to name the colours of depression-related, threat-related, trauma-related, happy, and categorised neutral words. Normal subjects, compared with the patient groups, exhibited faster colour naming overall. There were no significant differences on the colour naming of different categories of words and the performance across the three groups was not significantly different for different categories of words. O 1997 Elsevier Science Ltd
Keywords: childhood depression, information processing, cognitive biases, Stroop task
There is now convincing evidence that children and adolescents can and do suffer from depressive disorders which are similar to those of adults in many aspects (e.g. Rutter, 1987). Several research studies have used self-report measures (questionnaires) to investigate the cognitive characteristics of depression in children and adolescents. Results of these studies have shown that depressed children and adolescents display the same kind of cognitive characteristics as those observed in depressed adults, including depressive attributional style, low self esteem, cognitive biases, and less positive self schemata (e.g. Seligrnan, Peterson, Kaslow, Tanenbaum, Alloy & Abramson, 1984). Recently researchers have started to use experimental approaches to investigate cognitive biases of emotional disorders in children and adolescents. One of the most frequently used of these tasks is the modified Stroop paradigm. This task requires subjects to name ink colours in which emotional and non-emotional words are written. The emotional Stroop effect refers to the finding that colournaming times are slower for emotional words than for non-emotional words. This task has been used in several studies with non-clinically and clinically anxious and depressed adults. There is considerable evidence that anxious subjects, but not normal controls, have longer reaction times on colour naming of threatening words in the modified Stroop task (Williams, Mathews & Macleod, 1996). These effects seem much clearer than in the case of depression. In the case of depressed subjects, researchers have used different methods of presentation (card, tachistoscope, and computer), different groups of words (general negative vs. negative self-descriptive adjectives), and they have studied different groups of subjects (clinical, subclinical, mood induced, mixed depressed/anxious). These studies have led to different results. Some have found more Stroop interference for negative or depressed words relative to neutral or positive words (Gotlib & Cane, 1987; Gotlib & MacCann, 1984; Kinderman, 1994; Klieger & Cordner, 1990) while others have not (Hill & Knowles, 1991; Mogg, Bradley, Williams & Mathews, 1993). There are a few studies using the modified Stroop task in children and adolescents. Schneider, Unnewher, Florin, and Margraf (1992) found that children of phobic patients exhibited higher
*To whom all correspondence should be addressed at present address: Department of Psychology, University of Isfahan, Chaharbag Bala, Hezarjerib, Isfahan, Iran. 753
754
Hamid T. Neshat Doost et al.
interference scores to phobia-relevant words compared with children of patients with panic disorder and children of parents without any psychiatric morbidity. Martin, Horder, and Jones (1992) found that spider-phobic children were significantly slower to colour-name spider words than control words. Taghavi (19~6) studied the performance of children and adolescents with Generalised Anxiety Disorder on a modified Stroop task. Results of this study showed that, compared to normal controls, children and adolescents with Generalised Anxiety Disorder showed longer colour-naming times for threat-related and trauma-related words, with longer colour-naming times for threat-related words than trauma-related words. At the time of writing there is no published study on the performance of depressed children and adolescents on the modified Stroop task and the purpose of the present study, therefore, is to investigate the modified Stroop effect in clinically depressed children and adolescents. A third group of children and adolescents with mixed depression/anxiety was used to investigate content-specificity effects with depression-related and anxiety-related material.
METHODOLOGY Subjects
There were three groups of subjects: depressed, mixed depressed/anxious, and normal controls. All subjects were aged between 9 and 18 yr and their primary language was English. The age range was 10.92-18.33 yr for the depressed group, 9.50-18.00 yr for the mixed depressed/anxious group, and 13.25-18.17yr for the control group. The patient groups were recruited from a variety of sources including hospital outpatients' and inpatients' services in the UK (Maudsley Hospital, London; Leigh House, Eastleigh; Thelma Golding Centre, London; Bethlem Hospital, London; Brixton Child Guidance Clinic, London; Bloomfield Centre, London; St. Thomas' Hospital, London, and Camberwell Child Guidance Clinic, London). There were 19 subjects (11 girls and eight boys) in the depressed group; the selection criteria were as follows. (a) A primary diagnosis of Major Depressive Disorder according to DSM-IV (Diagnostic and Statistical Manual of Mental Disorders, 4th edition, American Psychiatric Association, 1994) criteria or ICD-10 (International Classification of Diseases, World Health Organization, 10th revision, WHO, 1992) criteria for Depressive Disorder. The diagnosis was determined in a clinical assessment by psychiatrists and psychologists before the patients were introduced to the experimenter. (b) Scores of higher than cut-off on all three depression scales: the Children's Depression Inventory (CDI; Kovacs, 1982), the Depression Self-Rating Scale (DSRS; Birleson, 1981) and the Mood and Feeling Questionnaire (MFQ; Angold, Costello, Pickles & Winder, 1987). There were 19 subjects in the mixed depressed/ anxious group (12 girls and seven boys). The selection criteria were diagnoses of both Depressive Disorder and Anxiety Disorder according to DSM-IV, or a primary diagnosis of Mixed Anxiety and Depressive Disorder according to ICD-10 criteria. The control group consisted of 26 subjects (16 girls and 10 boys), with no known history of emotional disorder. Subjects with vision problems, colour blindness, or brain damage were excluded. The patients were generally asked to participate in the study before the commencement of any medication regime. The results of the questionnaires showed that the depressed children and adolescents were very depressed at the time of the study. In order to control for the effects of confounding variables, the groups were matched for sex, age, vocabulary level, and reading ability. Measures
1. 2. 3. 4. 5.
The Depression Self-Rating Scale (DSRS; Birleson, 1981); The Children's Depression Inventory (CDI; Kovacs, 1982); The Mood and Feeling Questionnaire (MFQ; Angold et al., 1987); The Revised Children's Manifest Anxiety Scale (RCMAS; Reynolds & Richmond, 1978); The British Picture Vocabulary Scale (BPVS Short Form, Dunn, Dunn, Whetton & Pintilie, 1982); 6. The Wechsler Objective Reading Dimensions (WORD; Basic Reading; Rust, Golombok & Trickey, 19933.
