SELF-CONTROL OF EMOTIONAL REACTIONS BY YOUNG CHILDREN

/ Child. PsychoL Psychiat. Vol. 27, No. 3, pp. 357-366, 1986 0021-9630/86 $3.00 + 0.00Printed in Great Britain. Pergamon Journals Ltd.

© 1986 Association for Child Psychology and Psychiatry.

SELF-CONTROL OF EMOTIONAL REACTIONS BY YOUNGCHILDREN

MARK MEERUM TERV^^OGT,* JAN SCHENE* and PAUL L . HARRISI

*Free University, Amsterdam, The Netherlands and fUniversity of Oxford, U.K.

Abstract—Six-year-old children were instructed to adopt a detached or an involved attitude while theylistened to a sad story. The effects of these instructions became clear by the children's self-report usinga non-verbal emotion sccile, their way of reproducing the story they had to listen to, and theirachievements on an unrelated picture-memory task. Differences between boys and girls were foundand discussed as well as the nature of some of the self-control strategies reported by the children.

Keywords: Emotion, control, children, strategies

INTRODUCTION

IN THE last 20 years a number of empirical studies on the control of emotion—itsreduction and enhancement—has appeared. Lazarus and his co-workers, in particular,have provided us with a substantial amount of systematic knowledge about adultstrategies for coping with emotional situations (e.g. Averill, 1973; Folkman, Schaefer& Lazarus, 1979; Koriat, Melkman, Averill & Lazarus 1972; Lazarus, 1966,1968, 1975; Lazarus, Averill & Opton, 1970; Lazarus, Coyne & Folkman,1982). In many everday situations a person has no alternative but to endure apotentially noxious stimulus. In some of these cases behavioural control (e.g. escape)will be possible. In other cases, however, direct action on the environment is blockedand the only way left open will be cognitive control. The way a person construesor cognizes his/her relationship with the environmental context of the moment hasa pivotal significance for an emotional reaction. An extensive body of literature (see,for example, Lazarus, 1966) makes it clear that such appraisals are able to increaseas well as reduce the emotional impact of a situation. In children, however, thiscognitive control is still relatively unexplored.

In an interview-study (Harris, Olthof & Meerum Terwogt, 1981) even six-year-old children claimed that they have some control over the display of emotion[i.e. they could hide feelings from others or pretend to feel differently (Saarni, 1979)],and also over their actual emotion. Naturally, we have no guarantee that their claimsare correct. Moreover, when these children were asked what they would do toaccomplish a change of emotion, about half of the children failed to answer. Theother half, almost exclusively, proposed changing the external situation (e.g. "callin some friends") or the response (e.g. "dry your tears"). Thus Harris et al. (1981)

Requests for reprints to: Dr M. Meerum Terwogt, Department of Developmental Psychology, FreeUniversity, Koningslaan 22, 1075 AD Amsterdam, The Netherlands.

Accepted manuscript received b June 1985

357

3 5 8 MARK MEERUM TERW^OGT, JAN SCHENE AND PAUL L. HARRIS

described them as 'young behaviourists'. The six-year-olds hardly ever mentionedthe mental strategies, such as inner redirections of thought, that were described byolder children (Harris et al, 1981) and adults (Koriat et al, 1972).

In the present experiment we wanted to study young children's self-control, whena mental strategy was the only one available. Briefly, six-year-olds were instructedto become detached from or involved in a sad story. Their success in these effortswas assessed by registering the effects of the resulting emotion on subsequent cognitivetasks, as well as by self-report.

METHODSubjects

The subjects were 160 children between 6 years, 3 months and 7 years, 1 month old (mean = 6 years,7 months). There were equal numbers of each sex. The children were randomly selected from fourstate schools recruiting children from a predominantly middle class socio-economic background.

MaterialsThe children were instructed to listen to a taped story about six minutes in length (i.e. 730 words).

