Journal of Aging Studies 20 (2006) 177 – 191

www.elsevier.com/locate/jaging

Using adult attachment theory to differentiate adult children's

internal working models of later life filial relationships
D. Shemmings
Middlesex University, London, UK, EN3 4SA
Received 29 September 2004; received in revised form 15 May 2005; accepted 10 June 2005

Abstract

During the past fifty years, research based on attachment theory has found that when relational partners' attachment systems are
activated, significant differences emerge between the ways individuals respond to each other. These different attachment styles are
related to the ways individuals characterise and conceptualise close relationships generally, referred to as ‘internal representations’.
Internal representations of close relationships depend heavily upon whether individuals have a secure or insecure attachment style.
Until recently, most attachment-based research has focused either on the parent–child relationship during infancy, or on adult
romantic relationships. Attachment researchers are now turning their attention to the parent–‘child’ relationship during the later
stages of life. Later life filial relationships are of intrinsic interest to attachment researchers because they concern the same adults
who were instrumental in forming the attachment organisation of the young child. This study considers filial attachments from the
perspective of the adult ‘child’. Twenty-four participants were selected using the Attachment Style Questionnaire (ASQ) to include
equal numbers of the three main attachment organisations. Six robust factors emerged, accounting for 71% of the variation.
Confident Resolution and Resolved Yearning incorporated the secure attachment organisation. Distant Irritation and Dutiful Loyalty
captured the insecure-avoidant style, with Unresolved Yearning and Entangled Resentment comprising insecure-ambivalent
individuals.
© 2005 Elsevier Inc. All rights reserved.

Keywords: Later life filial relationships; Adult attachment theory; Q-methodology; Internal representations of close relationships

1. Introduction

In addition to needing practical help and financial security during later life, individuals also tend to value the
emotional support of close family and friends (Cicirelli, 1983; Seeman & Crimmins, 2001). This appears to be
especially true for the ‘oldest-old’ cohort (Krause, 2004). Research based on attachment theory (Bowlby, 1969, 1973,
1979) has consistently produced robust findings demonstrating that emotional availability towards others is strongly
associated with early attachment experiences (Van IJzendoorn & Bakermans-Kranenburg, 1997). Consequently, the
emotional availability and interplay of attachment patterns between adult ‘children’ and their parents is of particular
interest, because the latter usually performed the key roles in the attachment development of the former.
For the most part, however, gerontological research has tended to overlook the differential effects of secure and
insecure attachment organisation on relationships during later life (Bradley & Cafferty, 2001; Consedine & Magai,

E-mail address: d.shemmings@mdx.ac.uk.

0890-4065/$ - see front matter © 2005 Elsevier Inc. All rights reserved.
doi:10.1016/j.jaging.2005.12.001
178 D. Shemmings / Journal of Aging Studies 20 (2006) 177–191