Childhood depression and the modified Stroop test
755
Table 1. Means and standard deviations (SD) of age, reading score, vocabulary score, and depression and anxiety measures for the three groups of subjects Depressed
Age (yr) WORD BPVS DSRS CDI MFQ RCMAS
Mixed
Non-patients
Mean
SD
Mean
SD
Mean
SD
15.57 101.84 96.16 23.35 30.60 44.50 21.15
1.58 15.46 10.66 5.83 8.46 12.78 4.01
14.84 104.10 100.05 18.84 23.42 36.47 18.63
2.48 13.53 13.48 7.26 12.00 16.72 6.05
15.15 98.50 91.42 7.50 6.80 13.57 8.11
1.44 14.59 18.53 3.90 6.40 9.06 5.68
WORD, Wechsler Objective Reading Dimensions (Basic Reading); BPVS, British Picture Vocabulary Scale; DSRS, Depression Self-Rating Scale; CDI, Children's Depression Inventory; MFQ, Mood and Feeling Questionnaire; RCMAS, Revised Children's Manifest Anxiety Scale.
Instrumentation An IBM (Thinkpad 755C TF1) portable computer with a 26.41 cm active-matrix T F T colour LCD screen was used to present the stimuli. A voice key was used to register the vocal response of the subjects automatically. The voice key connected to an adjustable head microphone (Bandridge EH620, Dynamic Headsets) worn by the subject. Five buttons (one for each colour: red, green, blue and yellow, and another button for changing the word on the screen when it did not disappear automatically following the subject's vocal response) were also used.
Materials One hundred and twenty words were used as verbal stimuli. These consisted of 12 words from each of five word categories: happy (e.g. pleased), categorised neutral (animals; e.g. sheep), depression-related (e.g. helpless), threat-related (e.g. terrified), and trauma-related (e.g. casualty). Each word was presented once in each of two colours chosen at random from the four colours: green, yellow, red, blue. The words were selected from a source of words produced by 231 primary and secondary school pupils suitable for the age range of subjects in the present study (Neshat Doost, Moradi, Taghavi, Yule & Dalgleish, submitted). All categories were matched for length and frequency. The words were presented in capital letters 8 mm high. The visual angle subtended was approximately 1°
Procedure The test was carried out individually. Each subject was first asked to complete the reading and vocabulary tests. Then s/he was asked to sit in front of the computer, 50 cm from the screen, in a quiet room, without any disruptions. Before the main task, there were some practice trials, which were the same as the experimental trials except that the stimulus words were all uncategorised neutral words. In the main task, 120 stimulus words were presented in a new fully random order for each subject. The presentation time for each word was 1.7 sec. The inter-trial interval was 2.0 sec. The subject was asked to ignore the word and say out loud the colour name into the microphone as quickly as possible. After 40 trials subjects were given a short rest. The task took about 20 min in total. Following the Stroop task the subjects were asked to fill in the depression and anxiety scales. RESULTS
Subject characteristics Means and standard deviations were calculated separately for age, reading, vocabulary, and depression and anxiety scales for the two patient groups and the control group (see Table 1). One-way analyses of variance were used to compare the three groups of subjects on age, reading, vocabulary, depression, and anxiety. There were no significant differences between the three groups for age ( F < 1), vocabulary (F(2,61)= 1.758 NS) and reading ability ( F < 1), but the groups scored significantly differently on the measures of depression (DSRS: F(2,63)=49.036, P < 0.0001; CDI:
756
Hamid T. Neshat Doost et al. Table 2. Mean reaction times and standard deviations (SD) in seconds per word for the five groups of words in each of three groups of subjects Group
Happy
Neutral
Depression
Threat
Trauma
Depressed
Mean SD
0.83 0.16
0.83 0.14
0.84 0.14
0.85 0.15
0.84 0.16
Mixed
Mean SD
0.81 0.18
0.83 0.18
0.83 0.18
0.82 0.17
0.84 0.20
Control
Mean SD
0.69 0.14
0.69 0.13
0,69 0.15
0.69 0.14
0.69 0.14
F(2,63)=44.181, P<0.0001; MFQ: F(2,63)=35.227, P<0.0001) and RCMAS (F(2,63)=39.510, P<0.0001) as expected. To clarify differences between the three groups, post hoc multiple comparison tests (Student-Newman-Keuls; P < 0 . 0 5 ) were used. On the anxiety scale (RCMAS) the depressed group, and the group which was diagnosed as mixed depressed/anxious had significantly higher scores than the control group (P<0.05). Although the depressed group showed a higher anxiety score than the mixed depression/anxiety group, this difference was not significant. On the depression scales, the depressed group and the mixed depressed/anxious group had significantly higher scores than the control group ( P < 0.05). Also, the depressed group had significantly higher scores on the DSRS, and the CDI than the mixed depressed/anxious group (P<0.05) but not on the MFQ. Colour-naming reaction times
The mean ( _ SD) number of colour naming errors was 2.07 + 2.26 across all subjects. The mean numbers of colour naming errors for the depressed, mixed depressed/anxious and control groups were 3.58_+ 3.22, 2.05_+ 1.81 and 1.27_+ 1.12, respectively. A one-way analysis of variance showed that there were significant differences between the three groups on the number of colour naming errors (F(2,63) = 6.509, P < 0.01). To clarify differences between the three groups, Student-NewmanKeuls tests were again used. The results showed that the depressed group made significantly more colour naming errors than the control group (P < 0.05). There were no significant differences between the depressed group and the mixed depressed/anxious group or between the mixed group and the control group. Colour naming latency data were excluded from the analyses on those trials on which subjects made colour naming errors. Also, to minimise the influence of outlying data points, colour naming latency data were excluded from the analyses on those trials when the response time was less than 100 msec or more than 2000 msec. Response times less than 100 msec are impossibly fast and suggest that subjects had vocalised prematurely. Response times of more than 2000 msec could be due to lapses of attention or the computer failing to register a response. The mean colour naming latencies taken by the three groups for threat-related words, depression-related words, happy words, neutral words and trauma-related words are shown in Table 2. A repeated measures analysis of variance was carried out with Group (3: depressed, mixed depressed/anxious, and control) as a between-subjects variable and Word Type (5: depressionrelated, threat-related, trauma-related, happy, and categorised neutral) as a within-subjects variable. The results showed a significant main effect of Group (F(2,61)= 6.76, P < 0.05). The effect of Word Type was not significant ( F < 1) and no significant interaction was found among Group and Word Type ( F < 1). To examine the main effect of Group, Student-Newman-Keuls tests were used. The results showed that, compared with the control group, the depressed group and the mixed depressed/anxious group had significantly longer colour-naming times (P < 0.05). Correlational analyses
Correlation coefficients (Pearson) were calculated between anxiety and depression scores and interference scores from the four categories of words: depression-related, threat-related, traumarelated, and happy. Interference scores were calculated by subtracting the mean latency for categorised neutral words (animal names) from that for the relevant word type. Results showed that the correlations between anxiety and depression scores and mean interference scores of the four categories of words were not significant.