The theme was as follows: a child of about the same age as the subject meets his best friend in theschoolyard for the last time; tomorrow the friend will, against his will, move with his parents to anothertown; after the conversation, the boy feels lonely and walks the streets for a while; consequently heis late for dinner; unable to tell his parents what has happened, he is sent to his room without dinner;at this point he feels completely deserted by everyone. Female subjects heard an identical version ofthe tape featuring two girlfriends (Feshbach & Roe, 1968).

In order to establish the impact of the story on the children's mood, we used several indices. Asa subjective index, we asked the children to indicate their mood on a non-verbal emotion scale. Todo this, they had to select one of five schematic drawings of faces whose expression matched their ownfeelings at that moment. The faces formed a five-point scale, with expressions ranging from very happyto very sad.

Two tasks were selected that could index the intensity of sadness in the child more objectively. Thefirst index depended on the interferring effect of negative emotion on cognitive achievement (Easterbrook,1959). A study by Masters, Barden and Ford (1979) showed that young children's performanceon a memory task was influenced even by moderate emotion. We used essentially the same task: thechild had to look at 14 picture-pairs (e.g. an elephant and a bus) at a presentation-rate of one pairevery one and a half seconds. During this procedure, the experimenter randomly pointed to one ofeach pair on presentation. After a delay of 15 seconds, the child saw the same picture-pairs, in thesame order, for a second time and was asked to indicate which alternative the experimenter had pointedto in the previous presentation.

The second index was obtained with another memory task: we asked subjects to reproduce the story'as precisely as possible'. This time, the induced mood was expected to work in the opposite direction.Instead of the interference we expected for the unrelated task, here emotional mood should serve asa distinctive context for the storage and retrieval of specific memories (Bower, Monteiro & Gilligan,1978; Izard, Wehmer, Livsey & Jennings, 1965). In other words, in terms of Bower's (1981)'mood-state-dependent theory', there is a mood-congruity effect in learning and recalling affectivelycharged narratives. When a sad mood provides a helpful retrieval cue for sad material, we would expectsad children to have better recall of negatively charged passages.

ProcedureAll children were tested individually. Subjects were randomly assigned to four groups, each comprising

20 boys and 20 girls. A no-story (NS) control group, that only got the picture-memory task, servedas a standard for the achievements of this age group on such a task under normal conditions (i.e. withoutemotion-induction). The procedure for the three experimental groups differed only with respect to theinstruction introducing the sad story. All children were told that they had to listen very carefully sothat they could retell the story afterwards. The detached (D) group was also told that the forthcoming

SELF-CONTROL OF EMOTIONAL REACTIONS 359

story was a sad one, but that they had to listen to it in such a way "that you won't become sad yourself.The involved (I) group was given different instructions. They had to try to listen in such a way thatthey would feel the sadness themselves. Finally, a control (C) group was added in which children listenedto the story without any further introduction other than "to listen carefully".

Just before and after listening to the story, the children had to indicate their current mood on thenon-verbal emotion scale. Immediately afterwards, they all completed, without delay and in the samesequence, the picture-memory task, and the story-recall. This session ended with a short interview.All children were asked if they thought the story had a really sad theme and if they were really movedby it. The D- and I-groups were also asked what they had done in order to comply with the particularinstructions that they had been given.

ScoringNon-verbal emotion scale. The schematic expressions were given a score from 1 (very sad) to 5 (very

happy) and the positions of each child on this five-point scale, before and after listening to the story,were recorded. In the analysis, two indices were taken into account. These were the absolute moodof the child immediately after listening to the story and the change of mood that was caused by thestory. For this, the second measurement was corrected in terms of the original mood-state in whichthe child happened to be before the story was presented.

Picture-memory task. The number of pictures correctly chosen out of a maximum of 14 was recorded.Story reproduction. The completeness of the reproductions was assessed with two indices: the total length

of the stories (number of words used) and the number of information units that were still present inthe reproduction. For this last index, the original story was divided into 76 propositions (a sentenceor part of a sentence), each being considered as a separate information unit (Mandler & Johnson, 1977).Two independent judges decided on the number of propositions referred to in the reproductions bythe children (interjudge agreement: 0.87). In addition to these indices of completeness, the reproducedstories were scored for their emotional impact. Each explicit emotion label (e.g. "sad", "shocked"of "crying") received a score of 2, while each fragment that carried a clear connotation of an emotionalstate (e.g. "he fell down on his bed") received a score of 1 (interjudge agreement: 0.90). Disagreementsbetween judges were resolved by discussion.

