British Journal of Medical Psychology (1979).

52, 1-10 Printed in Great Britain 1

A Parental Bonding Instrument

Gordon Parker, Hilary Tupling and L. B. Brown

The concept of a ‘bond’ between a parent and a child is generally accepted despite, as Bowlby
(1%9) and Rutter (1972) have indicated, the lack of a satisfactory definition of the concept.
Theoretically, it might be proposed that parent-child bonds would be broadly influenced by
characteristics of the child (e.g. individual differences in attachment behaviour), characteristics
of the parent or care-taking system (e.g. psychological and cultural influences) and by
characteristics of the reciprocal, dynamic and evolving relationship between the child and the
parent. While that reciprocal interrelationship has resisted definition it is probably fair to suggest
that most research has examined the influence of single variables instead of attempting to
identify the principle dimensions of bonding. The aim of the research presented here was to
examine the parental contribution to a parent-child bond and to attempt to define and measure
the constructs of significance.
In reviewing the literature on parental qualities associated with normal development Rutter
(1972) isolated characteristics said to be necessary for adequate mothering: a loving relationship,
leading to an unbroken attachment to one specific person in the family who provides adequate
stimulation. As Rutter points out, studies are needed to confirm or refute many of these
‘characteristics’ and some have been disputed extensively. In discussing mother-child
interaction Bowlby (1%9) emphasized the retrieval behaviour of the mother which is concerned
with reducing the distance between infant and mother, so serving a protective function. Bowlby
noted that retrieval behaviour has been considered under many headings including ‘mothering’,
‘maternal care ’ and ‘nurturance ’. During the child’s development retrieval behaviour diminishes
until maintenance of proximity is left almost entirely to the growing juvenile. Ainsworth et al.
(1975) postulated that mothers ‘despite a massive overlay of learned behaviours ’, are biased to
respond to infants in species-characteristic ways which, in accord with Bowlby’s view, serve a
significant protective function. In their paper Ainsworth et al. drew attention to four dimensions
of maternal behaviour which were reflected in the balance of attachment and exploratory
behaviour in the children. The dimensions were labelled: sensitivity-insensitivity,
acceptance-rejection, cooperation-interference, and accessibility-ignoring. Paternal
characteristics have been neglected or only briefly considered by these authors.
One way of defining the parental contribution to bonding is to measure parental behaviours
and attitudes, and to determine source variables by factor analytic or other techniques. Roe &
Siegelman (1%3) factor analysed the responses of several independent samples of children and
adults to items dealing with their parents’ behaviour in their childhood. The first factor was
bipolar and considered to be ‘clearly one of affection and warmth, as contrasted with coldness
and rejection’. The second factor was entitled ‘casualdemanding’, with items at the latter end
suggesting strict regulation and intrusiveness together with demands for high accomplishment
and obedience. A third factor, which accounted for little of the variance, was unipolar and
suggested a protective concern for the child that was not necessarily affectionate. In their review
of nine other studies (variable based on observations of parents and their children, interviews of
mothers, or on children’s reports of present or past behaviours) they noted that eight of those
studies reported a factor ‘which is clearly one of affection and warmth, as contrasted with
coldness and rejection’. A similar factor to their second factor was reported in many of the
other studies but a factor similar to their third factor was isolated in only a few of the studies.
Schaefer (1%5) factor analysed responses of independent samples of children and adults in a
similar study. The first factor was bipolar and was labelled ‘acceptance versus rejection’- with
0007-1 129/79/0301-OOO1/$02.00/0 @ 1979 The British Psychological Society
1 MPS 52
2 Gordon Parker, Hilary Tupling and L . B. Brown
one pole defined by positive evaluation, sharing, expression of affection, emotional support and
equalitarian treatment and the other pole by ignoring, neglect and rejection. The second factor,
also bipolar, was labelled ‘psychological autonomy versus psychological control ’, with the latter
pole defined by items of intrusiveness, parental direction and control through guilt, suggesting
covert, psychological methods of controlling the child which do not permit it to develop as an
individual. The third factor was labelled ‘firm control versus lax control’, and appeared to be a
dimension reflecting the degree to which parents made rules and set limits. A more recent study
of 254 normal adults (Raskin et al., 1971) found three factors which resembled Schaefer’s three
conceptual dimensions. The largest percentage of common or shared variance was accounted for
there by the first two factors, while the third dimension seemed ‘to have less heuristic
significance’.
Findings from the studies reviewed here suggest that the parental contribution to bonding may
be influenced by two principal source variables. The first variable clearly appears to be a care
dimension. The second variable does not appear to be so readily definable but suggests a
dimension of psychological control over the child. These findings suggested that in developing a
parental bonding instrument it would be important to attempt to define the second dimension
quite precisely and that it would be unwise to attempt to use more than a two-dimensional
model.