Childhood depression and the modified Stroop test
757
DISCUSSION The purpose of this study was to compare the performance of three groups of children and adolescents (depressed, mixed depressed/anxious and normal controls) on their performance on the modified Stroop task. The results showed that, compared with the control group, patient groups (depressed and mixed depressed/anxious) were in general significantly slower on Stroop colour-naming. This finding replicated the findings of other studies with adults which have found similar main effects (Hill & Knowles, 1991; Kinderman, 1994; Williams & Broadbent, 1986). This retarded colour naming could be due to the effects of psychomotor retardation which is one of the general symptoms of depression. According to Kinderman (1994), this slowness in colour naming may be interpreted as a general effect of psychiatric disturbance. The clinical groups were not significantly slower on the depression-related, threatening, and trauma-related words. These results are inconsistent with Beck's (Beck, 1976) and Bower's (Bower, 1981) models of cognitive processing of emotion which predict that emotion-congruent biases operate in all stages of processing in depression and anxiety, with these biases differing only in terms of content-specificity. In other words, anxious individuals selectively process anxiety-related information, whereas depressed individuals selectively process depression-relevant material. However, this finding can be taken as providing some support for the model proposed by Williams, Watts, MacLeod, and Mathews (1988), which proposes that "depression is less associated with biased pre-attentive priming and more with bias in the active strategic aspect of encoding" (p. 178). The results of the present study with young subjects are therefore consistent with the results of some of the studies on depressed adults which showed that depressed subjects did not display interference effects for negative information on the modified Stroop task (Hill & Knowles, 1991; Mogg et al., 1993). The present study with children and adolescents shows a different profile of results to previous studies which have shown evidence of interference effects of negative information in depressed adult subjects (Gotlib & Cane, 1987; Gotlib & MacCann, 1984; Kinderman, 1994; Klieger & Cordner, 1990). However, as well as different subject populations there were a number of methodological differences between these latter studies and the present one. In some of the adult studies anxiety was not assessed, even though there is often comorbidity between anxiety and depression. Therefore, the Stroop effect could possibly have been due to anxiety rather than depression. Further support for this hypothesis is presented by Mathews and MacLeod (1985) who also showed that anxious patients who showed disruption on an emotional Stroop task were also more depressed than control subjects. However, partial correlational analysis showed that it was anxiety and not depression which was associated with longer colour naming of negative words. Furthermore, some of the adult findings were obtained with a modified version of the Stroop task in which each category of stimulus words (e.g. depression-related words) was presented in blocked format on a card. Accordingly, it is likely that the blocked presentation increased elaborative processing of the semantic content of the negative words and led to longer colour naming times for these words. In the present study, the words were presented individually in random order. Therefore, the possibility of elaborative processing was reduced. In the present study, child and adolescent patients had very high scores on an anxiety scale and in the mixed group they met criteria for an anxiety disorder; however, they did not show a Stroop bias for negative information despite their typically high level of anxiety. Results of several studies of adults have shown that, compared with normal controls, anxious patients and subclinically anxious individuals have significantly more interference for negative words (especially for threatrelated words) than for neutral words (e.g. Dalgleish, 1995). Furthermore, results of studies on anxious children and adolescents (e.g. Taghavi, 1996) have shown that anxious children and adolescents showed longer colour-naming times for threat-related words related to their emotional disorders. Taken together, it can be concluded that the presence of a high level of anxiety is not sufficient to elicit a preattentive bias for negative information. Then, the question of importance becomes why depressed subjects do not show a preattentive bias for negative information despite their high level of anxiety? This is particularly pertinent because in the Williams et al. (1988) model referred to above, anxiety and depression operate at different loci. Consequently, high levels of anxiety should affect processing regardless of the co-existence of depression.
758
Hamid T. Neshat Doost et al.
O n e possible answer is the difference between anxiety a n d depression. A n x i e t y is a state in which the i n d i v i d u a l is m o t i v a t e d to detect p o t e n t i a l l y aversive stimuli in o r d e r to t a k e a p p r o p r i a t e a c t i o n (Oatley & J o h n s o n - L a i r d , 1987). In contrast, depression is a state in which p e r h a p s the i n d i v i d u a l does n o t have e n o u g h m o t i v a t i o n . In o t h e r words, d e p r e s s i o n m a y be r e g a r d e d as a state that reflects helplessness ( A b r a m s o n , Seligman & Teasdale, 1978) a n d hopelessness ( A b r a m s o n , M e t a l s k y & Alloy, 1989). Thus, it m a y be t h a t high levels o f depression inhibit a n x i e t y - r e l a t e d processing. H o w e v e r , this does n o t o v e r c o m e the p r o b l e m o f different loci o f o p e r a t i o n in the W i l l i a m s et al. model. Implicit in the model, though, is the n o t i o n t h a t the time course o f processing is generally slower as a function o f depression. The i m p l i c a t i o n o f this w o u l d be t h a t in a t a s k such as the S t r o o p , b y the time a n y a t t e n t i o n a l effects due to anxiety are b r o u g h t to b e a r on the c o l o u r - n a m i n g task, it has a l r e a d y been c o m p l e t e d (Williams, p e r s o n a l c o m m u n i c a t i o n ) . This is clearly a m a t t e r for empirical validation; for example, by v a r y i n g the a s y n c h r o n y between the onset o f the c o l o u r a n d the onset o f the word. A further issue o f i m p o r t a n c e concerns the question o f w h e t h e r the t a s k is sensitive e n o u g h to detect a processing bias at all. T h e fact t h a t it is sensitive is i n d i c a t e d by the use o f a similar m e t h o d o l o g y , using r a n d o m p r e s e n t a t i o n , voice response, a n d w o r d s f r o m the same stimulus source by T a g h a v i (1996) which revealed a clear S t r o o p bias in f a v o u r o f t h r e a t - r e l a t e d m a t e r i a l in clinically a n x i o u s children a n d adolescents. In sum, the findings o f the present study have shown t h a t depressed children a n d adolescents do n o t exhibit selective S t r o o p interference for m o o d - c o n g r u e n t material. This finding m a y therefore be t a k e n as p r o v i d i n g p a r t i a l s u p p o r t for the new a p p r o a c h e s to c h i l d h o o d depression ( H a r r i n g t o n , 1993; K o v a c s , 1986; R u t t e r , 1987) which e m p h a s i s e the similarities between depressive d i s o r d e r s in children a n d adults. Acknowledgements--We would like to thank Dr Sue Goode, Dr Eric Fombonne and their colleagues at the Children's Department of the Maudsley Hospital, Dr Neil Harris and his colleagues at Leigh House, Dr Veiol Bailey and her colleagues at the Thelma Golding Centre, Dr Margaret Murphy and her colleagues at the Bethlem Hospital, Dr. Anula Nikapota and her colleagues at the Brixton Child Guidance Clinic, and Dr. Julie Withecomb and her colleagues at the Bloomfield Centre (Guy's Hospital), for allowing us to see the patients under their care. We would also like to thank L. N. Law for help with computer programming and Dr Gisli Gudjonsson and an anonymous referee for their comments on an earlier draft of this paper. H. T. N. D., M. R. T. and A. R. M. are grateful for generous financial support by the Ministry of Culture and Higher Education of Islamic Republic of Iran during this study.
