THE ARRIVAL OF A SIBLING Dunn1980

You might also like

Download as pdf or txt
Download as pdf or txt
You are on page 1of 15

/ Child Psychol. Psychiat . Vol. 21, pp. 119 lo 132. 0012 9630/80/0401 0119(02.

00/0
Ptigamor Press Lid. 1980. Printed In Greal Britain.
' Associaiion for Child Psychology and Psychiairy.

THE ARRIVAL OF A SIBLING: CHANGES IN PATTERNS OF


INTERACTION BETWEEN MOTHER AND
FIRST BORN CHILD
JUDY DUNN* and CAROL KENDRICK
M.R.C. Unit on the Development and Integration of Behaviour,
Madingley, Cambridge

INTRODUCTION
UNTIL RECENTLY studies of social development in young children have focussed
almost exclusively upon the relationship between mother and child. In the last three
or four years increasing interest has been shown in the role of fathers (Lamb, 1976),
but the importance of other family members in the child's development (either direct
or mediated through their influence on the mother) has been largely neglected in
systematic studies (but see Lewis and Rosenblum, 1978). There has also been little
attempt to examine the ways in which a young child's relationships with his family
change over time, or how they are affected by changing or difficult circumstances
other than separation.
This gap is important, for two connected reasons. Firstly, such changes as
alterations in the size, structure, social or economic circumstances, and pattern of
living of the family are very common in the experience of young children, and they
are thus plausible sources of significant stress for many children. Secondly, it is
known that individual children vary considerably in their responses to, and in their
vulnerability towards, potentially stressful circumstances. But as yet little is known
about the nature of such individual differences in vulnerability—a matter which is
potentially of considerable practical importance for child psychiatrists.
This paper is concerned with the changes in interaction between mother and first-
born from before to after the birth ofthe sibling. The importance ofthe arrival of a
sibling is much discussed by clinicians (Levy, 1934, 1937; Petty, 1953; Thomas and
Chess, 1977); but in spite ofthe significance which many clinicians have attached to
it, the impact of the event itself has been little studied since Levy's early studies.
There is as yet little systematic knowledge of the incidence and severity of
disturbances in behaviour in first-born children over this period, and little reliable
understanding ofthe long-term importance of such disturbance. It is often assumed
that maternal attention to the new baby must be of prime importance in the
aetiology of such disturbance. But since the arrival of a sibling is so often
accompanied by a constellation of changes for the first-born, and since no systematic
•Requests for reprints to: J. Dunn, M.R.C. Unit on the Development and integration of Behaviour.
Madingley, Cambridge CB3 8AA, U.K.
Accepted manuscript received 2 October 1979

119
120 JUDY DUNN AND CAROL KENDRICK

effort has yet been made to distinguish the causal impact of these changes from one
another, it-is at the moment quite unclear which of them are of major importance. In
this paper the general question of how the interaction between mother and first child
changes with the birth of a sibling is considered, in light of observations carried out
before and after the sibling birth.

METHODS
Procedures
Forty-one families were studied over a period from 1 to 3 months before the birth of a second child
until the second child was 14 months old. An outline ofthe structure ofthe study is given in Table 1.

TABLE 1. OUTLINE OF STUDY

2 pre-sib birth observations: child with mother and father


Mother pregnant with second child Interview
Rating of temperamental characteristics
Birth of second child , .
2 weeks post birth Post-sib-birth observation Child with mother + sibling
3 weeks post-birth Observation (inc. feed)
Interview (± father)
2nd child 8 months old 2 observations: siblings with mother (± father)
Interviews
Rating of temperamental characteristics
2nd child 14 months old 2 observations: siblings with mother (± father)
Interviews

In this paper the two observations carried out during the pregnancy are referred to as the pre-sib-
birth observations, and those carried out in the 2nd and 3rd week after the birth of the sibling as the
post-sib-birth observations. The first-born child is referred to as the child, and the 2nd child as the
baby, or the sibling.