2003; Kraus & Haverkamp, 1996; Sorensen, Webster, & Roggman, 2002). Given that prevalence studies examining the
distribution of attachment styles in large populations (e.g. Mickelson, Kessler, & Shaver, 1997: n = 8098) regularly
report a 60 : 40 split between attachment security and insecurity respectively, potentially significant findings may have
been cancelled out as a result of not differentiating later life relationships according to attachment organisation. And
because attachment insecurity subdivides further into the relatively evenly distributed ‘avoidant’ and ‘ambivalent’
styles, each one recognisable by marked different internal representations of relationships and attachment behaviour,
the argument for differentiation is strengthened (Bowlby, 1979; 1988).
During infancy, attachment styles operate pre-consciously and instinctually, but as the growing child acquires
cognitive skills s/he also develops internalised expectations – ‘internalised mental representations’, or ‘internal
working models’ (Bretherton & Mulholland, 1999) – about how s/he should behave, and how others are likely to
respond, when the attachment system is activated. Such activation occurs in situations when the child is anxious,
experiences separation or needs to feel close to an attachment figure. As a result, how individuals perceive, process and
eventually resolve stressful events and circumstances are closely connected to the developing infant's early formative
experiences of how they first learned to achieve proximity to primary caregivers (Shaver & Mikulincer, 2004).
Consequently, in adulthood, attachment theory and research routinely differentiates individual close relational
preferences (Main, Kaplan, & Cassidy, 1985), affect regulation (Feeney, 1995), and information processing (Fraley &
Shaver, 1997) into one of the three attachment ‘styles’.
Attachment researchers have developed a growing interest in later life filial relationships (Bradley & Cafferty, 2001;
Consedine & Magai, 2003; Magai, Consedine, Gillespie, O'Neal, & Vilker, 2004) precisely because they usually
involve the same individuals who were present at birth; but, as adults, each is able to give, receive and reciprocate
within the relationship, and therefore different attachment dynamics are produced. Studies have already been
undertaken into attachment and felt obligation/subjective burden (Cicirelli, 1993; Finley, Roberts & Banahan, 1988;
Hamon & Blieszner, 1990; Stein et al., 1998), and caregiving and attachments between adult offspring and their older
parents (Antonucci, 1994; Cicirelli, 1981, 1983, 1993; Crawford, Bond, & Balshaw, 1994; Sorensen et al., 2002) —
some of which have focused upon caregiving situations involving people with dementia (e.g. Miessen, 1993; Steele,
2000). Each study confirms the importance of distinguishing between secure and insecure attachments when analysing
outcomes.
This article reports on research exploring adult children's attachment styles and internal representations of their
relationship to their parent during later life. I conclude by discussing some of the implications for adult children's
emotional availability in situations when their attachment system is activated.

1.1. Differences in individuals' attachment organisation

Having broken away from a prevailing psychoanalytic tradition relying on instinctual drive-reduction mechanisms
to explain infant bonding, John Bowlby went on to articulate two complementary components of the attachment
process. Initially functioning to protect the child from harm or threat, a primary attachment figure also provides a secure
base from which the child can explore the environment (Bowlby, 1969, 1973, 1979); but it was the interdependence of
security and exploration that constituted Bowlby's focal discovery. The growing infant's facility for creatively
discovering its immediate world of people and environment varies directly with the extent to which s/he can routinely
‘touch base’, knowing that s/he will not be rebuffed (Cassidy & Shaver, 1999). The gentle guidance and unobtrusive
supervision offered by a sensitive primary caregiver helps create the conditions for the child's subsequent forays into a
potentially dangerous environment, whereas consistent, or inconsistent, rejection by a primary caregiver disrupts this
dynamic and results in the child experiencing insecurity about how best to maintain proximity to the caregiver, which in
the early years of an infant's life is essential to its survival. Furthermore, ‘the manner in which the caregiver responds,
or does not respond, to the child's emotional signals constitutes one of the most influential factors in the quality of
attachment that develops’ (Magai et al., 2004, p. 390).
Three ‘styles’ of attachment were found through extensive observation which consistently accounted for a range of
attachment-related behaviour. They comprised one secure and two insecure patterns — the anxious–avoidant and
anxious–ambivalent styles (Sperling & Berman, 1994). The Secure pattern is characterised by children who are able to
depend upon and trust adults and who, simultaneously, develop perceptions of self as lovable. Children who develop an
Avoidant pattern turn away from others when under threat because, from an early stage, primary caregivers consistently
discouraged or rejected the display of feelings. In times of crisis they de-activate attachment-related behaviour by
 D. Shemmings / Journal of Aging Studies 20 (2006) 177–191 179