REFERENCES Abramson, L. Y., Metalsky, G. I. & Alloy, L. B. (1989). Hopelessness depression: A theory-based subtype of depression. Psychological Review, 96, 358-372. Abramson, L. Y., Seligman, M. E. P. & Teasdale, J. D. (1978). Learned helplessness in humans: Critique and reformulation. Journal of Abnormal Psychology, 87, 49-74. American Psychiatric Association (1994). Diagnostic and Statistical Manual of Mental Disorders, 4th edn., DSM-IV. Washington: American Psychiatric Association. Angold, A., Costello, E. J., Pickles, A. & Winder, F. (1987). The development of a questionnaire for use in epidemiological studies of depression in children and adolescents. Unpublished manuscript, University of London. Beck, A. T. (1976). Cognitive therapy of the emotional disorders. New York: New American Library. Birleson, P. (1981). The validity of depressive disorder in childhood and the development of a self-rating scale: A research report. Journal of Child Psychology and Psychiatry, 22, 73-88. Bower, G. H. (1981). Mood and memory. American Psychologist, 36, 129-148. Dalgleish, T. (1995). Performance on the emotional Stroop task in groups of anxious, expert and control subjects: A comparison of computer and card presentation formats. Cognition & Emotion, 9, 341-362. Dunn, L. M., Dunn, L. M., Whetton, C. & Pintilie, D. (1982). British Picture Vocabulary Scale. London: The Cromwell Press. Gotlib, I. H. & Cane, D. B. (1987). Construct accessibility and clinical depression: A longitudinal investigation. Journal of Abnormal Psychology, 96, 199-204. Gotlib, I. H. & MacCann, C. D. (1984). Construct accessibility and depression: An examination of cognitive and affective factors. Journal of Personality and Social Psychology, 47, 427-439. Harrington, R. C. (1993). Depressive disorder in children and Adolescents. Chichester: John Wiley & Sons. Hill, A. B. & Knowles, T. H. (1991). Depression and the emotional Stroop effect. Personality and Individual Differences, 12, 481-485. Kinderman, P. (1994). Attentional bias, persecutory delusions and the self-concept. British Journal of Medical Psychology, 67, 53-66. Klieger, D. M. & Cordner, M. D. (1990). The Stroop task as a measure of construct accessibility in depression. Personality and Individual Differences, 11, 19-27. Kovacs, M. (1982). The Children's Depression Inventory. Unpublished manuscript. University of Pittsburg. Kovacs, M. (1986). A developmental perspective on methods and measures in the assessment of depressive disorders: The
C h i l d h o o d depression a n d the modified S t r o o p test
759
clinical interview. In M. Rutter, C. E. Izard and R. B. Read (Eds.), Depression in young people: Developmental and clinical perspectives (pp. 435-465). New York: Guilford. Martin, M., Horder, P. & Jones, G. V. (1992). Integral bias in naming of phobia-related words. Cognition & Emotion, 6, 479-486. Mathews, A. & MacLeod, C. (1985). Selective processing of threat cues in anxiety states. Behaviour Research and Therapy, 23, 563-569. Mogg, K., Bradley, B. P., Williams, R. & Mathews, A. (1993). Subliminal processing of emotional information in anxiety and depression. Journal of Abnormal Psychology, 102, 304-311. Neshat Doost, H. T., Moradi, A. R., Taghavi, M. R., Yule, W. & Dalgleish, T. (submitted). The development of a corpus of emotional words produced by children and adolescents. Behavior Research Methods, Instruments & Computers. Oatley, K. & Johnson-Laird, P. N. (1987). Towards a cognitive theory of emotions. Cognition and Emotion, 1, 29-50. Reynolds, C. R. & Richmond, B. D. (1978). What I think and feel: a revised measure of children's manifest anxiety. Journal of Abnormal Child Psychology, 6, 271-280. Rust, J., Golombok, S. & Trickey, G. (1993). Wechsler Objective Reading Dimensions. London: The Psychological Corporation, Harcourt Brace Jovanovich. Rutter, M. (1987). The role of cognition in child development and disorder. British Journal of Medical Psychology, 60, 1-16. Schneider, S., Unnewher, S., Florin, I. & Margraf, J. (1992). Cognitive characteristics of children of panic patients. Paper presented at the International Congress of Psychology, Brussels. Seligman, M., Peterson, C., Kaslow, N., Tanenbaum, R., Alloy, L. & Abramson, L. (1984). Attributional style and depressive symptoms among children. Journal of Abnormal Psychology, 93, 235-238. Taghavi, M. R. (1996). The cognitive characteristics of children and adolescents with anxiety. Unpublished doctoral thesis, University of London. Williams, J. M. G. & Broadbent, K. (1986). Distraction by emotional stimuli: Use of a Stroop task with suicide attempters. British Journal of Clinical Psychology, 25, I01-110. Williams, J. M. G., Mathews, A. & Macleod, C. (1996). The emotional Stroop task and psychopathology. Psychological Bulletin, 120, 3 24. Williams, J. M. G., Watts, P. N., MacLeod, C. & Mathews, A. (1988). Cognitive psychology and emotional disorders. New York: Wiley. World Health Organization (1992). The ICD-10 Classification of Mental and Behavioral Disorders: Clinical Descriptions and Diagnostic Guidelines. Geneva: WHO.