TTie observations
The observations, carried out in the home, were unstructured. One hour of observation was made on
at least two occasions at each of the four age-points of the study: pre-sib-birth, post-sib-birth, when
younger sibling was 8 months, and when younger sibling was 14 months. The observations were car-
ried out by the authors; only one observer was present at each observation. In the pre-sib-birth visits
(made between 1 and 3 months before the expected date of delivery), for each family, the observer ar-
ranged to sample periods when the mother was busy with house-work, and periods when she was more
relaxed. In the majority of cases this meant that the visits were made during the morning. In the post-
sib-birth visits, the observation included at least one session when the mother was feeding the baby in
the presence of the first-born.
The observations were made in the form of a running record, in a lined notebook with each line
representing 10 seconds. A time marker was provided by an electronic bleeper. Categories of child
behaviour, mother behaviour, and of interaction were pre-coded and are given in the Appendix. The
recording technique allowed the observer to record in narrative form details of the child's play, and the
objects played with, as well as further details of control incidents, or behaviour towards the new baby.
During Ihe observations, the verbal exchange between child and family members, and comments made
by the child as he played ("self-talk") were recorded on a portable stereo tape-recorder. Immediately
after the observation the observer transcribed the tape-recording.
The presence of an observer
It is inevitable that the presence of an observer will affect the interaction of family members. One
likely effect is that the parents' behaviour will be modified in a direction which they believe will be
THE ARRIVAL OF A SIBLING 121

socially approved. Patterson (1974) quotes studies showing that in normal families with older chiidren
parents were, on instruction, able to make their children "look good". However our experience with
observing families with children aged between 18 months and 3 years in this and other studies (Dunn
and Wooding, 1977), and in previous studies of mothers and infants (Richards and Bernal, 1972)
indicates that the child's behaviour, while it may initially involve friendly overtures to the observer, does
not become particulariy conforming or obedient. (See also Lytton and Zwimer (1975) for comparable
experience.)
In the great majority of families where the mother has agTeed to take part in a longitudinal study the
observer is treated in a very relaxed way; the mother is too busy with housework and children to change
her habitual style of coping in order to devote unusual attention to the children during the many hours
that the observer is in the home.
In the present study, to minimise the intrusive effect ofthe observer's presence, the observer paid at
least one visit to the family before conducting the first observation, and did not begin recording until at
least 10 minutes after her arrival for the observation visit. A premium was put on ensuring that a
rappon was established between observer and family members. The advantages and disadvantages of
home observation are funher discussed in Richards and Bernal (1972).

The population studied


The families taking part in the study were collected through the help of General Practionen and
Health Visitors. The families were largely working class; the occupation of the father, classified
according to the Registrar General, was IV for 7 families. III manual for 12 families. III white collar
for 14 families, and II for 8 families. Thineen of the mothers delivered the second baby at home.
There were 22 first-hom boys and 19 first bom girls. The age range of the children was from 18 to 43
months at the birth of the sibling (median age 25 months).

Reliabilities
The stability of individual differences on the observation measures of mother and child behaviour
described and analysed in this paper was assessed by comparing the rank orders on these measures, on
two observations 1-2 weeks apart, for 28 mother-child pairs. The Spearman rank correlations ranged
from 0.64 to 0.86. AU correlations but two were significant above the 1 % level; the remaining two were
significant at the 5% level.
Interobserver reliability was assessed by comparing the records of the two observers made during
home observations when both observers were present (range of ratio of agreement to agreement pltis
disagreement: 0.57 to 1.00, mean = 0.84).
Interobserver agreement on the coding of conversational episodes, from the transcripts of the tape-
recordings of verbal interaction ranged from 0.85 to 1.00 (agreement/agreement plus disagreement).
The median ratio of agreement to agreement plus disagreement was 0;90. mean 0.91.