downplaying emotional responsiveness. The Ambivalent style results from inconsistent and unpredictable early
caregiving and leaves the child with considerably mixed feelings about relationships. They tend to worry that others do
not really care for them, they are difficult to please or placate and are often viewed by other people as clingy and
‘needy’. They over-activate attachment-related behaviour in situations when they feel distressed, which are frequent
and experienced as intense.
Internalised representation of close relationships, or ‘internal working models’ (Craik, 1943), emerge from the
biography of the child's ‘event-based relationships’ (Main et al., 1985) with primary caregivers. They are the result of
accumulated memories of successive attempts to achieve proximity when under threat. They operate like an ‘internal
guidance system’ (Belsky, 2002) with regard to relationships and they are thought to begin forming between 9 and 18
months (Diamond & Blatt, 1994). They continue developing throughout the life span (hence, the term ‘working’) into
complex mental representations of how ‘self’ and ‘other’ interact.
Internalised relational representations are templates of expectations about how other people will react when an
individual's attachment system is activated. They are intrinsically bound up with attachment styles. In addition, as
Berlin and Cassidy (1999) point out, ‘what distinguishes attachment theory from … other theories is the specificity of its
predictions about individual differences and its arguments that mental representations … underlie the associations
between early attachments and subsequent close relationships’ (p. 689). Consequently, internal representations ‘reflect
not an objective picture of the “parent”, but rather the history of the caregiver's actions or intended actions with/toward
the attachment figure’ (Main et al., 1985, p.85).
Historically, adult attachment theory proceeded from work on infant development, firstly, into adolescence, and then
accelerated dramatically when the focus shifted to the study of close relationships in adulthood. A comprehensive
literature now exists (for detailed summaries, see Cassidy & Shaver, 1999; Feeney & Noller, 1996; Sperling & Berman,
1994) which explores topics such as romantic relationships, relationships between older people, and loss and
bereavement.
Consistent among the findings are trends that secure attachment experiences are associated with positive
affect and well-being (Mikulincer & Florian, 1998), lower levels of depression (Roberts, Gotlib, & Kassel,
1996), as well as reduced loneliness (Hazan & Shaver, 1987), anxiety (Mikulincer, Florian, & Weller, 1993)
and hostility (Mikulincer, 1998). Additionally, securely attached individuals are able to rely on an open, flexible
style of emotion regulation when facing relational stress (Magai, Hunziker, Mesias, & Culver, 2000). On the
other hand, avoidantly attached individuals tend to short-circuit negative emotion from consciousness (Magai et
al., 2000), whereas ambivalent individuals have a heightened style of affect regulation (Shaver & Mikulincer,
2002).
There are some similarities between the concept of ‘ambivalence’ within attachment theory and that of
‘psychological ambivalence’ but Connidis and McMullin (2002) point out that ‘ambivalence’ has been,
‘conceptualised at two levels: at the structural level, sociological ambivalence has been viewed as contradictory
normative expectations that occur in institutional resources and requirements … At the individual level,
psychological ambivalence has been referred to as contradictions that are primarily subjective and evident in
cognitions, emotions, and motivations …’ (Connidis & McMullin, 2002, pp.558–559)
The approach adopted in the present study is clearly at the ‘individual’ level, as it explores internal representations
of relationships, rather than social structures. Although disagreement exists about whether ‘ambivalence’
characterises all relationships at times (Connidis & McMullin, 2002) or whether it is a feature only experienced
within specific dyadic relationships (Luescher & Pillemer, 1998), from an attachment perspective ‘psychological
ambivalence’ is a distinct relational pattern among individuals noticed when their attachment system is activated.
Interestingly, although he often refers to a ‘significant proportion’ of relationships as characterised as ‘ambivalent’
(e.g. Pillemer & Suitor, 2002), a closer examination of Pillemer's research reveals that the actual proportion is
between 15–20%, the figure reported routinely in attachment-based research (see, for example, Mickelson et al.,
1997).