Person. individ. Diff. Vol. 23, No. 5, pp. 753 759, 1997 © 1997 ElsevierScienceLtd. All rights reserved Printed in Great Britain
PIh S0191-8869(97)00097-4
0191-8869/97 $17.00+0.00
THE P E R F O R M A N C E OF C L I N I C A L L Y DEPRESSED C H I L D R E N A N D ADOLESCENTS ON THE M O D I F I E D STROOP P A R A D I G M Hamid T. Neshat Doost, 1. Mohammad R. Taghavi, ~Ali R. Moradi, ~William Yule 1 and Tim Dalgleish2 ~Department of Psychology, Institute of Psychiatry, De Crespigny Park, Denmark Hill, London, England and 2MRC Applied Psychology Unit, Cambridge, England
(Received 25 February 1997) Summary--The study investigated selective processing of emotional information in childhood depression using a computerised version of the modified Stroop colour naming task. Three groups of children and adolescents--clinicallydepressed (n = 19), patients with mixed depression and anxiety (n = 19), and normal controls (n = 26)---were required to name the colours of depression-related, threat-related, trauma-related, happy, and categorised neutral words. Normal subjects, compared with the patient groups, exhibited faster colour naming overall. There were no significant differences on the colour naming of different categories of words and the performance across the three groups was not significantly different for different categories of words. O 1997 Elsevier Science Ltd
Keywords: childhood depression, information processing, cognitive biases, Stroop task
There is now convincing evidence that children and adolescents can and do suffer from depressive disorders which are similar to those of adults in many aspects (e.g. Rutter, 1987). Several research studies have used self-report measures (questionnaires) to investigate the cognitive characteristics of depression in children and adolescents. Results of these studies have shown that depressed children and adolescents display the same kind of cognitive characteristics as those observed in depressed adults, including depressive attributional style, low self esteem, cognitive biases, and less positive self schemata (e.g. Seligrnan, Peterson, Kaslow, Tanenbaum, Alloy & Abramson, 1984). Recently researchers have started to use experimental approaches to investigate cognitive biases of emotional disorders in children and adolescents. One of the most frequently used of these tasks is the modified Stroop paradigm. This task requires subjects to name ink colours in which emotional and non-emotional words are written. The emotional Stroop effect refers to the finding that colournaming times are slower for emotional words than for non-emotional words. This task has been used in several studies with non-clinically and clinically anxious and depressed adults. There is considerable evidence that anxious subjects, but not normal controls, have longer reaction times on colour naming of threatening words in the modified Stroop task (Williams, Mathews & Macleod, 1996). These effects seem much clearer than in the case of depression. In the case of depressed subjects, researchers have used different methods of presentation (card, tachistoscope, and computer), different groups of words (general negative vs. negative self-descriptive adjectives), and they have studied different groups of subjects (clinical, subclinical, mood induced, mixed depressed/anxious). These studies have led to different results. Some have found more Stroop interference for negative or depressed words relative to neutral or positive words (Gotlib & Cane, 1987; Gotlib & MacCann, 1984; Kinderman, 1994; Klieger & Cordner, 1990) while others have not (Hill & Knowles, 1991; Mogg, Bradley, Williams & Mathews, 1993). There are a few studies using the modified Stroop task in children and adolescents. Schneider, Unnewher, Florin, and Margraf (1992) found that children of phobic patients exhibited higher
*To whom all correspondence should be addressed at present address: Department of Psychology, University of Isfahan, Chaharbag Bala, Hezarjerib, Isfahan, Iran. 753
754
Hamid T. Neshat Doost et al.
interference scores to phobia-relevant words compared with children of patients with panic disorder and children of parents without any psychiatric morbidity. Martin, Horder, and Jones (1992) found that spider-phobic children were significantly slower to colour-name spider words than control words. Taghavi (19~6) studied the performance of children and adolescents with Generalised Anxiety Disorder on a modified Stroop task. Results of this study showed that, compared to normal controls, children and adolescents with Generalised Anxiety Disorder showed longer colour-naming times for threat-related and trauma-related words, with longer colour-naming times for threat-related words than trauma-related words. At the time of writing there is no published study on the performance of depressed children and adolescents on the modified Stroop task and the purpose of the present study, therefore, is to investigate the modified Stroop effect in clinically depressed children and adolescents. A third group of children and adolescents with mixed depression/anxiety was used to investigate content-specificity effects with depression-related and anxiety-related material.
METHODOLOGY Subjects
There were three groups of subjects: depressed, mixed depressed/anxious, and normal controls. All subjects were aged between 9 and 18 yr and their primary language was English. The age range was 10.92-18.33 yr for the depressed group, 9.50-18.00 yr for the mixed depressed/anxious group, and 13.25-18.17yr for the control group. The patient groups were recruited from a variety of sources including hospital outpatients' and inpatients' services in the UK (Maudsley Hospital, London; Leigh House, Eastleigh; Thelma Golding Centre, London; Bethlem Hospital, London; Brixton Child Guidance Clinic, London; Bloomfield Centre, London; St. Thomas' Hospital, London, and Camberwell Child Guidance Clinic, London). There were 19 subjects (11 girls and eight boys) in the depressed group; the selection criteria were as follows. (a) A primary diagnosis of Major Depressive Disorder according to DSM-IV (Diagnostic and Statistical Manual of Mental Disorders, 4th edition, American Psychiatric Association, 1994) criteria or ICD-10 (International Classification of Diseases, World Health Organization, 10th revision, WHO, 1992) criteria for Depressive Disorder. The diagnosis was determined in a clinical assessment by psychiatrists and psychologists before the patients were introduced to the experimenter. (b) Scores of higher than cut-off on all three depression scales: the Children's Depression Inventory (CDI; Kovacs, 1982), the Depression Self-Rating Scale (DSRS; Birleson, 1981) and the Mood and Feeling Questionnaire (MFQ; Angold, Costello, Pickles & Winder, 1987). There were 19 subjects in the mixed depressed/ anxious group (12 girls and seven boys). The selection criteria were diagnoses of both Depressive Disorder and Anxiety Disorder according to DSM-IV, or a primary diagnosis of Mixed Anxiety and Depressive Disorder according to ICD-10 criteria. The control group consisted of 26 subjects (16 girls and 10 boys), with no known history of emotional disorder. Subjects with vision problems, colour blindness, or brain damage were excluded. The patients were generally asked to participate in the study before the commencement of any medication regime. The results of the questionnaires showed that the depressed children and adolescents were very depressed at the time of the study. In order to control for the effects of confounding variables, the groups were matched for sex, age, vocabulary level, and reading ability. Measures
1. 2. 3. 4. 5.
The Depression Self-Rating Scale (DSRS; Birleson, 1981); The Children's Depression Inventory (CDI; Kovacs, 1982); The Mood and Feeling Questionnaire (MFQ; Angold et al., 1987); The Revised Children's Manifest Anxiety Scale (RCMAS; Reynolds & Richmond, 1978); The British Picture Vocabulary Scale (BPVS Short Form, Dunn, Dunn, Whetton & Pintilie, 1982); 6. The Wechsler Objective Reading Dimensions (WORD; Basic Reading; Rust, Golombok & Trickey, 19933.
