This article was downloaded by: [University of Stellenbosch]

On: 04 October 2014, At: 18:47

Publisher: Routledge
Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office:
Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK

Aging & Mental Health

Publication details, including instructions for authors and subscription
information:
http://www.tandfonline.com/loi/camh20

The Parent Adult-Child Relationship

Questionnaire (PACQ): The assessment of
the relationship of adult children to their
parents
C. PEISAH , H. BRODATY , G. LUSCOMBE , J. KRUK & K. ANSTEY
Published online: 09 Jun 2010.

To cite this article: C. PEISAH , H. BRODATY , G. LUSCOMBE , J. KRUK & K. ANSTEY (1999) The Parent
Adult-Child Relationship Questionnaire (PACQ): The assessment of the relationship of adult children to
their parents, Aging & Mental Health, 3:1, 28-38, DOI: 10.1080/13607869956415

To link to this article: http://dx.doi.org/10.1080/13607869956415

PLEASE SCROLL DOWN FOR ARTICLE

Taylor & Francis makes every effort to ensure the accuracy of all the information (the
“Content”) contained in the publications on our platform. However, Taylor & Francis, our
agents, and our licensors make no representations or warranties whatsoever as to the
accuracy, completeness, or suitability for any purpose of the Content. Any opinions and views
expressed in this publication are the opinions and views of the authors, and are not the views
of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon and
should be independently verified with primary sources of information. Taylor and Francis shall
not be liable for any losses, actions, claims, proceedings, demands, costs, expenses, damages,
and other liabilities whatsoever or howsoever caused arising directly or indirectly in connection
with, in relation to or arising out of the use of the Content.

This article may be used for research, teaching, and private study purposes. Any substantial
or systematic reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or
distribution in any form to anyone is expressly forbidden. Terms & Conditions of access and
use can be found at http://www.tandfonline.com/page/terms-and-conditions
 Aging & mental Health 1999; 3(I): 28± 38

ORIG INA L ART ICLE

The Parent Adult± C hild Relationship Questionnaire (PAC Q):

the assessm ent of the relationship of adult children to their parents

1 1,2 1,2 3 4
C. PEISAH, H. BRODATY, G. LUSCOM BE, J. K R UK & K. ANSTEY
1 2
Academic Department of Psychoger iatrics, Prince H enry Hospital, Sydney, School of Psychiatr y, University of N ew
3 4
South Wales, Formerly Neuropsychiatric Institute, Prince Henry Hospital, Sydney, Formerly Academic D epartm ent
of Psychoger iatrics, Prince H enr y Hospital, Sydney, Australia
Downloaded by [University of Stellenbosch] at 18:47 04 October 2014

Abstract
We review current understanding of the parent adult± child relationship, describe the development of an instrument to
measure this relationship and propose key themes which de® ne the relationship. A 97-item self-report questionnaire was
administered to two convenience samples (N = 302, N = 100) comprising primary care patients and comm unity residents.
The m ost interpretable solutions to Principal Com ponents Analyses of a reduced pool of items were, for the mother, an
orthogonal two-factor solution (regard and responsibility , comprising 13 items) and, for the father, a three-factor solution
(regard, responsibility and control , comprising 13 items). The Parent Adult± Child Questionnaire (PACQ) was found to be valid
and reliable as a self-report m easure of the ® lial relationship. We suggest that PAC Q has both clinical and research utility.

Introduction choice, close ness, power and statu s (M an cin i &

Blieszn er, 1989) , bu t these var y g re atly am o n g
The role of early child± parent relationships in the fam ilies. D e Vaus (1994) identi® ed four broad types
genesis of adult psychopathology has been extensively of parent adult± child relationships: `attached’ , char-
investigated, yet the im pact of parent adult± child acterized by m utual interest, support and autonomy;
relationships on adult m ental health has been little `remote’ , characterized by m utual disinterest; `parent-
explored. Relationships with parents are often a focus centred’ , w he re parents, pe rceive d as excessively
of psychiatric treatment, whether in the context of demanding, expect attentiveness and conform ity from
psychotherapy, in the treatment of depression or anxiety, their adult children who struggle to individuate; and
or in suppor ting the fam ily of an elderly patient with `child-centred’ , where the children’s lives provide the
psychiatric problems. D espite the frequency with which focus and the parents rem ain critical, interfering and
the relationship is ad dressed in ever yday clin ical unable to give up the parenting role.
practice, and the acknowledgement that interpersonal T hese variations in parent adult± child relation-
relationships have an im pact on mental health (Hend- ships can be attributed to a variety of factors. Firstly,
erson et al., 1981), little research has focused speci® - the parent’s resources (i.e. physical, emotional and
cally on the parent adult± child relationship. D e® nable ® nancial) determ ine the am ount of support expected
and measurable characteristics of the parent adult± or needed and m ay som etimes in¯ uence the child’s
child relationship have yet to be determ ined and, satisfaction with the relationship. Parents w ho receive
im portantly, no instrum ent exists to measure the parent and expect high levels of ® lial support are m ore likely
adult± child relationship per se. This paper will review to be single, of low incom e and in poor health with
brie¯ y the parent adult± relationship in adulthood and low m orale (Seelbach & Saur, 1977). Well-educated,
relevant m easures and describe the development of a resou rcefu l, healthy elderly requ ire on ly rou tine
new instrument quantifying the relationship. interactions with their children, do not expect direct
T he evolution over m any years of the child± parent assistance except in the m ost extrem e circum stances
relationship, together w ith the natural diversity of and are m ore concerned with remaining close without
hum an relationships, have m ade it difficult to de® ne interfering (Blieszner & M ancini, 1987). M oreover,
m odels or a typology of fam ily function in adult- active and self-su fficient parents are m ore likely to
hood. K ey constructs appear to be obligation and receive positive appraisals from their adult children

Correspondence to: Professor Henry Brodaty, Academic Department of Psychogeriatrics, Prince Henry Hospital, Little Bay,
Sydney 2036, Australia. Tel: 61 2 9382 5007. Fax: 61 2 9382 5016. E-mail: research.adpg@ unsw.edu.au
Received for publication 28th November 1997. Accepted 13th August 1998.
1360-786 3/99/010028-1 1 $9.00 ½ Carfax Publishing Limited
 The parent adult± child relationship questionnaire (PACQ) 29