RESULTS
This paper is concerned with five interrelated questions. First, were there changes
in the frequency of occurrence of measures of mother and child hehaviour from the
pre-sib-birth to the post sih-birth observations? Had the mothers in the sample
changed in their behaviour towards the first-born child, for instance? (Section 1).
Second, what were the relations between the various measures of mother and child
behaviour in the pre-sih-birth observations? (Section 2). Third, what were the
relations between these measures in the post-sib-birth observations? Had the
pattems of association changed since the arrival of the sibling? (Section 3). Fourth,
were the individual differences between mothers, and between children, consistent
from the pre-sib-birth to the post-sib-birth observations? That is, did the rank
orders of mothers and children remain consistent on the different measures of be-
haviour? (Section 4). Fifth, did the time interval between the pre-sih-hirth and post-sib-
hirth observations contribute significantly to changes in interactions? (Section 5).
122 JUDY DUNN AND CAROL KENDRICK

(1) Changes in frequency of occurrence of measures of mother and child behaviour


from pre-birth to post-birth observations
(a) Observational measures. The Wilcoxon T test was used to assess how
measures of maternal and child behaviour changed from the pre-sib-birth to the
post-sib-birth observations. The changes in 35 measures were examined. There were
decreases in several measures which reflected maternal attention to the child. The
measures of time spent in joint play, the time for which the child was held by the
mother, maternal affectionate contact, matemal giving, showing or pointing out
objects, helping the child, or making suggestions which highlighted features of
whatever was the current focus of the child's attention all decreased by more than
24% of the mean pre-sih-birth observation level. The decreases in 3 of these measures
reached a 5% level of significance (see Table 2).

TABLE 2. SOME CHANGES FROM PRESIBBIRTH TO PDSTSIB BIRTH OBSERVATIONS

Wilcoxon T A^ = 41
Observation measures Pre-sib-birth Post-sib-birth P level

Mother helps 9.5 5.5 0.01


Mother shows 16.0 9.5 0.01
Mother suggests (highlights) ll.O 8.0 0.05
Mother prohibits 15.0 21.0 0.01
Mutual look (negative) 10.0 16.0 0.01
Time in confrontation 19.0 34.0 0.04
(Median no. 10-second units/1000.)

Of the 6 measures refiecting confrontation between mother and child, and


maternal efforts to control the child which had been included in this analysis: 4
increased in frequency, 3 significantly (Table 2), (the remaining 2 control measures
occurred very rarely). For some children there were marked increases in measures of
the child wandering, sitting without playing, fussing, making verbal demands, and
using comfort objects, but these did not reach a 5% level of significance for the
sample as a whole.
There was also a change in the balance of responsibility for starting joint
activities; Table 3 shows that there was a decrease in the proportion of joint play and
joint activity started by the mother.

(b) Changes in the initiation and subject of conversation. Changes in the initia-
tion and subject of conversational episodes were next examined, using the
transcripts from the observations. First, changes in the topic of conversation were
studied. Episodes of conversation were coded as concerned with control, play, books
or other topics (see Wootton, 1974). There was a significant increase in the number
of control episodes in the conversation (Table 4). There was no significant change
in the frequencies of episodes categorised as focussed on play, books, or other topics.
Analysis of the changes in nine categories of the initiation of verbal exchange
showed that there were significant changes in five of these: there was a significant
increase in the frequency of verbal exchange started by the mother prohibiting the
child (Table 4), and significant decreases in the frequencies of three of the sub-
THE ARRIVAL OF A SIBLING 123

TABLE 3. CHANGES IN BOUT MEASURES FROM PRE SIB BIRTH TO POST SIBBIRTH OBSERVATIONS

Wilcoxon T/V - 41
Pre-sib birth Post sib-birth
Bout measures observations ohservations P level
No. mother-initiated bouts
play/1000 10 sec units U.4 6.40 0.01
Ratio child-initiated bouts play
mother-initiated -(- child-
initiated bouts play 0.48 0.59 0.05
Ratio child-initiated bouts joint
attention
mother-initiated + child-
initiated bouts joint
attention 0.57 0.64 0.10

categories of mother's positive initiation of interaction (Table 4 (d) (e) (f)), and in
the percentage of verbal interactions which were started hy the mother's positive
comments.