Development of the instrument

To assist definition of the two principal dimensions, items suggesting parental behaviours and
attitudes of care, affection, sensitivity, cooperation, accessibility, indifference, strictness,
punitiveness, rejection, interference, control, overprotection, and encouragement of autonomy
and independence were generated from clinical notes and from the literature. As a pilot study the
114 items were given to 50 fifth year medical students who were asked to score each of their
parents on a four-point scale as remembered during their first 16 years. Their responses
suggested that many of the items were too direct and that the scale could be interpreted in
several ways. A further pilot study with a similar group suggested that with the items
reworded an acceptable set of scale scores could be obtained.
The revised inventory of 114 items was reduced further after administering it to 34 psychiatric
nurses and 19 medical students. In that sample there were 27 males and 26 females with a mean
age of 27 years. Two identical items were included to determine reliability of responses and a
correlation coefficient of 0.877 ( P < 0401) was obtained between them. The four items without
welldistributed responses were rejected. The remaining items were intercorrelated and further
reduced by deleting one item from those pairs with a high correlation coefficient. Responses to
the remaining 99 items were factor analysed, with a limitation of four factors, followed by a
varimax rotation, to reduce the item pool further. The first factor in that analysis accounted for
52 per cent of the variance, was strongly bipolar, and suggested a dimension of
‘carelinvolvement versus indifferencelrejection’. The second factor accounted for 29 per cent of
the variance and suggested a dimension of ‘control/overprotection/intrusionversus
encouragement of independence’. The third factor was bipolar, accounted for 11 per cent of the
variance and contained ‘overprotection’ and ‘encouragement of autonomy ’ items. The fourth
factor was difficult to interpret because of its diverse item content. The items which did not
contribute clearly to one of the first three factors were culled leaving 48 items for the principal
study. From the results of this analysis it was felt that a two-dimensional model might be used
having both a care dimension and a second dimension covering control or overprotection.
Validation concentrated on detailed consideration of these two dimensions.
 A parental bonding instrument 3
The principal study
The sample in the principal study comprised 65 medical students, 43 psychiatric nurses, 13
technical college students and 29 parents of children at a local school. There were 79 female and
71 male respondents ranging in age from 17 to 40 years (mean 25 years) and completed forms
were collected for 150 mothers and 148 fathers. A Likert method of scoring was used. Validation
of the scale scores was attempted against responses to a subset of Thematic Apperception Test
(TAT) cards, and by interview of 65 (43 per cent) of the sample. Three TAT cards (7GF, 6BM
and 2) were administered and a content analysis of responses was made by the third author.
However the responses consistently showed the expectation or stereotype that mothers are
controlling whether directly or indirectly, and suggested that TAT responses did not provide a
useful basis for validation. In interviews with the first and second authors acting as raters, each
subject was asked to discuss the ‘emotional relationship’they had with each parent, and the
likelihood that each parent would let them ‘do their own thing’. The two raters independently
assessed the content of these interviews and assigned a score from five to one for each parent’s
‘care ’ and ‘overprotectiveness’.
A factor analysis of the total samples’ responses to the 48 items was performed. N o limitation
was placed on the number of factors to be extracted. The first factor accounted for 27 per cent
of the total variance, the second for 14 per cent, while the third and fourth accounted for 5 and
3 per cent respectively. An attempt to define the first three factors was made after varimax
rotation. The first factor clearly involved a ‘care-indifferencelrejection’ dimension with factor
loadings from +0-763 to -0.671. The second was weakly bipolar with loadings from +0.720 to
-0.373 and comprised items at the positive pole that suggested overprotection and the fostering
of dependency. The third was weakly bipolar also, with loadings from +0.719 to -0-449 and
comprised items at the positive pole suggesting the encouragement of independence and
autonomy. Items weighting negatively on the second factor tended to weight positively on the
third factor and vice versa suggesting that these two factors could be collapsed into a single
factor. After culling the items which had poor loadings on the first three factors, 31 items were
left. A factor analysis was therefore performed with the factors to be extracted limited to two.
The first factor accounted for 28 per cent of the total variance and clearly suggested a ‘caring
versus indifferencelrejection’ dimension. It was therefore identified as a ‘care’ factor. The
second factor accounted for 17 per cent of the total variance, and its items suggested a
‘control/overprotection versus allowance of autonomy and independence’ dimension. It was
termed an ‘overprotection’ factor. While ‘overprotection’ appeared the most apposite term of
the several considered, it is not certain that all items are clearly confined by the construct of
overprotection.
The item loadings on each factor were examined and items with the highest loadings were
retained for the final scales. A graph of the two rotated factors is shown in Fig. 1 . Factor
analyses of responses to the 31 items were performed separately for the mothers and the fathers
to examine parents’ sex as a variable. Close agreement between the comparable mother and
father factors are shown in Table 1, suggesting that two dimensions are independent of the sex
of the parent.
The final scales consisted then of 25 items, comprising 12 ‘care’ items and 13 ‘overprotection’
items. Using a Likert scaling from 0 to 3, the 12 items of the ‘care’ scale allow a maximum
score of 36, and the 13 items of the ‘overprotection’scale allow a maximum score of 39. After
calculating scores for all respondents the properties of the final instrument were assessed.