Childhood depression and the modified Stroop test
755
Table 1. Means and standard deviations (SD) of age, reading score, vocabulary score, and depression and anxiety measures for the three groups of subjects Depressed
Age (yr) WORD BPVS DSRS CDI MFQ RCMAS
Mixed
Non-patients
Mean
SD
Mean
SD
Mean
SD
15.57 101.84 96.16 23.35 30.60 44.50 21.15
1.58 15.46 10.66 5.83 8.46 12.78 4.01
14.84 104.10 100.05 18.84 23.42 36.47 18.63
2.48 13.53 13.48 7.26 12.00 16.72 6.05
15.15 98.50 91.42 7.50 6.80 13.57 8.11
1.44 14.59 18.53 3.90 6.40 9.06 5.68
WORD, Wechsler Objective Reading Dimensions (Basic Reading); BPVS, British Picture Vocabulary Scale; DSRS, Depression Self-Rating Scale; CDI, Children's Depression Inventory; MFQ, Mood and Feeling Questionnaire; RCMAS, Revised Children's Manifest Anxiety Scale.
Instrumentation An IBM (Thinkpad 755C TF1) portable computer with a 26.41 cm active-matrix T F T colour LCD screen was used to present the stimuli. A voice key was used to register the vocal response of the subjects automatically. The voice key connected to an adjustable head microphone (Bandridge EH620, Dynamic Headsets) worn by the subject. Five buttons (one for each colour: red, green, blue and yellow, and another button for changing the word on the screen when it did not disappear automatically following the subject's vocal response) were also used.
Materials One hundred and twenty words were used as verbal stimuli. These consisted of 12 words from each of five word categories: happy (e.g. pleased), categorised neutral (animals; e.g. sheep), depression-related (e.g. helpless), threat-related (e.g. terrified), and trauma-related (e.g. casualty). Each word was presented once in each of two colours chosen at random from the four colours: green, yellow, red, blue. The words were selected from a source of words produced by 231 primary and secondary school pupils suitable for the age range of subjects in the present study (Neshat Doost, Moradi, Taghavi, Yule & Dalgleish, submitted). All categories were matched for length and frequency. The words were presented in capital letters 8 mm high. The visual angle subtended was approximately 1°
Procedure The test was carried out individually. Each subject was first asked to complete the reading and vocabulary tests. Then s/he was asked to sit in front of the computer, 50 cm from the screen, in a quiet room, without any disruptions. Before the main task, there were some practice trials, which were the same as the experimental trials except that the stimulus words were all uncategorised neutral words. In the main task, 120 stimulus words were presented in a new fully random order for each subject. The presentation time for each word was 1.7 sec. The inter-trial interval was 2.0 sec. The subject was asked to ignore the word and say out loud the colour name into the microphone as quickly as possible. After 40 trials subjects were given a short rest. The task took about 20 min in total. Following the Stroop task the subjects were asked to fill in the depression and anxiety scales. RESULTS
Subject characteristics Means and standard deviations were calculated separately for age, reading, vocabulary, and depression and anxiety scales for the two patient groups and the control group (see Table 1). One-way analyses of variance were used to compare the three groups of subjects on age, reading, vocabulary, depression, and anxiety. There were no significant differences between the three groups for age ( F < 1), vocabulary (F(2,61)= 1.758 NS) and reading ability ( F < 1), but the groups scored significantly differently on the measures of depression (DSRS: F(2,63)=49.036, P < 0.0001; CDI:
756
Hamid T. Neshat Doost et al. Table 2. Mean reaction times and standard deviations (SD) in seconds per word for the five groups of words in each of three groups of subjects Group
Happy
Neutral
Depression
Threat
Trauma
Depressed
Mean SD
0.83 0.16
0.83 0.14
0.84 0.14
0.85 0.15
0.84 0.16
Mixed
Mean SD
0.81 0.18
0.83 0.18
0.83 0.18
0.82 0.17
0.84 0.20
Control
Mean SD
0.69 0.14
0.69 0.13
0,69 0.15
0.69 0.14
0.69 0.14
F(2,63)=44.181, P<0.0001; MFQ: F(2,63)=35.227, P<0.0001) and RCMAS (F(2,63)=39.510, P<0.0001) as expected. To clarify differences between the three groups, post hoc multiple comparison tests (Student-Newman-Keuls; P < 0 . 0 5 ) were used. On the anxiety scale (RCMAS) the depressed group, and the group which was diagnosed as mixed depressed/anxious had significantly higher scores than the control group (P<0.05). Although the depressed group showed a higher anxiety score than the mixed depression/anxiety group, this difference was not significant. On the depression scales, the depressed group and the mixed depressed/anxious group had significantly higher scores than the control group ( P < 0.05). Also, the depressed group had significantly higher scores on the DSRS, and the CDI than the mixed depressed/anxious group (P<0.05) but not on the MFQ. Colour-naming reaction times
The mean ( _ SD) number of colour naming errors was 2.07 + 2.26 across all subjects. The mean numbers of colour naming errors for the depressed, mixed depressed/anxious and control groups were 3.58_+ 3.22, 2.05_+ 1.81 and 1.27_+ 1.12, respectively. A one-way analysis of variance showed that there were significant differences between the three groups on the number of colour naming errors (F(2,63) = 6.509, P < 0.01). To clarify differences between the three groups, Student-NewmanKeuls tests were again used. The results showed that the depressed group made significantly more colour naming errors than the control group (P < 0.05). There were no significant differences between the depressed group and the mixed depressed/anxious group or between the mixed group and the control group. Colour naming latency data were excluded from the analyses on those trials on which subjects made colour naming errors. Also, to minimise the influence of outlying data points, colour naming latency data were excluded from the analyses on those trials when the response time was less than 100 msec or more than 2000 msec. Response times less than 100 msec are impossibly fast and suggest that subjects had vocalised prematurely. Response times of more than 2000 msec could be due to lapses of attention or the computer failing to register a response. The mean colour naming latencies taken by the three groups for threat-related words, depression-related words, happy words, neutral words and trauma-related words are shown in Table 2. A repeated measures analysis of variance was carried out with Group (3: depressed, mixed depressed/anxious, and control) as a between-subjects variable and Word Type (5: depressionrelated, threat-related, trauma-related, happy, and categorised neutral) as a within-subjects variable. The results showed a significant main effect of Group (F(2,61)= 6.76, P < 0.05). The effect of Word Type was not significant ( F < 1) and no significant interaction was found among Group and Word Type ( F < 1). To examine the main effect of Group, Student-Newman-Keuls tests were used. The results showed that, compared with the control group, the depressed group and the mixed depressed/anxious group had significantly longer colour-naming times (P < 0.05). Correlational analyses
Correlation coefficients (Pearson) were calculated between anxiety and depression scores and interference scores from the four categories of words: depression-related, threat-related, traumarelated, and happy. Interference scores were calculated by subtracting the mean latency for categorised neutral words (animal names) from that for the relevant word type. Results showed that the correlations between anxiety and depression scores and mean interference scores of the four categories of words were not significant.
