Child
Abuse
and
Neylect,
Vol.
I, pp. 315-320.
Preliminary
Pergamo”
Press,
1977.
Study of Mother/Infant
Printed
in Great
Britain
Interaction
C.A. Hyman National Advisory Centre for the Battered Child, NSPCC and University of Surrey, England
INTRODUCTION
I am going to speak today about a videod observational study which we are carrying out at the NSPCC’s National Advisory Centre on the Battered Child. This study forms one of a number of different research projects which we have been carrying on at the National Advisory Centre for the past five years. One finding in particular emerged from the comparative use of the Bene Anthony Test of family relations as reported in 'At Risk' NSPCC Research Team (1976). This suggested that the pre-school battered child's perception of the mother is distorted. This small sample study of 13 matched pairs of children between the ages of three and five years of age showed that the battered children denied both positive and negative feelings towards and emanating from the mother. This in turn raised the question of the nature of the battered child's attachment to the mother.
PROCEDURE It was therefore decided to attempt a direct observational study of mother/child interaction, matching infants between six and twenty months, who had been referred to the NAC, with nonbattered children from similar home backgrounds. Thus, so far 10 children have been pair matched for age and ordinal position and their parents have been matched for age, ethnic group and type of living accommodation. Both experimental and control families were asked to co-operate in a session involving video recording although they were not actually told at what point the camera was to be switched on. We always show the mothers the video sequences after the session and would, if asked, wipe the tape.
METHOD
The observational set up is similar to Ainsworth's as reported in her "Individual differences in one year olds" (1971). The length of the strange situation has been cut to eight and a half minutes and the situation modified to exclude the leaving of the infant in total isolation for any length of time. It was felt that Social Workers would find a possibly difficult situation involving their clients even less acceptable if the trauma to the child were
315
C.A. Hyman
316
thought
to be so great.
The sequence Situation
thus consists
of:-
A
Mother and child are left together in a strange room without anyone else being present. They are told someone will come shortly. Situation
B
This lasts for thirty seconds during which person who is a stranger to them. Situation
they are joined by a
C
This lasts three minutes. The mother is invited into an adjacent room to talk to a second stranger, while the child is left in the first room with the original stranger. Situation
D
This is the final situation. The mother returns (without the second stranger) and joins her child and the first stranger. This lasts two minutes. The child, thus, stays in a high chair in the same room all the time and both strangers so called, are in fact psychologists working at the centre.
VERB ANALYSIS We are, like Ainsworth, who originated the technique, aiming to identify the type of attachment shown by the child to her mother. One approach is a simple frequency Two approaches are possible. analysis of the many behaviours emitted by both mother and infant during the eight and one half minute sequence. To do this a three second interval analysis has been adopted and a complete catalogue of all behaviours possible has been drawn up. The separate behaviours are then summed, using PRIMATE computer package (Humphreys, 19731, to see for example how much crying actually occurs for each child in each situation, or how much looking at the mother occurs in each situation, etc. Simple behaviours are subsequently grouped together under such headings as distress, Ainsworth found that the amount of greeting, manipulation, etc. greeting behaviour evinced by her sample of children in the reunion situation with the mother was her best single index of attachment. So far this has not emerged in our sample, i.e. there have been no apparent differe=s in greeting behaviour between the 2 groups, but there is a distinct tendency (though not reaching statistical significance) for more of our battered children to cry on reunion and to cry for longer than is true for the matched controls as shown in Tables 1 and 2 respectively.
Preliminary
TABLE
Battered
1
Study
Number
of Mother/Infant
of Children
Control
Children
Interaction
Who Cried on Reunion
Children
5
2
cried
5
8
did not cry
Fisher TABLE
2
Exact Test
Mean Number
317
p
.05. N.S.
of Intervals
in which Crying
Occurred Battered
Control
Children
Children 3
10.4
Mann Whitney
Test
u = 30
N.S.
In Ainsworth's group this was one identifying feature of a small sub-sample of what she called ambivalently attached children. If this were repeated in a larger sample it might be one confirmation of our suspicion of early malfunctioning in the battered child's early attachment, since the normal child may be expected to greet his mother in a positive way after such a short separation. INTERACTION
ANALYSIS
A
second way to analyse the data is to examine the type of interactions which occur between mother and infant. This has been a more fruitful way of analysing our material. This is both a qualitative type of analysis as well as a quantitative one in that the analysis looks at interactions of all kinds, whether they It is involve speech, visual regard or other sorts of behaviour. possible to contrast the mutual interactions, in which mother responds to child or vice versa, with other types of interaction which are non-reciprocal. When this is done we find that in Situation D the proportion of reciprocal interactions increases for the normals, while for the patients the number of nonreciprocal interactions predominates. These differencesare significant at the .05 level. This means that in behavioural terms the non-battering mothers were more likely to give an appropriate response to their infants than was true for the battering mothers.