(Robinson & T hurner, 1979), whilst poor health in a (1) the early child± parent relationship such as the
m other is associated w ith lower rapport with her Children’s Reports of Parental Behaviour Inven-
daughter (Baruch & Barnett, 1983). tor y (Schaefer, 1965); the Parental B ond ing
Secondly, the previous child± parent relationship or Instrument (Parker, 1983); the Parent Child Rela-
the perception of this m ay determ ine the child’ s tions Q uestionnaire (Siegelman & Rod, 1979);
preparedness to assu m e the caregiver role and any and the O wn M em ories of Childhood Rearing
asso ciated burden. Adult children w ho recalled early Experiences (Perris et al., 1980).
parental rejection tend to be less concerned about (2) the relationship in adolescence or young adult-
staying in close touch w ith their parents, m onitoring hood, w hen issues of individuation and separa-
their wellbeing or providing support (W hitbeck et al., tion predominate over issues of ® lial reciprocity
1994); although som e children rem ain com m itted to and responsibility. E xam ples include the Parental
their parents despite a lack of affection (H orowitz & Attachm ent Q u estion naire (K enny, 1987),
Schindelm an, 1983) or n egative feelin gs such as designed for ® rst-year college students, which
im patience, frustration and irritation (Cicerelli, 1983). assesse s dim ensions which are likely to be life
Such commitment m ay be driven by obligation, which cycle speci® c, i.e. Parents as a Source of Support,
su bsu m es factors such as gu ilt, culturally d riven Parents as Facilitators of Ind epen dence, and
expectations and m orality, as well as gre ed and Affective Q uality of the Parental Relationship,
subjugation. Conversely, feelings of closeness, love and does not distinguish between m aternal and
Downloaded by [University of Stellenbosch] at 18:47 04 October 2014

an d com p atibility increase attach m ent (visitin g , paternal attachm ent; the C ontinued Attachm ent
telephoning) and helping (e.g. personal care) behavi- Scale-Parent version (Berm an et al., 1994), which
ours (C icerelli, 1983). M any adult children welcom e m easures cognitive and emotional responses to
the opportunity to reciprocate a nurturant childhood the perceived separation from parents in college
and, accordingly, affe ction, gratitude and `equity’ drive students; and the Inventory of Parent and Peer
the ® lial relationship (Horl, 1989). Im portantly, close- A ttachm en t (A rm sden & G reenb erg , 1987),
ness does not always equate with a positive relation- which was developed to assess Trust, Availab ility
ship (O ’ C on nor, 1990). W hile care given in the and Security in parental and peer relationships
context of positive affe ct is perceived as less stressful during early to m iddle adolescence.
(H orowitz & Schindelman, 1983), the `closer’ the (3) non-speci® c adult or fam ily relationships, assu m -
em otional bond between child and parent the m ore ing a com m o nality in all ad ult attachm en ts,
stressful the caregiver role (Cantor, 1983). regardless of the attachm ent ® gure. For exam ple,
A third determ inant of the nature of the relation- the Attachm ent Style Inventory (Sperling et al.,
ship is the personality and gender of the child. W hen 1992) asks subjects to identify their relationships
an elderly parent becomes sick and frail, the child on with parents, friends and intimates from a number
w ho m responsibility for care fa lls is u su ally the of descriptors derived from attachment and object
daughter (Brody, 1981). Wom en caregivers (Brodaty relations theory, depicting D ependent, Avoidant,
& H adzi-Pavlovic, 1990; F itting et al., 1986; Hin- H ostile and Resistant/Am bivalen t attachm ent
r ichsen , 1991; M or ris et al., 1991), par ticu larly styles. The Adult Attachm ent Scale (West et al.,
daughters (Scho® eld et al., 1995; Tudor et al., 1995), 1987) is also a non-spe ci® c scale which m easures
have higher rates of psychological m orbidity and dim ensional aspects of attachm ent such as Feared
burden than their m ale counterpar ts, possibly because Loss and Secure Base, and adult styles of attach-
of an emotionally involved, hands-on provider approach ment such as Com pulsive Caregiving, Com pulsive
rather than an instrumental and managerial approach Careseeking, C ompulsive Self-reliance and Angry
to caregiving. Gender differences in non-caregiving W ithdrawal. Sim ilarly, Hazan and Shaver (1987)
parent adult± child relationships are mainly speculative, developed an instrument used to classify rom antic
anecdotal rather than empirical and largely focused on love as either S ecu re, Avo id an t or A n xiou s/
exploring speci® c aspects of m other± daughter relation- Am bivalent.
ships (Chodorow, 1978; Freud, 1961; Friday, 1981; (4) caregiving asp ects of relationships w here parents
Phillips, 1991; Rich, 1977). are sick or impaired. M any instrum ents m easure
In sum m ary, the parent adult± child relationship is fam ily burden in severe m ental illness (Schene et
com plex and m ay depend on past aspects of the al., 1994) and several focus on the im paired
relationship, the parent’s wellbeing, gender of the elderly (usually w ith dementia) (C airl et al., 1985;
parent and the child, as well as other factors. Comm it- Vitaliano et al., 1976; W hittick, 1989; Zarit et al.,
m ent to care and `closeness’ m ay be independent of 1980), though none is speci® c for child caregivers.
positive or negative feelings about the parent. K ey, Moreover, they m easure aspects such as economic
quanti® able constructs rem ain largely theoretical and costs, social networks, grief reactions, attitudes
untested em pirically, perhaps in part because there is and em otional and physical burden rather than
no instru m ent which m easures the parent adult± the relationship itself.
child relationship. Currently available instrum ents can (5) interactions between non-speci® c fam ily m embers
be broadly classi® ed into ® ve types which m easure: such as the Cam berwell Fam ily Interview (Leff
 30 C. Peisah et al.

& Vaughn, 1985; Vaughn & Leff, 1976), w hich Statistical analyses
derives Expressed E m otion scores with ratings of
Criticism , H ostility, Warm th, Positive com ments All analyses were carried out using SPSS software
and O ver-involvem ent; the M cM aster Fam ily (Norusis & SPSS Inc, 1993). The ® nal questionnaire
Assessm ent D evice (E pstein et al., 1983), w hich was derived by subjecting a reduced pool of items to
m easu re s P ro blem solvin g , C om m u n ication , fa ctor an alysis u sing the Principal C om p on en ts
Roles, Affective responsiveness, Affe ctive involve- m ethod of extraction. Because of possible gender
m ent, Behaviour control and general functioning; differences, solutions were also generated separately
and the Fam ily Environm ent Scale (M oos & for sons and daughters. Spearm an’s correlation coef-
M oos, 1981), w hich m easures C ohesion, Expres- ® cients (r s) were used to exam ine the asso ciation
siveness and Con¯ ict. between age and factor scores because age was not
In this study we aimed to identify constructs central norm ally distributed. Two aspects of reliability were
to the ® lial relationship and to develop an instrum ent assessed: internal consistency using Cronbach’s alpha
to m easure the parent adult± child relationship w hich and test-retest reliability using Spearm an’s correla-
was sim ple to adm inister and score. tion coefficient (r s) for dim ensional variables which
were not norm ally distributed. Spearm an’ s correla-
tion coefficient was also used to com pare instrum ent
M ethods
and interview scores to determine concurrent validity.
Downloaded by [University of Stellenbosch] at 18:47 04 October 2014