Reliability and validity of the final scales

To obtain some measure of the reliability of the responses of the sample, two identical items had
been included in the 48-item questionnaire. Responses to these two items were intercorrelated
producing a Pearson correlation coefficient of 0.704 (P< 0.001).
4 Gordon Parker, Hilary Tupling and L. B. Brown
Overprotection
items

9 23
20 13
10 19 8

14
16
4 2
Indifference 24 6 Care
care
rejection items items
Items
18 12
5
II 17

25 22
3
15 7

Allowance of
autonomy and
independence
items
Figure 1. Plot of the rotated factors for both parents.

Seventeen members of the sample completed the inventory on two occasions three weeks
apart to assess test-retest reliability. A Pearson correlation coefficient of 0.761 (P< 0.001) was
obtained for the ‘care’ scale and 0.628 (P< 0401) for the ‘overprotection’ scale.
The questionnaire was divided into two halves for a measure of the split-half reliability. A
Pearson correlation coefficient of 0.879 ( P < 0-001) was obtained for the ‘care’ scale and 0.739
( P < 0401) for the ‘overprotection’ scale.
After joint interview with 65 of the subjects the two raters (G.P. and H.T.) independently
assigned a ‘care’ and an ‘overprotection’ score for each parent. The inter-rater reliability
coefficient on the ‘care’ dimension was 0.851 ( P < 0401) and 0.688 ( P < 0.001) on the
‘overprotection’ dimension. As a test of the concurrent validity of the scales the raters’ scores
of ‘care ’ and ‘overprotection’ obtained at interview were correlated with those determined by
the scales. The Pearson correlation for the two ‘care’ measures were 0.772 ( P < 0.001) for rater
G.P. and 0.778 (P<0.001) for rater H.T., and for the two ‘overprotection’ scores were 0.478
( P < 0401) for rater G.P. and 0.505 (P<0401) for rater H.T.
The intercorrelation between scores on the ‘care’ and ‘overprotection’ scales for the 300
responses for separate parents was -0.238 (P<0401) suggesting that scores on the two
dimensions were not independent in the study group.

Analysis of sample data

Mean scores of 24.9 for the ‘care’ scale and 13.3 for the ‘overprotection’ scales were obtained
for the sample. The possible influence of sex, both of the parent and respondent, was examined
(Tables 2 and 3). The results suggested that the mothers were experienced as more caring and
somewhat more overprotective than the fathers, but that the sex of the respondent did not
influence the parents’ capacity to ‘care’ or to be ‘overprotective’.
 A parental bonding instrument 5

Table 1. Factor loadings after varimax rotation of scale items for both parents, and for mothers
and fathers considered separately

Factor loading
Item
number' Both parents Mother Father

Care dimension
Care items 5 +0-760 +0.713 +0.793
17 +0-755 +0.716 +0*733
6 +0*754 +0.737 +0-741
Factor 1 12 +0*730 +0.675 +0.739
1 +0*709 +0.724 +0.673
11 +0.702 +0.664 +0*690
Indifferencel 14 -0.673 -0.670 -0.701
rejection items 4 -0.668 -0.652 -0.653
2 -0.647 -0.597 -0.682
24 -0.636 -0-677 -0.598
18 -0.635 -0.561 -0.645
16 -0.595 -0.677 -0.525
Overprotection dimension
Overprotection 9 +0.707 +0.727 +0.713
items 23 +0.703 +0*704 +0-652
13 +0.675 +0.735 +0.600

IIFactor 2

Encouragement of
autonomy and
independence items
19
8
20
10
15
7
3
+069
+0.662
+0*650
+0-628
-0.590
-0.576
-0.489
+0.67 1
+0643
+0*600
+0.621
-0.661
-0.647
-0.533
+0*525
+0.658
+0*618
+0*592
-0.531
-0.573
-0.395
21 -0.483 -0.511 -0.461
25 -0.468 -0.390 -0.534
22 -0.386 -0.357 -0.423

See Appendix for full item.