TABLE 4. CHANGES IN VERBAL INTERACTION FROM PRESIB BIRTH TO POST SIB.BIRTH OBSERVATIONS

Wilcoxon T/V ^ 41
Verbal interaction Pre-sib-birth Post sib-birth P level
(a) No. of control episodes/100 min mother present (median) 8.0 10.0 0.04
(b) Negative verbal interactions initiated by mother/100
min mother present (median) 7.5 12.5 0.02
(c) Verhal interactions initiated by child/100 min mother
present (median) 17,0 S7.0 0.01
(d) Mother unsolicited positive comment on child action/
100 min mother present (median) 10.3 8.0 0.10
(e) Mother new suggestion child activity/100 min mother
present (median) 11.3 8.3 0.02
(f) Mother initiate verbal game/fantasy/100 min mother
present (median) 1.6 0.6 0.03
(g) Total positive verbal interactions initiated by mother
(as % verbal interactions) 35.1 30.1 0.05

The analysis of the conversational exchange illustrates again the alteration in


the balance of responsihility for starting interaction. Table 4 shows that the number
of verbal interactions started by the child increased while the number of positive
interactions started by the mother decreased. The number of negative initiations by
the mother increased, and the ratio of positive to negative initiations by the mother
decreased. (Signtest2 = 2.70, N = 41, P < 0 . 0 1 ) .

(2) Correlations between measures of mother and child behaviour in the pre-sib-birth
observations
T h e Spearman rank correlations between the measures of mother and child
behaviour formed a number of groups. Table 5a gives the correlation matrix for
124 JUDY DUNN AND CAROL KENDRICK

'3

SBnj'3 o l 00 • - <N N
o dddd

IIS-3
o oooo o

lO N O ^ ^- ^- O
d ddddd d

o o o o

,-, [Jl — OO — O
Ul SUIIJ_
oooooooooo

(MIN—- — l N C J O i n T t . . 9 . > O l M
i>. 31 O ^ SO SO OO "-^ ^ W O ^^
dddddddddddd

O TT O O O

fl. — , - . • * 0-1
oo o w o o

o o o o o o o o o o o o o o o o

»J CO
ie\d luiof
O O O O O oooooooooo dd

la o « n .e
S
CJ
E
<u
I,
.5 c . c T)
ohi

— s "ra
u 0
IS (^ I .a 1 1 a j=
b-
c •« 0-
a6 's
H S sE S >
H U u u u uu u
>
THE ARRIVAL OF A SIBLING 125

the 18 measures examined in this analysis. These 18 variables were selected either
on the basis of their importance as indices of those features of the mother-child
interaction which showed marked changes after the sibling birth (such as the
increase in confrontation), or for their intrinsic interest as aspects of the mother-
child relationship (such as the measure of affectionate tactile contact). The correla-
tions are represented diagrammatically (using a selection of the measures) in Fig. la.
(a) Pra-birth obsgrvotion .j . *
' ' —* M initiate witn
^ — " ^ Z prohibit •" -~ ^
. - ^ / / ^ ^ ' ^ . ^ ^ ^ ^ - M shew

^ \ V^-r--^ - ^ < - - / - \ - * Join

wander— — '^'.i
, JTcTiifd
Child fuss

(b) Post b-rth observation . , ^_^


' M initiate with

M suggest
(highlight)
wander ~ — _ __
Child initiate
Child luss

Positive correlotion
Negative correlation

FIG. 1. Correlations between measures of mother and child behaviour, before and after the binh of a
sibling.

It can be seen that there are significant positive correlations between measures of
maternal showing, making highlighting suggestions, and joint play between mother
and child, and that these measures were negatively correlated with a group of
measures reflecting maternal attempts to control the child. These "control" variables
were all closely related; the time spent in confrontation, verbal prohibitions, and
the proportion of verbal interactions which were started by the mother prohibiting
the child's actions.
The group of measures reflecting "control" was positively related to the amount of
fussing, wandering, sitting without playing, and to the incidence of the child
looking at the mother without the mother looking at the child.
The measures of physical affectionate contact by mother and by child were
correlated with each other, and with the time the child was held, but these measures
were not significantly related to the groups of attentive or play measures, nor
negatively related to the group of prohibiting and confrontation measures.
(3) Post-sib-birth observations
(a) Changes in correlations of mother and child behaviour. The Spearman rank
correlations between measures from the post sib-birth observations were next
126 JUDY DUNN AND CAROL KENDRrCK

sn 04 i c
IB j((>0] ' 3 — •»• S

IMfqo poeuup |e -3 q q—
d d d
?
oJ SS
Ol
a
d d CT" d d

on
o 94
lie 3 d d d d dd

— o ::
OO
-3 o
od d d d d d d

sieniut -3 O - -SI

\d pueuup |E
CT CT CT CT CT

o —

^- ^ ^
o
CT CT
?o o o a o
-0


88"