Derivation of items N in ety-® ve p er cen t co n ® d en c e in ter vals w ere

calculated for all correlations. N orm ality of variable
Item s depicting the gam ut of constructs asso ciated distribution was assesse d using skewness values (S)
with the parent adult± child relationships including (Tabachnick & Fidell, 1989). All tests were two-tailed.
® lial obligation, reciprocity, gratitude, anger, financial W ithin each sam ple, ® gures for m other and father
gain and concern for the physical and psychological differed if subjects had only one living parent.
wellbeing of the parent were generated from the
published literature, pilot interviews with both clinical
and non-clinical subjects, and surveys of psychiatrists Reliability and validity studies
who were asked to generate or m odify items. C linical
subjects were those w ho presented with non-sp eci® c Statistical analyses generated a re® ned instrum ent
difficulties in their relationship w ith their parents, called the Parent Adult± Child Questionnaire or PACQ
and children of depressed parents. Psychiatrists were (Appendix 1), w ith fewer items (see below) and from
senior general psychiatrists, psychotherapists, child which factors of regard, responsibility and control were
and adolescent psych iatrists and psychogeriatricians. differentially derived from m others and fathers (see
below). Test-retest reliability was tested using a third
Derivation of the samples and testing sam ple (Sample 3) of 26 non-clinical volunteers from
the com m unity w ith the sam e inclusion criteria used
An earlier and lengthier version of the ® nal question- for Sam ples 1 and 2. The re® ned instrum ent was
naire was administered to two sam ples. We used two adm inistered to Sam ple 3 and then readm inistered to
samples to strengthen our ® ndings and to counter any the sam e sam ple three to six weeks later. We assessed
response bias associated with the ordering of items in the validity of the instrum ent, as a m easure of the
such a lengthy questionnaire (the item order being rand- relationship perceived by the child, by interview ing a
om ized for Sam ple 2). Sam ple 1 com prised 302 fourth sam ple (Sam ple 4) of another 26 non-clinical
subjects: 200 prim ary care patients recruited from ten volunteers from the com m unity using the sam e inclu-
general practices and a convenience sample of 102 sion criteria as previously. The interview com prised
from the community. Sample 2 comprised a community six open questions (e.g. `Tell m e how you feel about
convenience sample of 100 subjects. The prim ary care your m other’ ) designed to elicit qualitative data, and
patients were consecutive routine attenders of general closed questions to assess speci® cally the factors of
practices across the Sydney metropolitan area, who regard (e.g. `H ow close are you to your m other?’ ) and
were aged 18 or over, with at least one living parent and responsibility (e.g. `How reliant on you is your mother?’ )
a reasonable command of English. Both community for m other, and the three dim ensio ns of regard,
convenience sam ples were volunteers recruited by the responsibility and control for father. Each of the closed
® rst author and colleagues. Subjects were asked to questions, of which there were six for regard, eight for
provid e d em ograp hic in for m ation an d rate their responsibility and four for control, was rated on a four-
responses to 97 items on a four-point Likert-type scale point ordinal scale. A global rating of each of the
ranging from 0 (`very true’ ) to 3 (`not true at all’ ). Two three factors was also rated on a four-po int ordinal
identical versions of the 97-item scale were administered scale. T he interview score for each factor was thus a
to each subject, one which referred to `mother’ and the com posite of the sum med ordinal ratings plus the
other to `father’ . Consenting subjects completed the global rating; the m axim al interview factor scores
questionnaires alone and anonymously and placed them thus being different from the m axim al PACQ factor
in collection boxes. scores. After the interview, the subjects were asked to
 The parent adult± child relationship questionnaire (PACQ) 31

com plete the PACQ and resultant scores were inter- of parental divorce and separation (`I know that my
correlated with the interview scores. All the interview s m other/father will be taken care of by my other parent
were conducted, au diotaped and rated by C P, a if I don’ t’ ); or to re¯ ect a general personality response
psychiatrist and fam ily therapist (`A’ ). Audiotaped set (`I hate it when my m other/father tells people how
interviews were rated by a second psych iatrist or a strong I am and how well I cope because I feel that I
psychologist (HB or G L) (`B’ ). don’ t’ ). T he residual item s were further reduced by
rem oving items which were highly intercorrelated (>
0.7) with the rest of the item pool or w hich did not
Results provide a sufficiently wide range of responses (item s
which had over 90% of responses on either the `agree’
D emog raphics or `disagree’ end of the scale). The rem aining item s
(74 for the m other solution and 32 for the father
Sam ple 1 com prised 219 fem ales (73% ), 82 m ales
solution) were subjected to a factor analysis and both
(27% ) and one subject whose gender was not speci-
or thogon al and obliquely rotated solu tion s were
® ed. T he m ean subject age was 32 years (SD = 9.4,
exam ined, as well as unforced and forced rotations of
range 18± 59 years) and the m ean parent age was 64
two, three, four, ® ve and six factors, using paired and
(SD = 12.8, range 39± 93 years). Subjects were profes-
unpaired items. The solutions chosen (i.e. two-factor
sional or m anagerial workers (37% ), trades, clerical,
for m other and three-factor for father) were those
Downloaded by [University of Stellenbosch] at 18:47 04 October 2014

sales or process workers (30%), students (13%), home

w hich provided the m ost co n cep tually co heren t
w orkers (4% ), pension ers (1% ) o r em ploye d in
factors while maxim izing percentage of the varian ce
another occupation (11%). O ccupational data were
explained and internal consistency. A factor loading
m issing in 4% of cases. Eighty-four per cent were
threshold of 0.6 determ ined the ® nal items. Scree
from Australian or other English-speaking back-
plots con® rm ed the num ber of factors chosen were
grounds, while 3% were from European, 4% from
app ropriate, with a substantial drop in eigenvalues
Asian and 7% from other backgrounds (e.g. African,
between the second and third factors for the m other
M iddle Eastern, South Am erican). Country of birth
was not speci® ed in 2% of cases. solution and the third and fourth factors for the father
Sample 2 com prised 67 fem ales and 33 m ales, solution.
w ith a m ean age of 31 years (SD = 10.5, range 18± 60
years) and a m ean parent age of 59 years (SD = 10.2, M other. For the m aternal relationship, an orthogonal
range 40± 83 years). Fifty-o ne percent were profes- two-factor solution appeared to be the m ost interpret-
sional or m anagerial workers, while 18% were trades, able. The ® nal solution com prised two factors which
clerical, sales or process workers, 4% did home duties, were labelled regard and responsibility. After exam ining
18% were students, 6% were pensioners and 1% factor loadings and deleting item s w hich showed
worked in other m iscellaneous occupations. O ccu- m ajor differences between sons and daughters (i.e.
pational data were m issing in 2% of cases. Seventy- item s which loaded on different factors for sons and
® ve per cent were from Australian or other English- daughters), the regard factor was lim ited to ® ve item s
speaking backgrounds, while 7% were from European, and the responsibility factor to eight items. The regard
4% from Asian and 12% from other backgrounds factor accounted for 29.3% (eigenvalue 3.81) and
(e.g. African , M idd le E astern , So uth A m erican). the responsibility factor for 26.7% of the varian ce
C ountry of birth was not speci® ed in 2% of cases. (eigenvalue 3.47). T he factor loadings for the item s
Sample 3 (reliability testing) was sim ilar dem o- contributing to these factors for sons and daughters
graphically to that of Sam ples 1 and 2, i.e. 17 fem ales com bined, as well as the factor loadings derived
(65% ), six m ales (23% ) and three (12% ) gender separately for sons and daughters, are provided in
unknown; m ean subject age 33 years (SD = 8.8, range rank order in Table 1.
19± 53 years). Parent age was not obtained for this
sam ple. Sam ple 4 (validity testing) was also similar Father. For the paternal relationship, an orthogonal
dem ographically, i.e. 20 fem ales (77% ) and six m ales three-factor solution was m ost interpretable. These
(23% ), with a m ean subject age of 33 (SD = 8.7, three factors consisted of regard and responsibility,
range 19± 60 years) and a m ean parent age of 63 conceptually sim ilar but w ith different item s to those
(SD = 8.4, range 49± 87 years). derived for m other, plus an additional factor term ed
control. After exam ining factor loadings and removing
overtly sex-spe ci® c item s, the ® nal solution for father
D erivation of the re® ned questionnaire generated four items on the regard factor, four on the
responsibility factor and ® ve on the control fa ctor
T he initial, exploratory factor analysis was perform ed (Table 2). The regard factor accounted for 18.9%
using data collected from Sam ple 1 (N = 302). Items (eigenvalue 2.46), the responsibility factor for 10.6%
were deleted which were found to be am biguous (e.g. (eigenvalue 1.37) and the control factor for 37.2% of
`I no longer feel that I have to be the right person for the variance (eigenvalue 4.84).
my m other/father’ ); to be contam inated by realities The control factor, w hich accounted for a relatively
 32 C. Peisah et al.