For each index, the differences between the three experimental groups were analysed using an ANOVAprocedure. In the case of the picture-memory task the performance of the control group was also includedin the analyses. Information from the introspective domain (i.e. scores on the emotion scale) was comparedwith the emotional effects on the cognitive tasks by means of correlational techniques.

RESULTS

Non-verbal emotional scaleOn the whole, the children appeared to enjoy the experimental situation. In Fig.

1, which indicates the mood of the three experimental groups before and after listeningto the story, we see that initially all groups were in the (fairly) happy section of thescale. Only the I-groups were affected by the story to the extent that they went belowthe neutral point. A 2 x 3 covariance analysis on the factors sex and instruction, in whichthe measurement before testing is checked, confirms this observation and producesa significant effect of instruction [F(2,113) = 8.48, P< O.OOl]. The other effects donot reach a significant level, though there is a tendency for girls to be more affectedby the story than boys.

Post hoc analysis of the instruction effect shows that the D- and C-groups do notdiffer significantly, though both are less sad than the I-group (Duncan's MultipleRange test, 0.01 level).

Picture-memory taskIn Table 1 the mean scores for the memory task are given. A 2 x 4 analysis of

360 MARK MEERUM TERW^OGT, JAN SCHENE AND PAUL L. HARRIS

Very happy 5

•aoo

3.5

Normal 3

'•^V^ • (b)

•••••.7.-;-(b)

" " • (g)

\• (b)

••(g)

Very sad 1Before

listeningAfter

listening

Measurement

Fig. 1. Mean scores of the non-verbal emotion of the three experimental groups, before and after listeningto the story, divided by sex. : D(etachment)-group; : (I(nvolvement)-group; :

C(ontrol)-group.

TABLE 1.

Sex

BoysGirls

MEAN NUMBER OF CORRECTLY REMEMBERED

FUNCTION OF SEX AND GROUP

D-group

11.7512.15

PICTURES

GroupI-group C-group

10.808.85

11.8510.50

(OUT OF 14) AS A

NS-group

12.5512.95

variance of sex x group reveals that there is a significant difference between the groups[F(3, 152) = 16.39, P < 0.01]. In addition, girls score less than boys [sex x group:F (3,152) = 3.86, P<0.05].

Post hoc analysis of the group effect (Duncan's Multiple Range Test) shows thatamong the boys, the I-group deviates only from the control group (P< 0.01). Amongthe girls, the I-group scores significantly lower than all other groups, while theC-group also scores lower than the NS-group ( J P < 0 . 0 1 ) .

Reports of a sad mood on the non-verbal emotion scale after listening to thestory correlate with poorer performance on the picture-memory task [r(l 18) = 0.52,P< 0.001 ] . This correlation increases to 0.56 if the degree of mood change is usedrather than final mood. This is true to almost the same degree for boys (0.54) andgirls (0.57). Thus both the mood expressed and the change of mood are goodindications of subsequent memory performance.

Story recallAn analysis of instruction x sex failed to produce significant results when applied

to the scores for the mean number of words used in the reproductions. In nearly everyprotocol seven propositions appeared which captured the heart of the story. On theother hand, about 30% of the minor propositions were not mentioned by anybody.