(O 04 a OO 04 OJ 00
o n CT — a
•n
e CT CT o o O
o o o
o
? ?
48" -0

s
19
06

09

06
14
30
11

CO
on a

~ o — — O— o
O
T CT CT CT
f 1" ? ?
-0
-0

.
OC o sn >n
00 O l O O CT
S
04 CT OJ OJ 04
o o CT o o o CT CT CT o o CT CT CT o

*
50"

60"

CT V CT DI go
luiof o O CT CT CT 2
CT on
CT CT CT O CT CT CT O CT O o
f •^
r-
18

oe O I- 04
luiof 04 -r sT T IC CT to o
o
o
-s V

11 I •:

i^-^
I
.5.E
uuuuuuu Su
THE ARRIVAL OF A SIBLING 127

examined; the relationships between the same 18 variahles studied in the pre-sib-
birth ohservations (reported in Section 2) were analysed (Tahle 5b).
There were some marked changes from the pre-sib birth ohservations in the
relationships between some of the measures. Among the most pronounced were
alterations in the patterns of association of some of the measures of child behaviour,
notably wandering, sitting and fussing, and verbal demands for objects (Fig. Ib): in
the pre-sih hirth observations for example a high frequency of wandering was
associated with a low frequency of joint play, a low frequency of maternal showing
and suggesting (highlighting), a high frequency of prohibition and confrontation,
and a high frequency of the child looking at the mother without the gaze heing
returned. In the post-sib-birth observations this pattern had changed. In the pre-
sih-birth observations then, families where there was much confrontation and
prohihition tended to have children who spent a relatively high proportion of
time wandering in a unfocussed way, in marked contrast with the children
of the relatively playful, permissive mothers. However after the sihling birth
the changes in the correlations reflect the fact that some children with relatively
playful and permissive mothers spent a relatively high proportion of their time
wandering around aimlessly or sitting without playing. These changes in the
tions. For all the variables except joint play (post-sib-birth and post-sih-birth observa-
correlation matrix were examined further by testing the significance of the changes
in correlation between pairs of measures.
(b) Changes in correlations between pairs of measures. The changes in the
correlation matrix suggested that the following correlations changed markedly after
the sibling birth, and were of particular interest:
(1) Child wander —mother prohibit
(2) Child wander — mother show
(3) Child wander—joint play
(4) Child sit —mother show
(5) Child sit—joint attention
(6) Child sit — child look M. without M. look child
(7) Child verbal demand object —child look M. without M. look child
(8) Child verbal demand object — Mother prohihit
(9) Child demand play—joint attention.
Since the tests for the significance of changes in correlations are parametric, and
assume normal distribution of the variables, the Koimogoroff-Smirnoff Test of
Goodness of Fit to Normal Distribution was first carried out for the variahles in the
correlations listed above (1-9), from both pre-sib-birth and post-sib-birth observa-
tions. For all the variables except joint play (post-sib-birth obs.) the Koimogoroff-
Smirnoff test showed that the data did not deviate significantly from a normal
distrihution. The changes in correlations (1) to (9) above were tested using Fisher's
Test (Fisher, 1970, p. 205) (omitting correlation (3) which involved the joint play
post-sib-birth measures). Table 6 shows the significance levels of the changes in the
correlations.
These changes in correlations could have resulted from a major change in the
variance of the measures of mother and child behaviour after the sibling hirth. The
change in variance of the following measures was tested: Mother show, prohibit.
128 JUDY DUNN AND CAROL KENDRICK

joint attention, joint play, confrontation, child wander, sit, verhal demand object,
verbal demand play, child look mother without mother look child (Ferguson, 1966,
p. 183). No significant differences were found hetween the variance on these
behaviour measures from pre-sib-birth to post-sib-birth observations, with the
exception of the measure Mother show, where there was a significant decrease in
variance in the post-sib-birth ohservations.
TABLE 6. SIGNIFICANCE OF CHANGES IN CORRELATIONS BETWEEN PAIRS
OF VARIABLES AFTER SIBLING BIRTH