TAB LE 1. Factor loadings for items contributing to the ® nal m easureÐ PAC Q (mother)

Item Factor loading

All Sons Daughters

Fac tor 1Ð R egard

I look forward to seeing my m other 0.86 0.85 0.86
I am glad to be able to repay my m other for all the love and care she 0.81 0.72 0.83
gave me as a child
I respect my mother’s opinion 0.80 0.70 0.83
M y m other is my best friend 0.79 0.66 0.82
M y m other shows her appreciation of me 0.78 0.79 0.79

Fac tor 2Ð R espon sibility

I feel like I parent my m other 0.76 0.76 0.75
M y m other relies on m e too much 0.75 0.80 0.71
I feel that I have to protect my mother 0.70 0.57 0.73
M y m other thinks I’ m good in a crisis so she calls on me all the time 0.67 0.68 0.67
I am the only one my mother can rely on 0.66 0.76 0.63
M y m other’s difficulty m aking decisions has been a burden on m e 0.61 0.68 0.59
I feel that I should take care of my m other because she has suffered so 0.61 0.48 0.66
Downloaded by [University of Stellenbosch] at 18:47 04 October 2014

much in her life

I feel responsible for my m other’s happiness 0.60 0.54 0.61

TAB LE 2. Factor loadings for items contributing to the ® nal measureÐ PACQ (father)

Item Factor loading

All Sons Daughters

Fac tor 1Ð R egard

I respect my father’s opinion 0.84 0.85 0.84
I look forward to seeing my father 0.83 0.74 0.85
I know I can rely on my father to help me if I need him 0.78 0.72 0.79
I don’ t mind putting myself out for my father 0.78 0.80 0.77

Fac tor 2Ð R espon sibility

Something will happen to my father if I don’ t take care of him 0.86 0.92 0.82
I feel responsible for my father’s happiness 0.81 0.80 0.80
If I don’ t see my father for a week I feel guilty 0.72 0.87 0.70
M y father thinks I’ m good in a crisis so he calls on m e all the time 0.60 0.74 0.54

Fac tor 3Ð C ontro l

If I don’ t do things my father’ s way he will nag me 0.85 0.83 0.86
I feel that my father tries to m anipulate me 0.80 0.81 0.81
M y father tries to dominate me 0.77 0.72 0.76
I feel that my father m akes too many demands on me 0.76 0.70 0.76
I don’ t discuss m uch with my father because I’ m afraid of being criticized 0.71 0.69 0.70

large percentage of the varian ce for fathers, did not Factor scores and age
appear in the m ultifactorial m other solutions. W hen
the control items derived from the father sam ple were Factor scores were then calculated for each of Sample
exam ined as a possible third (i.e. forced) factor in the 1 and Sam ple 2. After exam ining correlations between
m other sam ple, they emerged as a separate factor for subject age and factor score in the two sam ples we
daughters only and even then accounted for less than found no signi® cant associations (r s ranged from -0.17
3% of the varian ce. T hus, the control factor was to 0.10).
retained only for fathers.
T he factor analysis was con® rm ed in Sam ple 2
(N = 100). T he m other solution was identical except Reliability
that regard counted for slightly less of the varian ce
(24.0% , eigenvalue 3.12) and responsibility for slightly Internal con sistency as m easured by C ro nbach’ s
m ore (27.9% , eigenvalue 3.62). T he father solution alphas were for the m other solution, 0.87 for the
was also reproduced in the second sam ple except for regard factor and 0.82 for the respon sibility factor. T he
the dropping of one item from the responsibility factor father solution yielded a Cronbach’s alpha of 0.86 for
(eigenvalue 1.26, 9.7% of variance; control: eigen- the regard factor, 0.74 for the responsibility factor and
value 5.33, 41% of variance, regard: eigenvalue 2.41, 0.87 for the control factor.
18.5% of variance). T he reliability of the derived factors is re¯ ected
 The parent adult± child relationship questionnaire (PACQ) 33

further in the sim ilarity in the m ean scores across the that it is reliable and valid. The high correlations
two occasions (Table 3) and the size of the correla- between factor scores returned on our measure and
tion coefficients which were r s = 0.9 2 [C I (95% assessm ents of the dimensions by two independent
C on® dence Interval) = 0.82± 0.96; n = 24] for the raters support the concurrent validity of the m easure.
m other rega rd factor; r s = 0.9 3 (C I = 0.8 4± 0.9 7; Internal consistency as re¯ ected by Cronbach’ s alpha
n = 24) for the m other responsibility factor; r s = 0.89 was high, and m ean test scores for each factor were
(C I = 0.75± 0.95; n = 21) for the father regard factor; stable over tim e and test-retest correlation coef-
r s = 0.77 (CI = 0.50± 0.90; n = 21) for the father ® cients were high, supporting reliability.
respon sibility factor; and r s = 0.84 (C I = 0.63± 0.93; The sim ilarity between the dim ensions of regard,
n = 21) for the father control factor. All correlations responsibility and control identi® ed in this study and
were highly signi® cant at p < 0.001 level. constructs of care/affe ction, protection and control
previously identi® ed in studies of early child± parent
relationships (Arrindell et al., 1986; Bow lby, 1979;
Validity 1980; M urphy et al., 1997; Parker, 1983; Schaefer,
1965) su ggest that these them es per vade child±
Interrater reliability coefficients (rater `A’ and rater parent relations throughout the life cycle, regardless
`B’ scores intercorrelated) are outlined in Table 4. All of the descriptive or conceptual m odel used. Sim ilar
correlations were highly signi® cant at p < 0.001 level. constructs have been identi® ed in studies of intimate
Downloaded by [University of Stellenbosch] at 18:47 04 October 2014