Table 2. Relationship between scale scores and sex of parent in principal study

Sex of parent

Scale Female Male t test Significance

Care 264 22.9 4.22 < 0.001

Overprotection 14.7 11.9 3.27 < 0.01

Table 3. Relationship between scale scores and sex of respondent in principal study

Sex of respondent
~~

Scale Female Male t test Significance

Care 24.2 25.7 1.52 n.s.

Overprotection 13.4 13.2 0.17 n.s.
6 Gordon Parker, Hilary Tupling and L. B. Brown
Table 4. Relationship between scale scores and sex of parent in general practice study

Sex of parent

Scale Female Male t test Significance

Care 26.9 23.8 8.28 < 0.001

Overprotection 13.3 12.5 2.3 1 < 0.025

Table 5. Relationship between scale scores and sexes of patients in general practice study

Sex of patients

Scale Female Male t test Significance

Maternal care 27. I 26.9 0.25 n.s.

Paternal care 24.3 23.1 1.47 n.s.
Maternal overprotection 12.9 13.9 1.12 n.s.
Paternal overprotection 12.7 11.7 1.29 n.s.

Some general population norms

As the scales were developed on an unrepresentative group, another sample was used to obtain
normative data. A consecutive series of 500 patients attending three general practitioners in
Sydney was obtained. These general practitioners were chosen to ensure a reasonable
representation of all social classes. Patients attending the practices were asked to complete the
questionnaire anonymously, and to provide their age, sex and father’s occupation, the last to
determine social class.
Sixty-seven patients failed to return their forms while 23 incorrectly or only partially
completed the forms. The responses of the remaining 410 (32 per cent males, 68 per cent
females) were analysed. Their mean age was 36 years, and ranged from 12 to 74 years with five
patients less than 16 years. The social class distribution of their fathers, with the Sydney general
population norms (Vinson, 1974) in parentheses, was: Class A, 15.3 per cent (4 per cent); Class
B, 31-9 per cent (19.1 per cent); Class C, 42.2 per cent (56.6 per cent); Class D, 10.6 per cent
(20.4 per cent). When the mothers’ and fathers’ care and overprotection scores were
intercorrelated, they produced a coefficient of -0.469 for mothers and -0.360 for fathers. This
negative correlation suggests again that the two parental dimensions are not independent and
that ‘overprotection’ is associated with a lack of ‘care’. The possible influence of parents’ sex
and respondents’ sex was examined (Tables 4 and 5). The mean scores are similar to those
obtained in the principal study, as are the findings that mothers are experienced as more caring
and somewhat more overprotective than are fathers, and that the sex of the child does not
influence a parent’s capacity to ‘care’ or to be ‘overprotective’.
The relationship between social class and parental care and overprotection was examined
(Table 6). While a weak positive association between higher social class and a greater maternal
care score was suggested, there was no clear association between social class and parental care
and overprotection. The effect of age was examined (Table 7) but an association between the age
of the recipient and parental care and overprotection was not found. This is an important finding
since a positive association would suggest either a change in the report of parental attitudes over
time or that the further removed respondents are in time from childhood the more their
responses may be influenced by the effect of social desirability.
 A parental bonding instrument 7
Table 6. Relationship between social class of parents and their care and overprotection scores

Class

A B C D F ratio Significance

Maternal care 29.7 27.4 26.1 25.9 3.63 0.013

Paternal care 24-9 24.6 23.3 23,4 0.94 n.s.
Maternal overprotection 13.2 12.9 13.5 14-6 0.49 n.s.
Paternal overprotection 14.2 12.5 11.8 13.8 1 46 n.s.