By appropriate I mean that the mother reciprocates what the child has initiated. Inappropriate responses in this context often involve the mother in an interruption of a self sufficient activity of the child's which requires no response from her, as when the child is playing happily with a toy which the mother then takes
C.A. Hyman
318
Although the battering mothers from him to start up another game. actually appear to direct more behaviour overall toward the child, often it is inappropriate in the way just described. Thus the sequences of behaviour in the battering dyads tend to be more disjointed, with fewer maintained interchanges than we see in the normal dyads. A further finding which underlines what has already been said, is the greater degree of mis-match between battering and normal Thus the correlation between mothers' mother-child interactions. and babies' types of interaction is lower when battering mothers are compared with their babies than when non-battering mothers are compared with their babies.
TABLE Types Battering
3 Correlation of Interaction
Dyads
Between
Control
Mothers'
and Infants'
Dyads
-37
.66
General
.75
.92
Mutual
interactions interactions
Such a finding implies that normal mothers and babies respond to each other's signals at about the same rate while battering mothers and children both send out a higher proportion of behaviours to which the other member of the dyad does not respond. Because these calculations are based on a very short sample of behaviour within Situation D, the difference between the correlations do not quite reach significance. We shall be analysing the complete protocols of all twenty tapes to see whether the full time sample of behaviour confirms this trend.
DISCUSSION In making all these distinctions I should like to emphasise that so far we are at the beginning of a pilot study. Not only have we only just begun the formal analysis of tapes and computer processing, but also our sample is very small to date. We are therefore looking at no more than trends. We are in fact half-way through the first year of a two year grant from the Department of Health and Social Security to enable us to undertake this pilot study. If we can confirm the trends already emerging and show other differentiating signs based on this type of study, this will justify our carrying out a longer and larger project which would in the longer term enable us to discover manifestation of malfunctioning in mother-infant interaction before actual physical assault has occurred.
preliminary Study of Mother/Infant Interaction
319
DEMONSTRATION
demonstration tape which I am going to show consists of the last part of Situation A when mother and child are alone; the whole of the following Situation B when they are joined by the first stranger; then the beginning of Situation C when the There is then a break followed by an mother leaves the room. excerpt from the final reunion episode. First you will see a battering dyad then their matched pair. The whole tape runs for 5 minutes. Of particular interest is the difference in the two children's reaction to the mother's re-entry. The
DESCRIPTION
First tape M, experimental Situation
A
Child very distressed Situation Child
Child
quietens.
on stranger's
entry.
C
Child very distressed Situation
on entry, gradually
B
very distressed
Situation
case:-
on mother's
departure.
D
cries on reunion.
Referral
Notes
Mother self-referred Greek Cypriot. Father had shown very rough handling with child. Father eight years younger than mother. Poor housing, shared with mother's in laws. Six months treatment. Re-housed; nursery for child. Tape P, control Situation
case:-
A
Child squeals, mother looks round calmly. They play slowly together, quietly but with peaceful intervals, child explores room visually. Situation
B
Child greets Situation Child
stranger
with smile.
C
squeals
as mother
leaves,
reverts
immediately
to play,
the
320
regards
C.A. Hyman then smiles
Situation Smiles
at stranger,
long slow regards.
D
broadly
at mother's
return.
Accommodation 1 bed sitting room, mother's in laws.
shared bathroom
and kitchen,
flat rented
from
REFERENCES Ainsworth, M., Stayton, D. and Bell, S. Individual differences in In Schaffer, R. The Origin of Human Relationships one year olds. Academic Press, London, (1971). Humphreys, P. Primate Programme, obtainable from the author, Department of Psychology, Brunel University, Kingston Lane, Uxbridge, Middlesex, England. (1973). N.S.P.C.C. Battered Child Research Kegan Paul, London, (1976).
Team, At Risk, Routledge
and
Acknowledgement is due to Ruth Mitchell and Geraldine Connor for Particular their participation in the video observation sessions. thanks are offered to Robert Parr for his meticulous coding, transcription and computer processing of the video tapes.