1.2. Universal attachment needs

Even though attachment differences exists between individuals, they also share common attachment responses,
and these are especially noticeable during later life. Attachment theory has identified four universal patterns of
180 D. Shemmings / Journal of Aging Studies 20 (2006) 177–191

human need that arise when individuals are placed under stress or when they experience anxiety. Distinctive, but
overlapping, responses have been consistently noted when attachment security is threatened (Bowlby, 1973, 1980;
Feeney & Noller, 1996): we have a need to seek proximity to another individual, whom we perceive capable of
providing a temporary safe haven. This proximity seeking behaviour may be ‘real’, in the sense that we want to be
physically close to the attachment figure, or it can be ‘virtual’ — for example, we might imagine the person acting in
a calm and soothing manner, but who nevertheless is physically absent. Until the threat has passed, individuals will
exhibit separation protest if the attachment figure withdraws support prematurely, or if they withhold it. When the
threat has passed, individuals will, ideally, use the safe haven as a secure base from which to re-engage more fully
with the outside world.
Later life filial relationships hold the potential for meeting all four criteria of attachment relationships – proximity
seeking, separation protest, safe haven and secure base functions – for both relational partners, at times
simultaneously. Hence, relationships with aging parents are likely to qualify as ‘attachment relationships’ in the minds
of adult children, whether or not they are in regular contact and, in some cases, irrespective of the degree of
satisfaction they derive from them. From an attachment perspective these relationships are particularly intriguing when
they involve the same individuals who were present from birth because, during the later stages of the life course, roles
and reciprocities may become reversed, leading to new experiences for both parties. According to attachment theory, if
these experiences become stressful they will activate the attachment systems of both parties; again, possibly
simultaneously.

1.3. Attachment theory, emotion regulation and stressors in later life filial relationships

Because emotional support and availability are so important to most people, especially during times of stress or
anxiety, attachment researchers regularly explore how individuals experience and process negative emotion, and their
association with attachment styles and internal representations of close relationships. Because roles may sometimes be
reversed, expectations about affect expression, containment and reciprocity are likely impact upon their form and
trajectory of later life filial relationships.
The term ‘affect regulation’ refers to ‘acquired strategies of emotion self-management’ (Thompson, 1994, p.28)
and are most noticeable in stressful situations (Feeney, 1998). According to Consedine and Magai (2003)
‘individuals high in attachment security are said to have an open, flexible style of emotion regulation, which
means that they have access to a wide range of emotions and are able to adjust their emotional responses in ways
that are appropriate to prevailing situational contingencies’ (p.166). Secure attachment is also ‘ … indicative of the
ability to acknowledge and express emotional distress without becoming unduly disabled by it’ (p.178). Securely
attached people tend to be more tolerant of stressful events because they are less likely defensively to exclude
negative and potentially unpleasant feelings from consciousness; neither are they are overwhelmed by any ensuing
distress.
Research into individuals' expectations of how far each should restrict or express feelings indicates that, not only
does the non-declarative approach of avoidant individuals inhibit anger expression towards attachment figures, they
also generally report ‘controlling anger more than sadness, and sadness more than anxiety; furthermore, they
perceived their partners as behaving in this manner, and as wanting them to do the same’ (Feeney, 1995, p.145).
These findings have implications for affect regulation strategies within later life filial attachments. In particular, the
belief that the expression of emotion is or should be controlled on both sides may inhibit avoidantly attached adult
children from seeking out their parent as a safe haven (and vice versa). In situations involving filial caregiving for
example, affect suppression may become progressively problematic because, as Magai (1999) – referring to
Tompkins' (1993) pioneering theoretical insights into the nature of emotion – states, ‘humans have a fundamental
need to reduce affect inhibition, and that constricted affect results in “affect hunger” or the need to express affect’
(p.790). Similar problems are likely in the event of the excessive and unprocessed affect expression typical of
ambivalently attached individuals.
During the later stages of the lifespan, later life filial attachments may include, separately or in combination,
stressors such as the appearance of altered or unrecognisable changes in personality and behaviour within the
parent, thoughts of ‘anticipatory grief’ (Fulton, Madden, & Miniciello, 1996; Seale, 1998) and loss, as well as
feelings of obligation to offer more direct forms of personal care. From an attachment perspective, another
potential stressor is role reversal, described by Shaver and Mikulincer (2004) as ‘a process in which older adults
 D. Shemmings / Journal of Aging Studies 20 (2006) 177–191 181

with grown children rely on their children to serve some or all of the standard functions of attachment figures’.
They continue,