SELF-CONTROL OF EMOTIONAL REACTIONS 361

TABLE 2. AVERAGE EMOTION SCORE FOR THE REPRODUCEDSTORIES AS A FUNCTION OF SEX AND GROUP

GroupSex D-group I-group G-group

7.65 9.40 7.15(106.76) (105.50) (96.00)

Girls 7.40 12.95 9.05(97.50) (109.50) (92.00)

The average length of the stories, which is included in theanalysis as a covariable, is given between parentheses

Mean scores for the emotional impact of the story are presented in Table 2. Sincelonger stories would be more likely to contain emotion words or connotations, lengthof the story was included as a co-variable in an instruction x sex analysis of thesescores. Girls were more emotional in their reproduction of the story [F(l,113) = 7.58,P < 0.01 ] . Moreover, there was a significant effect of instruction [F(2,113) = 9.38,P < 0.001]. Post hoc analysis of this latter effect (Duncan, P < 0 . 0 1 level) showedthat the I-group gave more indications of emotion than the other two groups.

High (i.e. happy) scores on the non-verbal emotion scale after listening to the storycorrelate negatively with more indications of emotion in recall of the story:r(118) =-0.26, (P<0.01) . If the measurement after testing is corrected for themeasurement beforehand, the (partial) correlation stays almost the same: -0.27{P< 0.005). A correction of the emotional impact of the reproduced story for its lengthalso produces a correlation of -0.27. Nevertheless, an examination of these threecorrelations for boys and girls separately shows a distinct difference between the sexes.For boys there was no significant correlation at all, but for girls the correlationamounted to r(58) = 0.31, P < 0 . 0 1 ; r(57) = 0.36, P < 0.005 (corrected scale score);and r(57) = 0.31, P<0.01 (corrected story score) respectively.

InterviewChildren's answers to the questions: "Did you find the story sad?" and "Did

it make you sad?" are given in Table 3, divided into the categories "yes" , " n o "

TABLE 3. DISTRIBUTION OF ANSWERS TO THE QUESTIONS: "DID YOU FIND THE STORY SAD?"

AND " W E R E YOU SAD YOURSELF AFTER LISTENING TO THE STORY?" AS A FUNCTION OF SEX AND

GROUP

Type ofQuestion

Story sad?

Sad yourself?

Answer

YesNoI don't knowYesNoI don't know

D-groupBoys

61220

146

Girls

10827

103

GroupI-group

Boys Girls

8 174 08 36 133 0

11 7

G-groupBoys

8662

126

Girls

1442

1208

362 MARK MEERUM TERW^OGT, JAN SCHENE AND PAUL L. HARRIS

and " I don't know". Answers that could not be scored unequivocally were includedin the " I don't know" category.

Summarizing across the results for both sexes, we find that the groups evaluatethe story as sad to varying degrees [x^(4) = 16.75, P< 0.005]. A further partitioningshows that the differences can be attributed mainly to the "no"-category [x (̂2) = 11.53,P< 0.005], which, particularly in the I-group, is quite empty, except for a few boys.Collapsing across the groups, we find that boys also give different answers from girls[X^(2) = 12.20, P< 0.005]. They denied that they thought the story was sad andclaimed that "they did not know" more frequently than the girls. The question ofwhether they themselves had also become a little sad resulted in exactly the same pattern,though with even more pronounced differences. Here too, a different answer patternbetween the groups was found [x^(4) = 25.50, P<O.OO1], with, as expected, theI-group reporting itself the saddest and the D-group the least sad. Here, too, thedifferences proved to be significant only within the "no"-category [x^(2) = 17.09,P < 0.001 ] . The answer patterns between boys and girls diverge even more clearlythan for the previous question [x^(2) = 24.26, P < 0.001]. This is particularlyevident when we look at the C-group, in which the story is listened to without furtherinstruction and thus without any warning that it will be a sad story. Here, no lessthan 60% of the boys as compared with 0% of the girls denied that they were affected.Even in the I-group, in which the children were asked in the instruction specificallyto "become sad", some of the boys stuck to their "no"-answer, while quite a fewchose the " I don't know-response.