All PAC Q factor scores correlated highly (p < 0.001) relationships, leading to the untested hypo thesis that
w ith both judges’ ratings of the three dim ensions of individuals have a general tendency to care, protect
regard, responsibility and control. N inety-® ve per cent and control in all relationships (Hinde, 1979; W ilhelm
con® dence intervals for both sets of correlations are & Parker, 1988). These tendencies m ay be shap ed by
outlined in Table 4. the stage of the life cycle and the speci® c nature of
the relationship.
The regard factor corresponds closely to sim ilar
M ean scores and distribution of the data concepts of attachm ent or care, and encom passe s
® lial gratitude or reciprocity and perceived closeness
M ean scores of the ® ve factors derived from the reli- or com patibility (C icerelli, 1983; Eisdorfer, 1991;
ability and validity studies are outlined in Table 3. H orl, 1989). In addition, item s such as I respect my
M ost of the data were skewed to som e extent (i.e. mother’s/father’s opinion or I know I can rely on my
skewness < 2.2), but only father control and, to a father to help me if I need him suggest that Kenny’s
lesser ex ten t, m oth er re sponsibility were skew ed (1987) Paren tal S u ppo r t d im en sio n, w h ich w as
signi® cantly (both positively). developed for students, extends beyond the early
child± parent relationship. The need to be appreciated
probably encom passes the continuing need for posi-
D iscussion tive appraisal by parents as well as the need to feel
that caretaking efforts have been acknowledged. This
T he Parent Adult± Child Questionnaire (PACQ ) was is co n sisten t w ith C icerelli’ s (1983) co n cept of
designed to m easure the relationship between adults reciprocity. W hile the balance of care m ay gradually
and their parents as perceived by the adult child. favo ur the parents, m utuality is im portant and adult
C onsiderably shorter than other measures of relation- children still turn to their parents for suppor t and
ship s which range from 25 to 50 item s, the instru- validation.
m ent is sim ple to com plete, efficient and app licable The responsibility factor includes parental behaviours
to both sexes. Its psych om etric properties suggest as well as the adult child’s feelings of guilt, burden

TABLE 3. Norm ative data obtained in reliability and validity studies

Test Retest Validity study

Time 1 Tim e 2

PACQ scale (total score) N Mean SD Mean SD N M ean SD

a
PACQ M
Regard (15) 24 7.5 3.7 7.5 4.0 25 11.7 3.2
Responsibility (24) 24 4.7 4.8 4.6 4.9 25 5.6 5.7

PACQ F b
Regard (12) 21 8.1 3.1 8.1 2.9 21 8.9 3.6
Responsibility (12) 21 1.1 1.6 1.3 1.4 21 2.1 2.4
Control (15) 21 2.3 3.2 2.1 2.9 21 3.1 4.7
a
PACQM refers to PACQ for mother. b PACQ F refers to PAC Q for father.
 34 C. Peisah et al.

TAB LE 4. Concurrent validityÐ correlation coefficients

Rater A interview score Rater A interview score Rater B interview score

versus versus versus
Rater B interview score PACQ score PAC Q score
d
PACQ scale N r sc CI N rs CI N rs CI
a
PACQM
Regard 25 0.93 0.84± 0.97 25 0.75 0.50± 0.88 25 0.77 0.53± 0.89
Responsibility 24 0.93 0.85± 0.97 25 0.90 0.79± 0.96 24 0.88 0.73± 0.95

PACQF b
Regard 21 0.94 0.85± 0.98 21 0.83 0.63± 0.93 21 0.85 0.67± 0.94
Responsibility 21 0.95 0.88± 0.98 21 0.79 0.55± 0.91 21 0.81 0.59± 0.92
C ontrol 21 0.89 0.75± 0.96 21 0.78 0.53± 0.91 21 0.71 0.41± 0.88
a b c d
PACQ M refers to PACQ for m other. PACQF refers to PACQ for father. r s = Spearman’s correlation coefficient. CI =
95% con® dence interval.

and protectiveness, and has elements of parenti® ca- evolution of parental power over the life cycle. Clearly
Downloaded by [University of Stellenbosch] at 18:47 04 October 2014

tion and the desire to com pensate for the parent’s an im portant factor in childhood (M inuchin, 1974;
past suffer ing (Boszorm enyi-N agy & Spark, 1984; M urphy et al., 1997), parental power dim inishes over
W hittick, 1989). Responsibility m ay vary according time (Benbow et al., 1990; C arter & M cGoldrick,
to the parent’s m ental state (Coyne, 1976; C oyne et 1980) as m ost parents gradually view their children
al., 1988; 1990), but whether feeling responsible for m o re as fr ien d s than su bo rd in ates (B lieszn er &
happiness is a norm ative concept in all ® lial relation- Mancini, 1987).Thus, the control factor m ay be m ore
ships or differentially applies to children of depressed useful in clinical populations than large com m unity
parents is uncertain. Responsibility as a construct screens, while the m ore norm al distribution of the
also is predicated upon culturally speci® c patterns of other factors su ggests that they m ay have w ider
kin relations and contact, the relevance of which will ap plicatio n . T h ese issu es m ig ht be clar i® ed b y
be lim ited by circum stances. For exam ple, feeling adm inistering the instrum ent to a larger com m unity
guilty about not seeing parents for a week will be sam ple.
relevant only for fam ilies living in reasonable proxim ity We concede som e lim itations to the developm ent
but not cohabiting. M oreover, the rapidly changing of the PAC Q. Firstly, the entire sam ple m ay have not
nature of fam ily structure, kin relationships and provi- been sufficiently heterogenous due to the use of
sion for the aged w ill alter m any of these concepts, convenience samples rather than probability sam ples
especially in western industrialized societies where of a known population. However, the fact that sim ilar
fam ily m em bers are becom ing increasingly inde- results were obtained in two different convenience
pendent (Kornhab er & Woodward, 1981). Changes sam ples and the sim ilarity between the dim ensions
are taking place also throughout East Asia and inter- identi® ed here and those identi® ed in other studies of
generational relations are approxim ating those of interpersonal relationships suggests the ® ndings are
western cultures, i.e. they are m ore affection based, ro bust. S econ d ly, th ere w as a con cer n th at the
convenience orientated and freer from intense ® lial `double-barrelled’ nature of som e of the items (e.g. I
piety (Sung, 1998). Interestingly, constructs such as feel that I should take care of my mother because she has
responsibility, affe ction and repaym ent identi® ed by suffered so much in her life) introduced am biguity, but
Sung (1998) as prim ary and shared between cultures this did not occur in practice. Thirdly, a difficulty
as diverse as Korean and Am erican, were sim ilar to with the principal com ponents approach towards
those identi® ed in this study. However, Sung (1998) qu estionn aire design is that discrim inatory item s
notes there is greater em phasis on respect, fam ily which are m ore unusual tend to em erge in the factor
harm ony an d sacri® ce in K orean cu lture. T hus, analysis, while m ore comm only held `everyday’ beliefs
although the PACQ was tested on a som ewhat cultur- about parents may be lost in the analysis. For exam ple,
ally homogeneous sam ple (its diversity lim ited by the M y mother is my best friend loaded highly on the regard
criterion of having adequate English to participate), scale, yet intuitively the item I treat my mother as a
the factors identi® ed rem ain relevant to other cultures friend m ay be m ore generally app licable. Such ques-
and the PACQ will provide fur ther scope to com pare tions are also potentially affe cted by response bias, a
® lial relationships across cultures. problem w ith any m easure of individuals’ percep-
T he control factor (relevant for fathers only) is akin tions of their relationship with their parents (W ilhelm
to that identi® ed in analyses of earlier relationships & Parker, 1988). Feelings of disloyalty and guilt may
(Parker, 1983; Siegelm an & Roe, 1979) but without in¯ uence responses to questions about parents as
the protection element. T his, together with the more `best friends’ or respecting their opinion.
positive skew of the factor, suggests the possibility of Fourthly, the effects of child gender, parent gender
a pathological entity, which m ay re¯ ect the naturalistic and their interaction are problem atic. A m easure
 The parent adult± child relationship questionnaire (PACQ) 35