Childhood depression and the modified Stroop test
757
DISCUSSION The purpose of this study was to compare the performance of three groups of children and adolescents (depressed, mixed depressed/anxious and normal controls) on their performance on the modified Stroop task. The results showed that, compared with the control group, patient groups (depressed and mixed depressed/anxious) were in general significantly slower on Stroop colour-naming. This finding replicated the findings of other studies with adults which have found similar main effects (Hill & Knowles, 1991; Kinderman, 1994; Williams & Broadbent, 1986). This retarded colour naming could be due to the effects of psychomotor retardation which is one of the general symptoms of depression. According to Kinderman (1994), this slowness in colour naming may be interpreted as a general effect of psychiatric disturbance. The clinical groups were not significantly slower on the depression-related, threatening, and trauma-related words. These results are inconsistent with Beck's (Beck, 1976) and Bower's (Bower, 1981) models of cognitive processing of emotion which predict that emotion-congruent biases operate in all stages of processing in depression and anxiety, with these biases differing only in terms of content-specificity. In other words, anxious individuals selectively process anxiety-related information, whereas depressed individuals selectively process depression-relevant material. However, this finding can be taken as providing some support for the model proposed by Williams, Watts, MacLeod, and Mathews (1988), which proposes that "depression is less associated with biased pre-attentive priming and more with bias in the active strategic aspect of encoding" (p. 178). The results of the present study with young subjects are therefore consistent with the results of some of the studies on depressed adults which showed that depressed subjects did not display interference effects for negative information on the modified Stroop task (Hill & Knowles, 1991; Mogg et al., 1993). The present study with children and adolescents shows a different profile of results to previous studies which have shown evidence of interference effects of negative information in depressed adult subjects (Gotlib & Cane, 1987; Gotlib & MacCann, 1984; Kinderman, 1994; Klieger & Cordner, 1990). However, as well as different subject populations there were a number of methodological differences between these latter studies and the present one. In some of the adult studies anxiety was not assessed, even though there is often comorbidity between anxiety and depression. Therefore, the Stroop effect could possibly have been due to anxiety rather than depression. Further support for this hypothesis is presented by Mathews and MacLeod (1985) who also showed that anxious patients who showed disruption on an emotional Stroop task were also more depressed than control subjects. However, partial correlational analysis showed that it was anxiety and not depression which was associated with longer colour naming of negative words. Furthermore, some of the adult findings were obtained with a modified version of the Stroop task in which each category of stimulus words (e.g. depression-related words) was presented in blocked format on a card. Accordingly, it is likely that the blocked presentation increased elaborative processing of the semantic content of the negative words and led to longer colour naming times for these words. In the present study, the words were presented individually in random order. Therefore, the possibility of elaborative processing was reduced. In the present study, child and adolescent patients had very high scores on an anxiety scale and in the mixed group they met criteria for an anxiety disorder; however, they did not show a Stroop bias for negative information despite their typically high level of anxiety. Results of several studies of adults have shown that, compared with normal controls, anxious patients and subclinically anxious individuals have significantly more interference for negative words (especially for threatrelated words) than for neutral words (e.g. Dalgleish, 1995). Furthermore, results of studies on anxious children and adolescents (e.g. Taghavi, 1996) have shown that anxious children and adolescents showed longer colour-naming times for threat-related words related to their emotional disorders. Taken together, it can be concluded that the presence of a high level of anxiety is not sufficient to elicit a preattentive bias for negative information. Then, the question of importance becomes why depressed subjects do not show a preattentive bias for negative information despite their high level of anxiety? This is particularly pertinent because in the Williams et al. (1988) model referred to above, anxiety and depression operate at different loci. Consequently, high levels of anxiety should affect processing regardless of the co-existence of depression.
758
Hamid T. Neshat Doost et al.
O n e possible answer is the difference between anxiety a n d depression. A n x i e t y is a state in which the i n d i v i d u a l is m o t i v a t e d to detect p o t e n t i a l l y aversive stimuli in o r d e r to t a k e a p p r o p r i a t e a c t i o n (Oatley & J o h n s o n - L a i r d , 1987). In contrast, depression is a state in which p e r h a p s the i n d i v i d u a l does n o t have e n o u g h m o t i v a t i o n . In o t h e r words, d e p r e s s i o n m a y be r e g a r d e d as a state that reflects helplessness ( A b r a m s o n , Seligman & Teasdale, 1978) a n d hopelessness ( A b r a m s o n , M e t a l s k y & Alloy, 1989). Thus, it m a y be t h a t high levels o f depression inhibit a n x i e t y - r e l a t e d processing. H o w e v e r , this does n o t o v e r c o m e the p r o b l e m o f different loci o f o p e r a t i o n in the W i l l i a m s et al. model. Implicit in the model, though, is the n o t i o n t h a t the time course o f processing is generally slower as a function o f depression. The i m p l i c a t i o n o f this w o u l d be t h a t in a t a s k such as the S t r o o p , b y the time a n y a t t e n t i o n a l effects due to anxiety are b r o u g h t to b e a r on the c o l o u r - n a m i n g task, it has a l r e a d y been c o m p l e t e d (Williams, p e r s o n a l c o m m u n i c a t i o n ) . This is clearly a m a t t e r for empirical validation; for example, by v a r y i n g the a s y n c h r o n y between the onset o f the c o l o u r a n d the onset o f the word. A further issue o f i m p o r t a n c e concerns the question o f w h e t h e r the t a s k is sensitive e n o u g h to detect a processing bias at all. T h e fact t h a t it is sensitive is i n d i c a t e d by the use o f a similar m e t h o d o l o g y , using r a n d o m p r e s e n t a t i o n , voice response, a n d w o r d s f r o m the same stimulus source by T a g h a v i (1996) which revealed a clear S t r o o p bias in f a v o u r o f t h r e a t - r e l a t e d m a t e r i a l in clinically a n x i o u s children a n d adolescents. In sum, the findings o f the present study have shown t h a t depressed children a n d adolescents do n o t exhibit selective S t r o o p interference for m o o d - c o n g r u e n t material. This finding m a y therefore be t a k e n as p r o v i d i n g p a r t i a l s u p p o r t for the new a p p r o a c h e s to c h i l d h o o d depression ( H a r r i n g t o n , 1993; K o v a c s , 1986; R u t t e r , 1987) which e m p h a s i s e the similarities between depressive d i s o r d e r s in children a n d adults. Acknowledgements--We would like to thank Dr Sue Goode, Dr Eric Fombonne and their colleagues at the Children's Department of the Maudsley Hospital, Dr Neil Harris and his colleagues at Leigh House, Dr Veiol Bailey and her colleagues at the Thelma Golding Centre, Dr Margaret Murphy and her colleagues at the Bethlem Hospital, Dr. Anula Nikapota and her colleagues at the Brixton Child Guidance Clinic, and Dr. Julie Withecomb and her colleagues at the Bloomfield Centre (Guy's Hospital), for allowing us to see the patients under their care. We would also like to thank L. N. Law for help with computer programming and Dr Gisli Gudjonsson and an anonymous referee for their comments on an earlier draft of this paper. H. T. N. D., M. R. T. and A. R. M. are grateful for generous financial support by the Ministry of Culture and Higher Education of Islamic Republic of Iran during this study.