Answers of the children from the I- and D-groups to the question "What did youdo when you listened so as 'to become/not to become sad? '" are categorized andpresented in Table 4. For most of the children the question proved rather difficultto answer. Over 40% could not give an answer at all and 20% claimed that nothingspecial was necessary in order to meet the instruction. Involvement, in particular,appeared to be a natural attitude (claimed by 25% of the boys and almost half ofthe girls). The remaining children formulated a clearly recognizable involvement ordetachment strategy. One girl in the I-group and two boys in the D-group even

TABLE 4. DISTRIBUTION OF ANSWERS TO THE QUESTIONS "WHAT DID YOU DO AS TO

BECOME SAD? (I-GROUP)/NOT TO BECOME SAD? (D-GROUP)", AS A FUNCTION OF SEX ANDGROUP

GroupBoys Girls Boys Girls

Don't knowNatural attitudeIt's just a story; not realWhat happened is not so terrible after allExpanding the story in a positive wayLinkage to similar (not so terrible

experience)Imagine that it is happening to oneselfImagine that it is happening to a friendLinkage to similar (terrible) experienceExpanding the story in a negative way

951

1

21

1

69

3

21

10263

1

9

712

1

SELF-CONTROL OF EMOTIONAL REACTIONS 363

mentioned two strategies. The strategies that were mentioned most were to remindoneself that it is just a story to obtain detachment and an identification with theprotagonist to obtain involvement. Only three children mentioned an inappropriatestrategy, i.e. a strategy that would be expected to work in the opposite direction.One of them was a D-group child who clearly realized her 'mistake' and describedher reaction as an involuntary association.

DISGUSSION

The results give a consistent picture showing that to a certain extent six-year-oldsare capable of manipulating their emotions on command. Children who were instructedto become sad with the help of the story (I-group) found the story sadder (emotionscale; interview) than children who had not been given these instructions. They didless well on the picture-memory task and they were more emotional in their recallof the story. Conversely, children who had been instructed not to let themselves beinfluenced by the emotionality of the story (D-group) were less sad, were less subjectto emotional interference in the picture-memory task and were less emotional in theirrecall of the story.

The measurable effect of the I-instruction indicates that the children could intensifytheir emotions. This phenomenon suggests that they have at their disposal an emotionalstrategy, for example a further elaboration of the situation. Admittedly, the answer"nothing special" was given often in the final interview, but the results of the C-group show that six-year-olds normally adopt a slightly detached attitude. When thesituation clearly encourages them to do so (I-group), they drop this restraint. Themeasurable effect of the D-instruction was also quite clear on the whole, though lessmarked. Compared with the C-group, these children showed themselves to be distinctlyless emotional. Nevertheless, as can be seen from the completeness with which thestory was reproduced, the instruction to listen carefully to the story was faithfullyobeyed by the D-group.

Nearly all dependent variables showed a systematic difference between boys andgirls. So far as the direct expression of emotion is concerned (interview, emotion scale,emotional elaboration of the story), we could attribute this to sex-role expectationsprohibiting boys from showing their sadness (Feshbach, 1978). However, there werealso systematic differences in performance on the picture-memory task. Under normalcircumstances, no differences are found between the sexes in this age group usingsimilar memory tasks (Flavell, Beach & Chinsky, 1966; Mathews & Fozard,1970; Ross & Youniss, 1969). Thus a more plausible explanation is that the accidentalchoice of story topic, namely the ending of a friendship, was more emotionally chargedfor girls than for boys. At this age, it appears that girls already make friends witha single person, whereas boys seek contact with a group (Elder & Hallihan, 1978;Maccoby & Jacklin, 1974). Friendships among girls seem to be more exclusive andintimate. Thus, given that identification with the protagonist depends on perceivedsimilarity between self and protagonist (Jose & Brewer, 1984), girls should identifymore readily with someone losing such a relationship, and would therefore be morelikely to feel sad as a result, particularly when instructions encourage such involvement.