applicable to sons and daughters m ay lose gender their husbands (Tower & K asl, 1996). The PACQ
speci® c item s and blur sex differences, but it allows will facilitate extension of such studies exam ining the
direct com pariso ns, de® nes central constructs better in teraction between ® lial variables, gen d er, an d
and has m ore utility in applied research. In this study, depressive or other sym ptoms in elderly people.
differences between sons and daughters were m inor Potential research applications include exam ining
and could be ignored, while the differences between correlations and interactions with other m easures of
m others and fathers were m ajor. C ontrol emerged as a psych ological function, such as m easures of per-
separate factor for fathers only and several item s were sonality, interpersonal sensitivity or locus of control
speci® c to m others, such as the parenti® cation or the to exam ine the determ inants of relationships. O ther
appreciation questions, or to fathers such as can rely on research applications include the assessm ent of parent
to help if needed. O ur ® ndings that the concepts of adult± child relations in speci® c clinical populations,
regard an d responsibility are applied d ifferently to such as adult children of parents w ith psychiatric
m others and fathers and that parental differences illness or children of survivors of traum a.
outweigh child gender differences accords with reports The PACQ m ay also prove a useful clinical instru-
by others (D e Vaus, 1994; M ain et al., 1985). Finally, m ent. It is widely acknowledged that assessm ent of
the PACQ ’s applicability across a wide range of the the aging patient m ust include assessm ent of the
life cycle is supported by the failure to ® nd correla- fam ily. The PACQ will provide an efficient estim ation
tions between age and factor scores. Potential differ- of the nature of the ® lial relationship during the assess-
Downloaded by [University of Stellenbosch] at 18:47 04 October 2014

ences between relationships w ith older and younger m ent phase, guiding the clinician to future strengths
parents were not borne out, suggesting that the PACQ or weaknesses in the patient’s support system . In a
is not `life-stage dependent’ . m ore form al sense, in either individual or fam ily
Future work will be directed at providing further psych otherapy it m ight be used as an assessm ent
validation of the instrum ent as well as testing its in str u m en t to u n d er sta n d fa m ily relatio n sh ip s.
Problem areas of relationships such as issues of control
application in various populations. We will extend
or responsibility m ight be identi® ed and used to plan
validity testing by com paring scores with objective
treatm ent goals, develop intervention strategies or
observer assessm ents of the adult± child relationship.
explore the outcom e of therapy. O ther general uses
A com pariso n with the parent’s perception of the
m ight be to identify or predict psychological m orbidity
relationship m ight be interesting but would not be a
or potential elder abuse in adult children forced into
useful assessm ent of validity because of the question-
the role of caregivers. For exam ple, a pattern of low
able `objectivity’ of fam ily m embers as raters (W ilhelm
regard and high responsibility (som etimes m issed or
& Parker, 1988). C oncurrent validity could be tested
denied on the ® rst interview ) m ight rapidly identify a
by com paring with instrum ents such as the Cam ber-
child caregiver in need of respite or future m onitoring.
well Fam ily Interview (Leff & Vaughn, 1985; Vaughn
Furtherm ore, such patterns m ight confer particular
& L eff; 1976), the Adult Attachm ent Scale (West et
vulnerability to depression, anxiety or pathological
al., 1987) and the Attachment Style Inventory (Sper-
grief when the parent dies.
ling et al., 1992). Continuity and evolution of relation-
In conclusion, the PACQ (Appendix 1) is a useful
ship s m ight be explored by com paring results on the
addition to the literature in providing a sim ple and
PAC Q w ith m easures of the early relationship such
effective m eans of m easuring constructs central to
as the Parental Bonding Instrum ent (Parker, 1983). the parent adult± child relationship with potential
A body of research regarding the in¯ uence of primary app lication in both clinical and non-clinical popula-
fam ily relationships on the mental health of elderly tions.
people is emerging. Vulnerability in caregivers and
aspects of fam ily relationships are associated with
dysfunctional behaviour in patients with Alzheim er’s Acknowledgem ents
disease (Vitaliano et al., 1993), elder abuse (Sadler et
al., 1995) and poor recovery, relapse (Hinrichsen & M rs Sue Sproule entered the data and M s Jenny
Hernandez, 1993) and suicidality (Zweig & Hinrichsen, Grice assisted with the preparation of the m anuscript.
1993) in depressed elderly subjects. Caregivers have Professo rs G ordo n Parker, Wayne H all an d K ay
been exam ined largely as a heterogenous group of W ilhelm assisted with the m ethodology. D rs John
spo uses an d children , and in terperson al facto rs Brennan, C aroline Q uadrio and Brian D raper prof-
m easured in a general way. As a speci® c, well-derived fered suggested item s. D rs M ichael D avis, An na
m easure for assessing the parent± child relationship, G ranger, D avid L im , M ichael Lowy, Alan Pollack,
the PACQ perm its exam ination of the interaction D ick Q uan , A ndrew Sm all, D avid Tan usdisastro,
between variables related to the parent adult± child Stephanie Way and Shane Woods allowed access to
relationship and m ental health variables of the parent. their patients for collection of data. T he volunteers in
For exam ple, it has been found that elderly husbands the validity study gave their time to be interviewed,
were less depressed when their wives were emotion- often about distressing issues. D r Peisah was sup-
ally in d ep en d en t, w hile eld erly w ive s were less ported by an Eli Lilly Fellowship in the ® rst year of
depressed when they felt em otionally im por tant to the project. All are thanked w ith gratitude.
 36 C. Peisah et al.

References F ITTING , M ., R ABINS , P., L UC AS , M .J. & E ASTHAM , J. (1986).

C aregivers of d em en tia patien ts. A com pariso n of
A R M SD EN , G.G. & G RE EN BE RG , M .T. (1987). The Inven- husbands and wives. The Gerontolog ist, 26 , 248± 252.
tory of Parent and Peer A ttachm ent: individual differ- F REU D , S. (1961). Some psychical consequences of the
ences and their relationship to psycholog ical well- anatomical distinction between the sexes. In: J. S TRAC HEY
being in adolescence. Jo ur na l of You th and A dolescen ce, (Ed.), The standard edition of the complete psycholog ical
16 , 427± 454. works of Sigm und Freud, Volume 9. London: Hogarth
A R R IND EL L , Q.A ., P ER R IS , C ., P ER R IS , A.M ., E ISE M AN N , M ., Press.
VA N D ER E N DE , J. & VO N K N O R R IN G , L. (19 86). C ross- F R IDAY, N. (1981). M y M other, my self . Glasgow: Williams
national invariance of dimensions of parental rearing Collins.
behaviour: comparison of psychometric data of Swedish
H AZAN , C. & S H AVER , P. (1987). Romantic love conceptual-
depressives and healthy subjects with Dutch target ized as an attachment process. Jour nal of Personality and
ratings on the EM BU. B r itish Jou r n al of P sych iatr y,
Social Psychology, 52 , 511± 524.
148 , 305± 309.
H EN DER SON , S., B YR NE , D.G. & D U NC AN -JONES , P. (1981).
B ARU C H , G. & B AR NETT , R.C. (1983). Adult daughters’
Neurosis and the socia l env ironm ent . Canberra: Academic
relationships with their m others. Jour na l of M arr iage and
Press.
the Fam ily , 45 , 601± 606.
H IND E , R.A. (1979). Towa rds und erstan ding relationships .
B ENBOW , S., E G AN , D., M AR R IOTT , A., TREGAY, K. ET AL .
London: Academic Press.
(1990). U sing the family life cycle with later life families.
H INR IC H SEN , G.A. (1991). Adjustment of caregivers to
Jour na l of Family Therapy, 12 , 321± 340.
d epressed o lder ad u lts. P sy ch olo g y & A g in g , 6 (4),
B E R M AN , W.H ., H EIS S , G.E. & S P ER LIN G , M .B. (1994).
631± 639.
Downloaded by [University of Stellenbosch] at 18:47 04 October 2014