Value of
Correlated variables z, — z-, P value

n - 3

C. wander and M. prohibit 2.06 0.02


C. wander and M. show 1.29 0.10
C. sit and M. show 2.92 0.002
C. sit and joint attention 1.61 0.06
C. sit andC. look M, without M. lookC. 1.80 0.04
C. verbal demand object and C. look M.
without M. look C. 1.59 0.06
C. verbal demand object and M. prohibit 2.62 0.004
C. verbal demand play and joint attention 2.02 0.02

(4) Consistencies and changes in rank orders from pre-sib-birth to post-sib-birth


observations
The analyses described in sections (2) and (3) of this paper showed marked
changes in the patterns of association between measures of maternal and child
behaviour. These altered patterns of association could have heen produced by
changes in individual differences in maternal behaviour alone, or in child behaviour
alone, or in both together. Spearman rank order correlations were used to assess
the extent to which measures of mother and child behaviour, though changing in
value (section 1), retained rank order consistency.
Tahle 7 shows that in spite of the changes in the measures of maternal attention
and prohihition described in section 1 the rank orders on these maternal measures
remained significantly similar. The rank orders on measures of child fuss, wander,
make verbal demand for object, sit without playing, however did not remain
significantly similar.
It is possible that age differences, or differences in the conversational skills of the
children in the sample were contributing to the correlations from pre- to post-sib-
birth observations in the conversational measures shown in the lower section of
Tahle 7. Since there was only a low positive correlation with age for the measure
of turns per hour (rs = 0.29, P < 0.10), it seems more probable that the con-
versational ahility of the children was of prime importance.
(5) Change in interaction pattems as a function of time between pre-sib-birth
observations and post-sib-birth observations
It is possible that the changes in interaction patterns between the pre-sib-birth
and post-sib-binh observations reflect developmental changes in interaction between
THE ARRIVAL OF A SIBLING 129

TABLE 7. SPEARMAN RANK CORRELATIONS ON MEASURES FROM PRE SIB-


BIRTH AND POSI SIB BIRTH oBSERVATiONs(A' = 41)

rs P level

Joint attention 0.75 0.001


Joint play 0.38 0.02
Time being held 0.55 0.04
M. shows 0.60 0.001
M. suggests (highlights) 0.61 0.001
M. prohibits 0.59 0.001
Time in confrontation 0.55 0.001
C. fuss 0.19
C. wander 0.07
C. sitting without playing 0.28
C. verbal demand for object 0.14
C. verbal demand for play 0.44 0.005
C. initiate verbal exchange 0.51 0.001
C. look M. without return of gaze 0.36 0.02
C, uses comfort object 0.51 0.05
Time close to M. 0.30
Mutual look (positive) 0.55 0.04
Mutual look (negative) 0.45 0.005
Measure from conversational analysis
No. of conversational turns/hour 0.85 0.001
No. of control episodes/hour 0.55 O.OOi
No. of episodes concerned with play, hook.
or other topic/hour 0.42 0.01
Ratio of control episodes/play, hook and
other episodes 0.59 0.001

mother and child, rather than the impact of the events surrounding the birth of the
sibling. It is also possible that the changes from pre-sib-birth to post-sib-birth
observations were influenced by the stage of the mother's pregnancy at the pre-sib-
birth visits.
We do not have sufficient data on a control group of families without a second
child to examine the first possibility; however, the sample did include 20 families
where there had been a very short time period between the pre-sib-birth and post-
sib-birth observations, namely less than two months. The data for these 20 families
were analysed, with Spearman rank correlations as reported in sections (1) and (2)
above. The results were without exception similar in direction, and the degree of
significance of the changes was at least as strong. This does not rule out the pos-
sibility that the changes from the pre-birth to the post-birth observations are
associated with age-changes, but it seems highly unlikely that such marked changes
in behaviour and interaction could he strongly, or solely, related to an increase in
age of less than 2 months.
DISCUSSION
Most children in the sample studied here experienced a decrease in maternal
playful attention with the arrival of the sibling. This decrease in maternal attention
was particularly evident in those measures which reflect the more subtle aspects
of sensitivity to the child's interests. For instance there was a decrease in highlighting
130 JUDY DUNN AND CAROL KENDRICK