We suspect that, just as insecure attachment relationships create problems for both children and parental attachment
figures early in life, insecure relationships between older adults and their grown children, now serving as
attachment figures, also create serious … problems for both relationships partners. We would expect to see
continuing conflicts related to feelings and insufficient care, anger at past treatment, attempts to coerce and control
… (p. 461)

‘Stress’, however, is not necessarily problematic per se because an event ‘may have a very different psychological
impact on a person, depending on how the person copes with it’ (Zhang & Labouvie-Vief, 2004, p.431). Thus, Shaver
and Mikulincer (2004), also point to more creative aspects of role reversal which ‘may create an opportunity to heal old
wounds and reconstruct a relationship on more secure terms. This may be a point at which informed … intervention
could increase the likelihood of positive outcomes’ (Shaver & Mikulincer, 2004, p.461).
Later life filial attachments may also be experienced stressfully if both relational partners are negotiating different
developmental lifespan stages. For example, Krause and Haverkamp (1996) note that ‘whereas the middle years are a
time when the middle-aged child is challenged by the biological changes of ageing, and the demands of his or her
family, as well as financial obligations, the older parent is also in the process of experiencing changes as a result of
retirement, decreased health, or the death of a spouse’ (p.84). Additionally, for the parent, changes in former life roles
are often associated with a complementary increase in the importance of family ties as a source of meaning and
affirmation (Myers, 1988). Retirement, for example, may bring with it expectations in the parent's mind of increased
and more frequent family involvement, but such hopes may collide with an adult child's life, replete with career
aspirations and, in many cases, obligations to their own offspring.
Later life filial relationships can also create new and unfamiliar attachment-related experiences for adult sons and
daughters, such as additional stress for adult children (for which they may be especially unprepared) when persistent or
unreasonable separation anxiety is expressed by the parent. Demands by a parent that s/he is not left alone can become
physically wearing and emotionally taxing, and are magnified considerably if the parent is living with their son or
daughter.

1.4. Attachment styles and representations: fixed or variable?

Finally, although the question of intergenerational transmission of attachment (Bretherton & Mulholland, 1999; Van
IJzendoorn & Bakermans-Kranenburg, 1997) does not feature in the present study, it clearly has implications for filial
relationships. Although Bowlby ‘viewed attachment style to be changeable in response to new relationship
experiences, attachment representations are postulated to tend more toward assimilating, rather than accommodating
to, later experiences. Pre-existing attachment representations tend to carried over to new relationships by guiding and
influencing an individual's perception, evaluations and behaviour’ (Zhang & Labouvie-Vief, 2004, p. 420).
Questions about whether people exhibit different attachment behavioural patterns to different individuals when
attachment systems are primed, as well as the extent to which individuals change their attachment behaviour over time,
remain the subject of considerable debate among attachment researchers. Most studies indicate that attachment styles
remain relatively constant during infancy and toddlerhood, when there is an 80% correspondence between the
attachment style of the primary caregiver with that of the child.
During adolescence and adulthood, an increasing number and range of relational influences available to individuals
develops, beginning with friends and teachers and then usually including romantic and other intimate relationships.
Although fewer studies exist that have focused on the relative stability of attachment styles during adulthood, three
consistent findings emerge:

1. Attachment style contains elements of both stability and flexibility. In one study over a six year period Zhang and
Labouvie-Vief (2004) confirmed earlier studies that ‘consistently show that attachment style is relatively stable but
the degree of stability is debatable’ (Zhang & Labouvie-Vief, 2004, p. 420).
2. Attachment styles can change from an insecure style to an secure style, or vice versa, but they tend not to change
from one insecure style to another (Rothbard and Shaver (1994).
182 D. Shemmings / Journal of Aging Studies 20 (2006) 177–191

3. Adults, ‘on average, seem to become less anxiously attached with age, at least as measured by self-report scales’
(Shaver & Mikulincer, 2004, p.459).