M easuring continued attachment to parents: the C on-

H INR IC HSE N , G.A. & H ER N AND EZ , N.A . (1993). Factors
tinued Attachment Scale-Parent Version. Psychological
associated with recovery from and relapse into m ajor
Reports, 75 , 171± 182.
depressive disorder in the elderly. Am erican Jour nal of
B LIESZN ER , R. & M ANC INI , J. (1987). Enduring ties: older
Psychiatr y, 150 , 1820± 1825.
adults’ parental role and responsibilities. Fam ily R ela-
H ORL , J. (1989). Looking back to caregiving: ® ndings from
tions, 36 , 176± 180.
case studies in Vienna, Austria. Jour nal of Cross Cultura l
B OSZO R M ENYI -N AG Y, I. & S PARK , G.M . (1984). Invisible loya l-
G erontology, 4 , 245± 256.
ties . New York: Brunner M azel.
H OROWIT Z , A. & S C HINDEL M AN , L.W. (1983). Reciprocity
B OW LBY, J. (1979). The m aking and breaking of affectional
and affection: past in¯ uences on current caregiving.
bonds . London: Tavistock.
Jour nal of G erontolog ical Socia l Work, 5 , 5± 20.
B OW LBY, J. (1980). A tta chm ent and loss.Volum e 3: Loss: stress
K ENNY , M . (1987). The extent and function of parental
and depression . New York: Basic Books.
B RODATY , H. & H ADZI -P AVLOVIC , D. (1990). Psychosocial attachment among ® rst-year college students. Jour na l of
effects on carers of living with persons with dementia. Youth and A dolescence, 16 , 17± 29.
A ustra lian & N ew Z ealand Jou r na l of Psychiatr y, 24 , K OR NHABER , A. & WO ODWARD , K.L. (1981). G randparents/
351± 361. grandchildren: the vital connectio n . New York: Anchor Press/
B RODY , E. (1981). Women in the middle and family help to Doubleday.
other people. The Gerontologist, 21 , 471± 480. L EFF , J. & VAU GH N , C . (1985). E xpressed em otion in families .
C AIRL , E., P FEIFFER , E. & K ELLER , D.M . (1985). Geriatr ic New York: G uilford Press.
assessm ent testing an d evaluatio nÐ C ost of C are Index . M AIN , M ., K APLAN , N. & C ASSIDY, J. (1985). Security in
Florida: Suncoast Gerontology Centre. infancy, childhood and adulthood: a m ove to the level of
C ANTOR , M .H. (1983). Strain among caregivers: a study of representation. In: I. B RETHER TON & E. WATERS (Eds),
experience in the U nited States. The Gerontolog ist, 23 , M onog raphs of the Society for Research in C hild Develop-
597± 604. m ent, 50 (1± 2, Serial No. 209), 66± 104.
C AR TER , E. & M C G O LDRICK , M . (1980). The family life cycle: M ANC INI , J. & B LIESZN ER , R. (1989). Aging parents and
a framework for fam ily therapy . New York: Gardner. adult children: research themes in intergenerational rela-
C H O D O R O W , N. (197 8). T he reprod u cti on of m oth er in g: tions. Jour na l of M arr iage and the Fam ily, 51 , 275± 290.
psych oana lysis and the sociolog y of gender . University of M INU CHIN , S. (1974). Fam ilies and family therapy . Bristol:
Berkeley: California Press. Tavistock.
C IC ERELLI , V.G. (1983). Adult children’s attachment and M O OS , R.H. & M O OS , B.S. (1981). Fam ily E nvironm ent Scale
helping behaviour to elderly parents: a path model. Jour nal m anua l . Palo Alto, CA: Consulting Psychologists Press.
of M ar riage and the Fam ily, 45 , 815± 825. M O RR IS , R.G., WOO DS , R.T., D AVIES , K.S. & M OR R IS , L.W.
C OYNE , J.C. (1976). Towards an interactional description of (1991). G en d er d iffe ren ces in carers of dem en tia
depression. Psych iatr y, 39, 28± 40. sufferers. B r itish Jour na l of Psych iatr y, 158 (Suppl. 10),
C OY NE , J.C ., E L LA RD , J.H . & S M ITH , D.A .F. (1990). Social 69± 74.
support, interdependence, and the dilemmas of helping. M U R PHY, E., B REW IN , C .R. & S ILKA , L. (1997). The assess-
In: B.R. S A RA SO N , I.G. S AR AS O N & G.R. P IERC E (eds), ment of parenting using the Parental Bonding Instru-
S ocia l sup port. A n intera ction al view . New York: John ment: two or three factors? Psychological M edicine, 27 ,
W iley. 333± 342.
¾ TM
C OYNE , J.C., WO R TM AN , C.B. & L EHM AN , D.R. (1988). The N OR U SIS , M.J. & SPSS I NC . (1993). SPSS for W indow s
other side of support: emotional overinvolvement and Release 6.0. C hicago: SPSS Inc.
m iscarried helping. In: B.H. G O TTLIEB (Ed.), M arshalling O’ C ON NOR , P. (1990). The adult mother/daughter relation-
socia l sup port . Beverley, CA: Sage. ship: a uniquely and universally close relationship?
D E VA U S , D. (19 94). L ett in g go . M elb ou rn e: O xford Sociological Review, 38 , 293± 323.
U niversity Press. P A RK ER , G. (19 83). Pa renta l over protecti on : a r isk factor
E ISDO RFER , C. (1991). Caregiving: an emerging risk factor in p sychosocia l d evelo p m en t . N ew York: G ru n e an d
for em otional and physical pathology. B ulletin of the Stratton.
M enninge r Clinic, 55 , 238± 247. PERR IS , C., JACOBSSON , L., L INDSTR OM , H., VON K NO RR ING , L.
E PSTEIN , N.B., B ALD WIN , L.M . & B ISHO P , D.S. (1983). The & PER RIS , H. (1980). Development of a new inventory
M cMaster Family Assessment Device. Jour nal of M arriage for assessing memories of parental rearing behaviour.
and Family Therapy, 9 (2), 171± 180. Acta Psych iatrica Scandinavica, 61 , 265± 274.
 The parent adult± child relationship questionnaire (PACQ) 37