REFERENCES Abramson, L. Y., Metalsky, G. I. & Alloy, L. B. (1989). Hopelessness depression: A theory-based subtype of depression. Psychological Review, 96, 358-372. Abramson, L. Y., Seligman, M. E. P. & Teasdale, J. D. (1978). Learned helplessness in humans: Critique and reformulation. Journal of Abnormal Psychology, 87, 49-74. American Psychiatric Association (1994). Diagnostic and Statistical Manual of Mental Disorders, 4th edn., DSM-IV. Washington: American Psychiatric Association. Angold, A., Costello, E. J., Pickles, A. & Winder, F. (1987). The development of a questionnaire for use in epidemiological studies of depression in children and adolescents. Unpublished manuscript, University of London. Beck, A. T. (1976). Cognitive therapy of the emotional disorders. New York: New American Library. Birleson, P. (1981). The validity of depressive disorder in childhood and the development of a self-rating scale: A research report. Journal of Child Psychology and Psychiatry, 22, 73-88. Bower, G. H. (1981). Mood and memory. American Psychologist, 36, 129-148. Dalgleish, T. (1995). Performance on the emotional Stroop task in groups of anxious, expert and control subjects: A comparison of computer and card presentation formats. Cognition & Emotion, 9, 341-362. Dunn, L. M., Dunn, L. M., Whetton, C. & Pintilie, D. (1982). British Picture Vocabulary Scale. London: The Cromwell Press. Gotlib, I. H. & Cane, D. B. (1987). Construct accessibility and clinical depression: A longitudinal investigation. Journal of Abnormal Psychology, 96, 199-204. Gotlib, I. H. & MacCann, C. D. (1984). Construct accessibility and depression: An examination of cognitive and affective factors. Journal of Personality and Social Psychology, 47, 427-439. Harrington, R. C. (1993). Depressive disorder in children and Adolescents. Chichester: John Wiley & Sons. Hill, A. B. & Knowles, T. H. (1991). Depression and the emotional Stroop effect. Personality and Individual Differences, 12, 481-485. Kinderman, P. (1994). Attentional bias, persecutory delusions and the self-concept. British Journal of Medical Psychology, 67, 53-66. Klieger, D. M. & Cordner, M. D. (1990). The Stroop task as a measure of construct accessibility in depression. Personality and Individual Differences, 11, 19-27. Kovacs, M. (1982). The Children's Depression Inventory. Unpublished manuscript. University of Pittsburg. Kovacs, M. (1986). A developmental perspective on methods and measures in the assessment of depressive disorders: The
C h i l d h o o d depression a n d the modified S t r o o p test
759
clinical interview. In M. Rutter, C. E. Izard and R. B. Read (Eds.), Depression in young people: Developmental and clinical perspectives (pp. 435-465). New York: Guilford. Martin, M., Horder, P. & Jones, G. V. (1992). Integral bias in naming of phobia-related words. Cognition & Emotion, 6, 479-486. Mathews, A. & MacLeod, C. (1985). Selective processing of threat cues in anxiety states. Behaviour Research and Therapy, 23, 563-569. Mogg, K., Bradley, B. P., Williams, R. & Mathews, A. (1993). Subliminal processing of emotional information in anxiety and depression. Journal of Abnormal Psychology, 102, 304-311. Neshat Doost, H. T., Moradi, A. R., Taghavi, M. R., Yule, W. & Dalgleish, T. (submitted). The development of a corpus of emotional words produced by children and adolescents. Behavior Research Methods, Instruments & Computers. Oatley, K. & Johnson-Laird, P. N. (1987). Towards a cognitive theory of emotions. Cognition and Emotion, 1, 29-50. Reynolds, C. R. & Richmond, B. D. (1978). What I think and feel: a revised measure of children's manifest anxiety. Journal of Abnormal Child Psychology, 6, 271-280. Rust, J., Golombok, S. & Trickey, G. (1993). Wechsler Objective Reading Dimensions. London: The Psychological Corporation, Harcourt Brace Jovanovich. Rutter, M. (1987). The role of cognition in child development and disorder. British Journal of Medical Psychology, 60, 1-16. Schneider, S., Unnewher, S., Florin, I. & Margraf, J. (1992). Cognitive characteristics of children of panic patients. Paper presented at the International Congress of Psychology, Brussels. Seligman, M., Peterson, C., Kaslow, N., Tanenbaum, R., Alloy, L. & Abramson, L. (1984). Attributional style and depressive symptoms among children. Journal of Abnormal Psychology, 93, 235-238. Taghavi, M. R. (1996). The cognitive characteristics of children and adolescents with anxiety. Unpublished doctoral thesis, University of London. Williams, J. M. G. & Broadbent, K. (1986). Distraction by emotional stimuli: Use of a Stroop task with suicide attempters. British Journal of Clinical Psychology, 25, I01-110. Williams, J. M. G., Mathews, A. & Macleod, C. (1996). The emotional Stroop task and psychopathology. Psychological Bulletin, 120, 3 24. Williams, J. M. G., Watts, P. N., MacLeod, C. & Mathews, A. (1988). Cognitive psychology and emotional disorders. New York: Wiley. World Health Organization (1992). The ICD-10 Classification of Mental and Behavioral Disorders: Clinical Descriptions and Diagnostic Guidelines. Geneva: WHO.