2. Study aims

Findings are presented and discussed from a ‘theory-led qualitative study’ (Hayes, 2000) into adult sons' and
daughters' internal working models of their attachment toward the parent to whom they feel closest. The primary
research questions were:

1. How do adult children mentally represent their past, present and future attachment relationships with the parent (to
whom they feel closest)?
2. How are different internal representations and attachment styles likely to contribute to adult children's emotional
availability towards a parent when their attachment system is activated?

3. Method

Adult attachment research is often conducted with convenience samples of university students, but the typical age
profile of a student population would have been too low for the present study. Instead, volunteers were drawn from
health and social care professionals. The only restriction placed upon participants was for the parent selected to be a
surviving birth -, adoptive- or step-parent, and that s/he was part of the son's/daughter's primary caregiving system as a
young child.
Volunteers were partitioned equally into the three main attachment groups, using the Attachment Style
Questionnaire (ASQ; Feeney, Noller, & Hanrahan, 1994) which classifies attachment organisation into the secure,
avoidant and ambivalent styles.. The ASQ comprises 40 items to which participants are asked to indicate their level of
agreement on a 6-point Likert-type scale (1 = Totally Disagree; 6 = Totally Agree). The ASQ was selected because it
possesses robust psychometric properties, in terms of both validity and reliability (Shaver & Mikulincer, 2002). Twenty
four participants were recruited in this way. Ages ranged between 41 and 57 (M = 48.6) and the selected parent was aged
between 64 and 83 (M = 76.8). Because the study was not focused primarily upon caregiving, participants were
excluded if they were living with their parent, or if they were providing significant amounts of direct caregiving during
visits. Four of the parents were living in residential homes at the time of the study.

UNTRUE FOR ME TRUE FOR ME

Fig. 1. Q-sort grid.

 D. Shemmings / Journal of Aging Studies 20 (2006) 177–191 183

Two approaches were used to explore attachment representations. Firstly, a Q-sort of 44 statements (see
Addendum), devised specifically around later life filial attachments, was completed by each participant. Q-sorts are a
standard method of measurement used in attachment research (see Waters, 1995). The procedure involves participants
arranging into a pre-determined, normally distributed grid (see Fig. 1) cards containing the written statements (see
Table 1). As an example, statements 1 and 17 read as follows:
1. As we've got older my (mother/father) and I have got closer

17. I'd never try and talk to my (mother/father) about a personal problem because s/he doesn't understand
The distribution is quasi-normal to encourage subjects to consider the items systematically and because it is assumed
that fewer items are held to be of greatest or least importance. Participants place each of the statements into the grid
according to the ‘conditions of instruction’, True for me and Untrue for me. Participants place one card (once only) in

Table 1
The 44 Q Statements
1. As we've got older my mother and I have got closer
2. I can be myself when I'm with my mother
3. My mother has always given me a hug when we see each other
4. I can talk about almost anything with my mother
5. I enjoy the time I spend with my mother
6. I miss my mother when we've not been in contact
7. My mother and I rarely argue
8. The things my mother and I do now together will leave me with many happy memories
9. I really respect and admire my mother as a person
10. The relationship I have with my mother is a very equal one
11. There seem to be quite a lot of things that are left unresolved between my mother and myself
12. My mother really irritates me most of the time
13. My mother rarely shows any physical affection towards me
14. I feel guilty about my feelings towards my mother
15. I really dislike being with my mother for any length of time
16. I can't have a conversation with my mother because she never really listens to what I say
17. I'd never try and talk to my mother about a personal problem because she doesn't understand
18. My relationship with my mother seems quite fragile and I get quite worried about it
19. My relationship with my mother has deteriorated over the years
20. Whenever I contact my mother she usually manages to upset me in some way
21. Both my mother and I are comfortable with the amount of contact we have with each other
22. If my mother became dependent I'd be happy to have her live with me
23. If my mother became dependent I'd be happy to do whatever I could to help pay for her care
24. I don't contact my mother as much as I'd like but that's only because there's always other things which compete for my time
25. My mother and I rarely bicker about who contacts who first
26. I stay in contact with my mother more out of duty and a feeling of obligation than anything else
27. I have very little contact now with my mother and that's the way I prefer it
28. I'm OK about the amount of time I see my mother … but she's not
29. I often help my mother with household chores etc. but I really resent doing it
30. I find it a burden to keep in contact with my mother
31. I never really worry about my relationship with my mother in the future
32. The thought of being without my mother makes me sad
33. I would say there's no ‘unfinished business’ between my mother and me
34. There's still a lot more my mother and I have to give to each other and I'm sure we will
35. By and large there's nothing I'd like to change about my relationship with my mother
36. I have so many happy memories as a child with my mother
37. As I child I always felt loved by my mother
38. I'd love my mother to get closer to me but I don't think that will ever happen and it makes me feel sad
39. I feel angry with my mother for not being the parent she should have been when I was a child
40. The thought of being without my mother is unbearable and I think about it a lot
41. I'd love one day to be able to get on better with my mother but I don't think that will ever happen
42. I don't think I would miss my mother if I never saw her again
43. When I was a child I never knew whether or not my mother loved me
44. I look back at my childhood as a time of unhappiness and I blame my mother quite a lot for this
184 D. Shemmings / Journal of Aging Studies 20 (2006) 177–191