suggestions, which are suggestions in which the mother comments on what is already
the focus of her child's attention; these often reflect a mother's interest and sensitivity
in entering her child's world. A second example was the decrease in the measures
of the mother's initiating conversations with verbal games or new suggestions for the
child's activity. So although the rank orders on the frequency with which mothers
made comments which were subtly tuned to their child's interests remained fairly
stable, there was for the majority of children much less of this sensitive attention
after the sibling was born. There was also a considerable rise in the negative
confrontation between mother and child: control issues began to dominate the
conversation to a much greater extent.
The relative rank order stability of individual differences in measures of maternal
behaviour, and the lack of stability in individual differences in child behaviour
section 4 suggests that the changes in pattems of association described in section
3 are primarily the result of changes in individual differences in child behaviour
after the birth of the sibling. For example, while unfocussed wandering, and sitting
without playing were in the pre-sib-birth period associated with a mother-child
interaction characterised hy a high level of maternal prohihition, and low level of
maternal attentiveness, in the post-sib-birth period some children who bad previously
spent little time wandering or sitting had now greatly increased the proportion of
time spent in this way; but their mothers were still, in relation to the other mothers,
relatively attentive, permissive and playful.
The changes in child behaviour may well of course be a response to the changes
in maternal behaviour whicb have been described for the sample as a whole (section
1). That is, the children may have been responding differently to the decrease in
maternal attention that generally occurred for most of the sample.
A general change in the balance of responsibility for starting interaction between
mother and child was noted in conversation between mother and child, in play, and
in attention to a common focus. In most families the children were, after the
sibling was born, responsible for a greater proportion of such initiation, and the
mothers for a correspondingly smaller proportion. This change links in an important
way with the individual differences in children's behaviour over this period. We
have reported elsewhere the results from the post-sib interview which show that
some children respond to the events surrounding the birth by withdrawing (Dunn
and Kendrick, in press). Now in these families, the child did not show the increase
in responsibility for, and in the frequency of initiating the exchange that occurred in
most families. If, as happened in some of these families, the mother did not increase her
own part in starting interaction but followed the general pattern of initiating
exchange less frequently, then the level of communicative exchange between
mother and child decreased very markedly.
Three further points should be noted. First, the individual differences in changes
in child behaviour may weli not be related directly to the particular changes in
maternal behaviour which are documented here. It is clear from the results of the
post-sib-birth interview, that there is a constellation of changes in the child's life
which may be implicated. However, some of the changes in child behaviour are
closely related to changes in maternal behaviour. For instance the increase in
verbal demands is related to an increase in prohibitions.
THE ARRIVAL OF A SIBLING 131

Second, while there are not detailed systematic studies of mother-child interaction
in the home environment with which these findings can be compared, it should be
noted that in a laboratory study of 8 mother-child pairs before and after the birth
of a sihling, Taylor and Kogan (1973) reported a consistent decrease in "warmth"
on the part of both mother and first born child, and Trause et al. (in press) noted
a significant increase in the use of angry and stern comments by mothers at home
after a period in hospital for the birth of a second child.
Third, the extent to which the changes in the interaction between mother and
child after the sibling birth are infiuenced directly by tbe mother's caregiving
activity with the new baby are examined in Kendrick and Dunn (in press).

SUMMARY
Interaction between mother and first-born child before and after the birth of a
second child was studied in 41 families, using home observations and interview
techniques. Decreases in maternal attention and play, increases in confrontation,
and changes in the balance of responsibility for initiating interaction were found.
Changes in pattems of association between maternal and child variables after the
birth of the sibling were primarily the result of changes in individual differences in
child behaviour.