364 MARK MEERUM TERWOGT, JAN SCHENE AND PAUL L. HARRIS

Harris et al. (1981) found that many six-year-olds claim that they are able, to acertain extent, to manipulate their emotions. The results of the present experimentconfirm their opinion. However, it is striking that the six-year-olds interviewed byHarris et al. (1981) almost exclusively mentioned changing the external situation:"If I do something else, my mood will also change". Mental strategies, which weremore often mentioned by 11- and 15-year-old children, did not occur among the 6-year-olds. In our experiment, the possibility of changing the concrete situation, e.g. bykeeping one's hands against one's ears, was not available to the children. Nevertheless,they proved capable of self-control, although it appeared from the interview that onlyabout 40% could report the mental strategies that they used to achieve this result.Once more it appears that children have skills in their repertoire of which they arenot yet aware, at least on a meta-cognitive level (Flavell, 1977; Harris & Olthof, 1982;Meerum Terwogt, 1984).

Among children who were able to report the use of self-control strategies, thereis a striking similarity with reports obtained from adults. Thus Koriat et al. (1972)found that the most frequently reported strategy among D-group subjects was to remindthemselves of the unreal nature of the stimulus material. This was also the mostfrequently mentioned strategy among D-group subjects in the present experiment.Similarly, Koriat et al. (1972) found that the most frequently reported strategy amongI-group subjects was an identification with the protagonist. Again, the same resultwas found among our six-year-old children. It is interesting to see that Koriat et al.(1972) marked these two strategies not only as the most frequently used but also asthe most effective strategies. Our results suggest, albeit tentatively, that young childrenare quite similar to adults in the strategies that they use. Thus, development mayconsist in a more conscious deployment of these strategies, rather than the acquisitionof qualitatively new strategies. It should be borne in mind, however, that the presentstudy, like that of Koriat et al. (1972), was concerned solely with the ability to complywith instructions to control emotional reactions to fictitious material. It remains tobe seen whether children have techniques for spontaneously controlling their emotionin the face of real events.

In conclusion, the present results indicate that the techniques developed for thestudy of self-control of emotion among adults can be used to study self-control amongchildren. These initial data suggest that techniques for involvement versus detachmentare available even at six years of age. Having identified the use of self-controltechniques among normal children, we may proceed to ask whether children sufferingfrom particular pathologies (e.g. impulsivity) differ from normal children in their abilityto use particular techniques (e.g. detachment). In the long term, such research mayhave implications for therapy [e.g. cognitive-behavioral interventions, as suggestedby Kendall and Finch (1979)] as well as diagnosis.

SUMMARY

Three groups of six-year-old children were instructed to listen to a sad story, insuch a way that (a) "you won't become sad yourself (detachment group); (b) "youfeel the sadness yourself (involvement group); or (c) without further instruction

SELF-CONTROL OF EMOTIONAL REACTIONS 365

(control group). Immediately afterwards, the effect of the instruction on two differentmemory tasks was tested. An unrelated picture-memory task showed that thedetachment group did better and the involvement group did worse than the controlgroup. However, when the subjects had to recall the story, all three groups were equallycomplete and accurate, but the involvement group included more emotional phrases.These results suggest that six-year-olds are capable of manipulating their emotionson command. This conclusion is supported by interview material and by the children'sself-report using a non-verbal emotion scale. Girls showed more instruction effectsthan boys. Possible explanations for this phenomenon were discussed.

REFERENGES

Averill, J. R. (1973). Personal control over aversive stimuli and its relationship to stress. PsychologicalBulletin, 80, 286-303.

Bower, G. H. (1981). Mood and memory. American Psychologist, 36, 129-148.Bower, G. H., Monteiro, K. P. & Gilligan, S. G. (1978). Emotional mood as a context of learning

and recall. Journal of Verbal Learning and Verbal Behavior, 17, 573-585.Easterbrook, J. A. (1959). The effect of emotion-on cue utilization and the organisation of behavior.