each of the spaces available until each of the 44 cards have been placed. Thus, participants in effect assign scores
according to their ‘internal frame of reference’ (McKeown & Thomas, 1988, p.22), a concept not dissimilar to the
‘internal working model’.
Q-methodology (QM) was used to factor analyse1 the data. The underlying principles of QM are radically different
from those underpinning conventional correlational matrix analysis as it is the 24 individuals who are factor analysed,
not the 44 statements. When participants sort the 44 statements there are potentially over 1054 different ways of
combining them. Consequently, due to the size of the probability that a pattern will not occur, if one does it indicates
that participants' are responding to the sorting task in organised ways. In Q-methodology, a sample size of 10 or over is
considered large enough to reveal underlying factors (Brown, 1980; McKeown & Thomas, 1988). The completed
Q-sorts were entered into QMethod, a software package designed specifically for analysing Q-sort data in line with
the principles of Q-methodology. After factors are identified, each individual's degree of concordance with each
factor (called factor loading) is established. The higher an individual's factor loading, the more s/he is representative of
that factor. To more precisely describe the dimensions of each factors, QMethod reproduces each of the 44 statements as
z-scores2 for each factor. The software programme also provides an analysis of the ‘distinguishing’ and ‘consensus’
statements for each factor. In this way comparisons and contrasts between the positioning of items can be examined,
which assist factor interpretation.
Secondly, participants took part in an extended version of the Adult Attachment Interview (AAI; George, Kaplan, &
Main, 1985). The AAI employs a narrative approach to transcript analysis and is routinely used in attachment research
to explore internalised working models of close relationships. Specifically,
‘the 60- to 90-minute interview asks interviewees to choose five adjectives to describe their relationship with their
mother/father, to supply anecdotes illustrating why these adjectives are appropriate, to speculate about why their
parents behaved as they did, and to describe changes over time in the quality of their relationships with their
parents …’. (Shaver & Mikulincer, 2002, p.136).
Because participants are reflecting upon early memories, defences are exposed when they encounter unwanted
thoughts. In most relational situations, individuals can ignore, repress, or deny feelings when their attachment system is
activated, but during the AAI participants are asked to think about loss, separation, being comforted and other early
memories. Priming the attachment system in this way is referred to as ‘surprising the unconscious’ (George et al., 1985)
and a considerable amount of attachment-related information is revealed. It is how participants perform and narrate
AAI tasks which forms the focus of subsequent analysis: ‘the AAI indicates defensive strategies and more emphasis is
placed on discourse properties … than on the propositional content of what is said’ (Shaver & Mikulincer, 2002, p.136).