Acknowtedgements - This work is supported by the Medical Research Council. We are very grateful to
the families in the sample for their generous help, to the Health Visitors of the Cambridge Area,
and to Robert Hinde and John Dunn for their helpful comments on the manuscript. We would like to
thank Pat Altham, Rosanne McNamee and Anne Campbell for statistical advice. We are also grateful
to Jeremy Cherfas, Irene Hudson and Christine Outhwaite for their help in the analysis.

REFERENCES
DUNN, J. and KENDRICK, C . The reaction of first born children to the hirth of a sibling: mother's
reports./. Child Psychol. Psychiat. (in press).
DUNN, J. and WOODING, C. (1977) Play in the home and its implications for learning. In The Biology
of Play (Edited hy TiZARD, B. and HARVEY, D . ) . S.I.M.P./Heinemann Medical Books, London.
FERGUSON, G. A. (1966) Statistical Analysis in Psychology and Education. McGraw-Hill, London.
FISHER, R. (1970) Statistical Methods for Research Workers, 14th Ed. Oliver & Boyd, Edinburgh.
KENDRICK, C . and DUNN, J. Caring for a second haby: effects on interaction between mother and
first child. Devi Psychol. (in press).
LAMB, M. E. (Ed.) (1976) The Role of the Father in Child Development. ]ohn Wiley. New York.
LEVY, D . M. (1934) Rivalry between children in the same family. Child Study 11,233-261.
LEVY, D . M. (1937) Studies in sibling rivalry. Amer. Orthopsychiatr. Ass. Res. Monogr, No. 2.
LEWIS, M . and ROSENBLUM, L . (1978) The Child and Its Family. Plenum Press, New York.
LYTTON, H . and ZWIRNER, W . (1975) Compliance and its controlling stimuli observed in a natural
setting. Devi Psychol. 11, 769-779.
PATTERSON, G. R . (1974) A basis for identifying stimuli which control behaviour in natural settings.
Child Dev. 45,900-911.
PETTY, T . A. (1953) The tragedy of Humpty Dumpty. The Psychoanalytic Study of the Child 8,
404-412.
RICHARDS, M . P. M. and BERNAL, J. F. (1972) An observational study of mother-infant interaction.
In Ethnological Studies of Child Behaviour (Edited by BLURTON-JONES, N . ) . Cambridge
University Press, Cambridge.
TAYLOR, M. K. and KOGAN, K. L . (1973) Effects of birth of a sibling on mother-child interactions.
Child Psychiat. Hum. Dev. 4 (1), 53-58.
132 JUDY DUNN AND CAROL KENDRICK

THOMAS, A. and CHESS, S. (1977) Temperament and Development. Brunner Mazel, New York.
TRAUSE, M. A., BOSLETT, M . , VOOS, D . , RUDD, C , KLAUS, M . and KENNELL, J. A hirth in the
hospital: the effect on the sihling. (Submitted for publication.)
WOOTTON, A. J. (1974) Talk in the homes of young children. Sociology 8, 277.

APPENDIX
Variables from pre sib- birth and post sib- birth observations
(Detailed definitions of observational categories can be obtained from authors.)
Joint attention (total)
bouts initiated hy mother, bouts initiated by child
Time held (total)
bouts initiated by mother, bouts initiated by child
Time close to mother (total)
bouts initiated hy mother, houts initiated by child
Time in joint play (total)
bouts initiated by mother, houts initiated by child
Mother caretakes
Mother helps
Mother shows
Mother suggests: highlighting, directing, redirecting
Mother affectionate physical contact
Mother prohibit
Mother restrain
Mother punish
Time in confrontation
Mutual look positive
Mutual look negative
Child look mother, without M. look C.
Child verbal demand (total)
Child verbal demand for play, attention
Child verbal demand for object
Child wander
Child sit without playing
Child fuss or cry
Child give to mother
Child affectionate physical contact
Child comply with mother suggestions
Child uses comfort object
Time not interacting
Mother initiate interaction with prohihition (as % of verbal initiations)
Child initiate verbal interaction (as % of verhal initiations)
Mother initiate verbal interaction, with positive orneutral comment (as % of verbal initiations)

You might also like