P HILLIPS , S. (1991). B eyond the myths. Sydney: Hampden TU DOR , A., JACKSON , H.J. & PATTISO N , P. (1995). Deter m ina nts
Press. of psychopatholog y in carers of relatives with dem entia . 7th
R ICH , A. (1977). O f women bor n . New York: Bantam. International Psychogeriatric Association Congress. Spirit
R O BINSON , B. & T HU R NER , M . (1979). Taking care of aged in Ageing. Sydney.
parents: a fam ily cycle transition. The Gerontologist, 19 , VAU G HN , C.E . & L EFF , J.P. (1976). The in¯ uence of family
586± 593. and social factors on the course of psychiatric illness.
S AD LER , P., K U R RLE , S. & C AM ERO N , I. (1995). Dementia B ritish Jour na l of Psychiatr y, 129 , 125± 137.
and elder abuse. Austra lian Jour nal on Ageing , 14 , 36± 40. V ITALIANO , P.P., R U SSO , J.,YOUNG , H.M., B ECKER , J. & M AIU RO ,
S C H A E F E R , E.S . (19 65). A con ® gu ration al analysis of R .D. (1991 ). The screen for caregiver burden. T he
children’ s reports of parental behaviour. Jo ur n al of G erontolog ist, 31 , 76± 83.
Consultin g Psych olog y, 6 , 552± 557. V IT ALIANO , P.P., YO U NG , H.M ., R U SSO , J., R O M A NO , J. &
S CHENE , A.H., TESSLER , R.C. & G AM AC HE , G.M . (1994). M AG ANA -A M ATO , A. (1993). Does expressed emotion in
Instruments measuring fam ily or caregiver burden in s po us es pred ict s ub sequ en t p ro blem s a m o n g care
severe mental illness. Socia l Psychiatr y and Psychiatric recipients with Alzheimer’s disease? Jour nal of Gerontolog y,
Epidem iolog y, 29 , 228± 240.
48 , 202± 209.
S CHO FIELD , H., H ER M AN , H., B LOCH , S. & S ING H , B. (1995).
WEST , M., S HELDON , A. & R EIF FER , L. (1987). An approach
H ealth and well-being of fam ily carers: a com parative stud y.
to the delineation of adult attachment: scale develop-
7th International Psychogeriatric Association Congress.
ment and reliability. Jour nal of Ner vous and M ental Disease,
Spirit in Ageing. Sydney.
175 , 738± 741.
S EELBAC H , W.C. & S AU R , W.J. (1977). Filial responsibility
W H ITBEC K , L., H OYT , D.R. & H U CK , S.M . (1994). Early
expectations an d m orale am ong aged paren ts. T h e
family relationships, intergenerational solidarity, and
Downloaded by [University of Stellenbosch] at 18:47 04 October 2014

Gerontologist, 17 , 492± 499.

S IEG ELM AN , M. & R O E , A. (1979). M anua l:The Parent± Child support provided to parents by their adult children.
Relations Questionn aire-II . New York: City College of New Jour na l of Gerontolog y, 49 , 585± 594.
York. W H ITTIC K , J.E. (1989). Dementia and mental handicap:
S PERLIN G , M .B., B ER M AN , W.H. & F AG EN , G. (1992). Clas- attitudes, emotional distress and caregiving. B ritish Jour nal
si® cation of adult attachment: an integrative taxonomy of M edical Psych olog y, 62 , 181± 189.
from attachment and psychoanalytical theory. Jour nal of W ILHELM , K. & PARKER , G. (1988). The development of a
Personality A ssessm ent, 59 , 239± 247. m easure of intimate bonds. Psych olog ical M edicine, 18 ,
S U NG , K. (1998). Filial piety in modern times: timely adapta- 225± 234.
tion and practice patterns. Austra lasian Jour nal on Ageing, Z ARIT , S.H., R EEVES , K.E. & B ACH -PETER SON , J. (1980). Rela-
17, (Suppl.) 88± 92. tives of the impaired elderly: correlates of feelings of
TABACH NIC K , B.G. & F IDELL , L.S. (1989). U sing m ultiva riate burden. The G erontologist, 20, 649± 655.
statistics . Northridge: HarperCollins Z W EIG , R.A. & H INR IC HSEN , G.A. (1993). Factors associated
TOWER , R.B. & K ASL , S.V. (1996). G ender, marital close- w ith suicide attem pts by depressed older adults: a
ness, and depressive symptoms in elderly couples. Jour nal prospective study. Am er ican Jour nal of Psychiatr y, 150 ,
of G erontolog y, 51 , 115± 129. 1687± 1692.

A ppendix 1
The Parent Adult± C hild R elationship Q uestionnaire

This form contains a number of statements about relationships between adults and their parents. Please read each statement
carefully, and decide how well it describes your relationship with your mother and your father as it has been over the last year.
Please circle the m ost appropriate response of the four possible responses.

Ver y true
Tick if you feel that the statement describes your relationship with your parent accurately.

M oderately tru e
Tick if you feel that the statement describes your relationship with your parent for the m ost part.

S om ew hat true
Tick if you feel that the statement describes your relationship with your parent to a small extent.

N ot true at all
Tick if you feel that the statement does not describe your relationship with your parent at all.

R elationsh ip with M other (PAC Q M ) Very true M oderately So m ewhat N ot true at

tru e true all

1. I look forward to seeing my mother (3) (2) (1) (0)

2. I feel responsible for my mother’s happiness. (III) (II) (I) (0)
3. I feel that I should take care of my mother because she (III) (II) (I) (0)
has suffered so much in her life.
4. My mother is my best friend. (3) (2) (1) (0)
5. My mother’s difficulty in making decisions has been a (III) (II) (I) (0)
burden on me.
6. My mother shows her appreciation of me. (3) (2) (I) (0)
 38 C. Peisah et al.

7. I am the only one my mother can rely on. (III) (II) (I) (0)
8. M y m other thinks I am good in a crisis so she calls on (III) (II) (I) (0)
m e all the time.
9. I respect my mother’s opinion. (3) (2) (1) (0)
10. I feel that I have to protect my m other. (III) (II) (I) (0)
11. M y m other relies on m e too much. (III) (II) (I) (0)
12. I am glad to be able to repay my m other for all the love (3) (2) (1) (0)
and care she gave me as a child.
13. I feel like I parent my m other. (III) (II) (I) (0)

Scoring: Rega rd Scale Arabic numerals

Responsibility Scale Roman numerals

R elation ship w ith Fath er (PAC Q F) Very true M oderately S om ew hat N ot tru e at
true tru e all

1. If I don’ t do things my father’s way he will nag me. (3) (2 ) (1) (0 )

2. I respect my father’s opinion. (3) (2) (1) (0)
3. Something will happen to my father if I don’ t take care (III) (II) (I) (0)
of him.
Downloaded by [University of Stellenbosch] at 18:47 04 October 2014

4. I feel that my father tries to m anipulate me. (3) (2 ) (1) (0 )

5. I look forward to seeing my father. (3) (2) (1) (0)
6. I feel responsible for my father’s happiness. (III) (II) (I) (0)
7. M y father tries to dominate me. (3) (2 ) (1) (0 )
8. I feel that my father makes too many demands on m e. (3) (2 ) (1) (0 )
9. I know I can rely on my father to help me if I need (3) (2) (1) (0)
him.
10. If I don’ t see my father for a week I feel guilty. (III) (II) (I) (0)
11. I don’ t discuss m uch with my father because I’ m afraid (3) (2 ) (1) (0 )
of being criticized.
12. I don’ t mind putting myself out for my father. (3) (2) (1) (0)
13. M y father thinks I’ m good in a crisis so he calls on m e (III) (II) (I) (0)
all the time.

Scoring: Rega rd Scale Arabic numerals

Responsibility Scale Roman numerals
Control scale Italic Arabic numerals