Psychological Review, 66, 183-201.Eder, D. & Hallihan, M. T. (1978). Sex differences in children's friendships. American Sociological Review,

43, 237-250.Feshbach, N. D. (1978). Studies of empathic behavior in children. In: B. A. Maher (Ed.), Progress in

experimental personality research (pp. 1-48). New York: Academic Press.Feshbach, N. D. & Roe, K. (1968). Empathy in six- and seven-year-olds. Child Development, 39, 133-145.Flavell, J. H. (1977). Cognitive development. Englewood Gliffs, NJ: Prentice Hall.Flavell, J. H., Beach, D. R. & Ghinsky, J. M. (1966). Spontaneous verbal rehearsal in memory tasks

as a function of age. Child Development, 37, 283-299.Folkman, S., Schaefer, G. & Lazarus, R. S. (1979). Gognitive processes as mediators of stress and

coping. In: V. Hamilton & D. M. Warburton (Eds), Human stress and cognition: an information-processingapproach (pp. 265-298). New York: Wiley.

Harris, P. L. & Olthof, T. (1982). The child's concept of emotion. In: G. Butterworth & P. Light(Eds), Social cognition, studies of the development of understanding {pY*. 188-209). Brighton: Harvester Press.

Harris, P. L., Olthof, T. & Meerum Terwogt, M. (1981). Ghildren's knowledge of emotion. yoMrna/of Child Psychology and Psychiatry, 22, 247-261.

Izard, G. E., Wehmer, G. M., Livsey, W. L. & Jennings, J. R. (1965). Affect awareness andperformance. In: S. S. Tomkins & G. E. Izard (Eds), Affect, cognition'and personality (pp. 2-41).London: Tavistock.

Jose, P.E. & Brewer, W. F. (1984). Development of story liking: character identification, suspenseand outcome resolution. Developmental Psychology, 20, 911-924.

Kendall, P. G. & Finch, A. J. (1979). Developing nonimpulsive behavior in children: cognitive-behavioralstrategies for self-control. In: P. G. Kendall & S. D. Hollon (Eds), Cognitive-behavioral interventions:theory, research and procedures (pp. 212-246). New York: Academic Press.

Koriat, A., Melkman, R., Averill, J. R. & Lazarus, R. S. (1972). The self-control of emotional reactionsto a stressful film. Journal of Personality, 40, 601-619.

Lazarus, R. S. (1966). Psychological stress and the coping process. New York: McGraw-Hill.Lazarus, R. S. (1968). Emotions and adaptation: conceptual and empirical relations. In: W. J. Arnold

(Ed.), Nebraska symposium on motivation (pp. 175-270). Lincoln: University of Nebraska Press.Lazarus, R. S.(1975). The self-regulation of emotion. In: L. Levi (Ed.), Emotions; their parameters and

measurement (pp. 47-67). New York: Raven Press.Lazarus, R. S., Averill, J. R. & Opton, E. M. (1970). Toward a cognitive theory of emotions. In:

M. Arnold (Ed.), Feelings and emotions (pp. 202-232). New York: Academic Press.

366 MARK MEERUM TERWOGT, JAN SCHENE AND PAUL L. HARRIS

Lazarus, R. S., Coyne, J . G. & Folkman, S. (1982). Gognition, emotion and motivation: the doctoringof Humpty-Dumpty. In: R. W. Neufeld (Ed.), Psychological stress and psychopathology {^p^p. 218-239).New York: McGraw-Hill.

Maccoby, E. E. & Jacklin, G. N. (1974). The psychology of sex differences. Stanford, GA: Stanford UniversityPress.

Mandler, J . M. & Johnson, N. S. (1977). Remembrance of things parsed: story structure and recall.Cognitive Psychology, 9, 111-151 .

Master, J. G., Barden, R. Gh. & Ford, M. E. (1979). Affective states, expressive behavior and learningin children, yoMrna/ of Personality and Social Psychology, 37, 380-390.

Mathews, M. E. & Fozard, J. L. (1970). Age differences in judgements of recency for short sequencesin pictures. Developmental Psychology, 3, 208-217.

Meerum Terwogt, M. (1984). Emotional development in middle childhood: a cognitive view. Doctoral dissertation.Free University, Amsterdam.

Ross, B. M. & Youniss, J. (1969). Ordering of nonverbal items in children's recognition memory.Journal of Experimental Child Psychology, 8, 20-32.

Saarni, G. (1979). Children's understanding of display rules for expressive behavior. DevelopmentalPsychology, 15, 424-429.