Table 7. Relationship between age of recipient and care and overprotection scores of their
parents

Maternal Paternal Maternal Paternal

Age Number care care overprotection overprotection

< 20 years 63 28.5 24.6 13.8 11.7

21-30 years 84 26-9 24.7 13.8 12.7
3140 years 110 26.6 23.7 14.1 13.7
41-50 years 71 25.7 22.1 12.9 12.6
51-60 years 36 28.9 25.4 11.8 11.4
261 years 25 26.3 24.9 11.6 11.3
Not recorded 21

Table 8. Relationship between parental dissonance and social class

Social class
~ ~ ~ ~

A B C D F ratio Significance

Mean parental dissonance 4.5 4.3 4.4 4.2 0.075 n.s.

scores (arbitrary units)

For each recipient, maternal and paternal scores were plotted on axes defining care and
overprotection. Measuring the distance between the position of each recipient’s mother and
father provides an arbitary measure of the degree of parental concordance or dissonance. The
possibility of an association between this measure (‘parental dissonance ’) and social class was
examined, but it was found to be non-significant (Table 8). Mean scores of 4.5 for male, and 4-6
for female, respondents were not significantly different ( t = 0.20).

Uses of the instrument

The scales we have developed may be used separately, or together as a bonding instrument.
Used together they allow five types of parental bonding to be examined (see Fig. 2): average
(defined statistically), high care-low overprotection (which might be conceptualized as optimal
bonding), low care-low overprotection (conceptualized as absent or weak bonding), high
care-high overprotection (conceptualized as affectionate constraint) and low care-high
overprotection (conceptualized as affectionless control). It is felt that the instrument could be
useful for considering optimal parental bonding and for examining the influence of parental
distortions on psychological and social functioning of recipients. For precision the
intercorrelation between respondents’ scores on the care and the overprotection scales should
8 Gordon Parker, Hilary Tupling and L. B. Brown
High
overprotection

‘Affectionless control ’ ‘Affectionate constraint ’

Low care High care

‘Absent or weak bonding’ ‘Optimal bonding’

Low
overprotection

F w e 2. The two scales of the Parental Bonding Instrument showing the conceptualized parental bonding
possibilities.
be examined. If there is no correlation the scale scores can be regarded as independent
measures. If there is a correlation then a correction is required to partial out any
overprotection contribution to a care score and vice versa to create orthogonal axes. In clinical
studies it might be most useful to use raw scale scores, despite any correlation between scores
on each scale; in studies assessing any relationship between care or overprotection and another
variable, independent scores, with any contribution by one scale to the other partialled out, may
be more useful. A third use is to examine the degree of concordance or dissonance between
parents’ principal attitudes by plotting and measuring the distance between each parent’s position
on the two axes, for comparison against other variables.
Discussion
Care has been identified theoretically, and supported empirically by factor analytic studies, as
the major parental dimension. The significance of an overprotection dimension has received little
theoretical consideration despite findings from factor analytic studies. Starting from that basis,
the present study sought to define both dimensions empirically and to produce scales which
would be short, reliable and valid measures of those dimensions. The first derived factor in the
present study involved one pole defined by affection, emotional warmth, empathy and closeness,
and the other by emotional coldness, indifference and neglect. Its items appeared to measure the
presence or absence of care and formed the care scale (see Appendix). The second factor has
one pole defined by control, overprotection, intrusion, excessive contact, infantilization and
prevention of independent behaviour, and the other defined by items that suggest allowance of
independence and autonomy. It was difficult to identify from the items a principal construct
underlying this dimension, but as Levy (1970) has suggested that overprotectiveness is
principally revealed clinically by excessive contact, infantilization and prevention of independent
behaviour, the items in this dimension were conceptualized as measuring the presence or
absence of overprotection and formed the overprotection scale (see Appendix). The difficulty
in defining the second dimension, compared to the first, is indicated by its lower inter-rater
 A parental bonding instrument 9
reliability and validation figures. The two dimensions do not appear to be independent as their
scores correlated negatively in both the principal study and in the general population study. The
results showed that overprotection is linked with lack of care.
Used together the scales form an instrument which aids study of optimal and distorted
parental bonding, the principal aim of their development. It is true that they rely on a subject’s
own responses, and that they assume some parental consistency over the lengthy period of
infancy, childhood and early adolescence. The importance of subjective experience is generally
accepted. The validity of the present instrument was therefore assessed at interview by
examining it against the internal consistency of the sample’s reported experiences, rather than
by any study of their parents. In considering the second point, consistency of parental behaviour
cannot be readily assumed. Variations in care and overprotectiveness, as a consequence of time
or family circumstances, are usual. Ainsworth et al. (1975) noted that infants are biased to leave
the mother to explore the world as soon as they are able, and that the dynamic balance between
that behaviour and care-taking behaviour is subject to many shifts. What is measured in the
scales that we developed is the recipient’s later judgement of a parent, which is an assessment
made after childhood and early adolescence. While it is unlikely that parental behaviours will
vary so capriciously and extensively as to alter scores significantly on the two scales, this
perhaps should be examined by performing serial test-retest measures over a period of time. We
have assumed that the scales reflect a moment, or product of innumerable specific experiences,
as well as involving some consistency over time.
Since bonding is an important concept we hope that these scales may assist research in
defining the parental contribution to optimal bonding and in considering the sequelae of distorted
parental bonding.

summary
Two new scales of parental care and overprotection, and their combination as a Parental
Bonding Instrument, are described. On measurements of reliability and validity the scales appear
to be acceptable, and are independent of the parent’s sex. It would appear that mothers are
perceived as significantly more caring and slightly more overprotective than fathers, but that
those judgements are not influenced by the sex of the child. Overprotection appears to be
associated with lack of care. The scales and scoring method are appended. Norms for a general
Sydney population are presented, and the possible influence of age, sex and social class
examined.

Acknowledgements
We wish to thank Professor L. G. Kiloh and Mrs Megan Neilson for their constructive advice; Dr Kevin
Bird for statistical assistance; Dr Earle Connolly, Dr Bruce Glass and Dr John Rolleston for testing the
general practice sample, and Di Frances for secretarial assistance.

References
AINSWORTH, M. D. S., BELL,S. M. & STAYTON, Factor analyses of normal and depressed patients’
D. J. (1975). Infant-mother attachment and social memories of parental behavior. Psychological
development: Socialisation as a product of Reports, 29, 871-879.
reciprocal responsiveness to signals. In ROE, A. & SIEGELMAN, M. (1%3). A parentahild
M. Richards (ed.),The Integration of the Child questionnaire. Child Development, 34, 355-369.
into a Social World. Cambridge: Cambridge RUITER,M. (1972). Maternal Deprivation
University Press. Reassessed. Harmondsworth: Penguin.
BOWLBY,J. (1%9). Attachment and Loss, vol. 1: SCHAEFER.E. S. (1%5). A configurational analysis
Attachment. London: Hogarth Press. of children’s reports of parent behavior. Journal
LEVY,D. M. (1970). The concept of parental of Consulting Psychology, 29, 552-557.
overprotection. In E. J. Anthony & T. Benedek VINSON,T. (1974). Crime, Correction and the
(eds), Parenthood. Boston: Little Brown. Public. Report No. 17, N.S.W.Bureau of Crime
RASKIN,A., BOOTHE,H. H., REATIG,N. A., Statistics and Research.
SCHULTERBRANDT, J. G . & ODLE,D. (1971).
10 Gordon Parker, Hilary Tupling and L. B. Brown
Received 1 June 1977

Requests for reprints should be addressed to Gordon Parker, School of Psychiatry, University of New South
Wales, PO Box 1 , Rozelle, NSW 2039, Australia.
Hilary Tupling is at the same address.
L. B. Brown is at the School of Psychology, University of New South Wales, Kensington, NSW 2033.

Appendix
The Parental Bonding Instrument and its scoring. Scores for the care scale are recorded in Arabic
numerals. Scores for the overprotection scale are recorded in Roman numerals.
Femalelmale parent form:
This questionnaire lists various attitudes and behaviours of parents. As you remember your Mother/Father in
your first 16 years would you place a tick in the most appropriate brackets next to each question.

Very Moderately Moderately Very

like
1. Spoke to me with a warm and friendly voice ( 3)
2. Did not help me as much as I needed . 0
3. Let me do those things I liked doing 0
4. Seemed emotionally cold to me 0
5. Appeared to understand my problems and worries ( 3)
6. Was affectionate to me ( 3)
7. Liked me to make my own decisions 0
8. Did not want me to grow up (111)
9. Tried to control everything I did (111)
10. Invaded my privacy (111)
1 1. Enjoyed talking things over with me ( 3)
12. Frequently smiled at me ( 3)
13. Tended to baby me (111)
14. Did not seem to understand what I needed 0
or wanted
15. Let me decide things for myself 0
16. Made me feel I wasn’t wanted 0
17. Could make me feel better when I was upset ( 3)
18. Did not talk with me very much 0
19. Tried to make me dependent on her/him (111)
20. Felt I could not look after myself unless (111)
she/he was around
21. Gave me as much freedom as I wanted 0
22. Let me go out as often as I wanted 0
23. Was overprotective of me (111)
24. Did not praise me 0
25. Let me dress in any way I pleased 0