4. Results

QMethod generated six factors (see Table 2). Asterisked coefficients indicate which participants significantly loaded
onto each factor (p b .05). Nineteen of the 24 participants (83%) loaded significantly onto one of the six QMethod
factors. (The p values of the five remaining participants were 0.058, 0.068, 0.069, 0.075 and 0.087, indicating the degree
to which they each approached significance). The higher an individual's loading on each factor, the more they represent
and typify that factor. Between them, the six factors accounted for 71% of the overall variation. The contribution of
individual factors to the overall variance is also given in Table 2. Because no individual loaded significantly on more
than one factor, evidence exists of factor integrity and independence. This is also reinforced by the low level of inter-
correlation, apparent from Table 3, and indicates that the matrix structure comprises non-overlapping factors.
The configuration of the Q-sorts for each factor suggested the following descriptions (which were obtained after
examining in depth the printouts of the Q-sort process):

Factor 1: Confident Resolution

Factor 2: Dutiful Loyalty
Factor 3: Distant Irritation

1
Factor analysis ‘identifies and highlights any underlying structure and patterns embedded within a set of variables’ (Howe, Shemmings, & Feast,
2001, p.348).
2
z-scores are re-calibrations of scores in terms of standard deviations from the mean and, hence, denote score dispersions.
 D. Shemmings / Journal of Aging Studies 20 (2006) 177–191 185

Table 2
Q-sort factor loadings (asterisks indicate statistical significance at p b .05)
Factor 1 Factor 2 Factor 3 Factor 4 Factor 5 Factor 6
1 − 0.0503 0.0283 −0.1077 0.7686⁎ − 0.1816 0.1573
2 − 0.0565 0.4185 0.0434 0.1608 0.5433⁎ 0.2611
3 0.6368⁎ 0.2140 −0.3413 0.2496 0.2750 − 0.2764
4 − 0.0408 0.3237 −0.0283 0.5358⁎ 0.2352 − 0.2151
5 − 0.3298 0.4539 0.3574 0.1959 − 0.0701 0.3372
6 0.0433 0.5762⁎ −0.2911 − 0.0902 − 0.2437 0.2897
7 0.4813 0.2212 0.0071 − 0.0214 0.6134⁎ 0.1679
8 − 0.0160 0.0694 0.0258 − 0.6124 − 0.6348 0.0791
9 0.7538⁎ 0.1455 −0.0897 − 0.1238 0.1519 − 0.1687
10 0.8313⁎ − 0.1642 −0.0455 0.1432 − 0.0714 0.0729
11 − 0.5024 0.3810 0.0192 0.0509 0.0349 0.3630
12 − 0.3517 − 0.0397 0.2308 − 0.0111 0.0124 0.7434⁎
13 0.3860 0.3258 −0.3469 0.3575 0.1731 0.0926
14 0.3566 0.1808 0.1179 0.7124⁎ 0.1931 0.2880
15 0.7207⁎ 0.0211 −0.1305 0.3603 0.0366 − 0.2844
16 0.4328 0.1523 −0.3049 − 0.0978 0.6677⁎ 0.0272
17 0.7418⁎ 0.0159 −0.3214 − 0.1372 0.2338 − 0.1891
18 0.2262 0.2202 −0.5029 0.2802 0.3058 − 0.0396
19 − 0.0993 0.2601 −0.0354 0.1541 0.1761 0.7181⁎
20 0.1160 0.9054⁎ 0.1196 0.1171 0.1460 0.0321
21 − 0.1410 0.0744 0.9311⁎ − 0.0056 − 0.0111 0.0829
22 − 0.1643 0.7425⁎ −0.1921 0.1294 0.3989 0.0184
23 − 0.1412 0.0743 0.9375⁎ − 0.0178 − 0.0061 0.0618
24 0.1323 0.9051⁎ 0.1216 0.1158 0.1461 0.0320
% expl variance 17% 15% 12% 10% 9% 8%

Factor 4: Unresolved Yearning

Factor 5: Resolved Yearning
Factor 6: Entangled Resentment.

4.1. Relationship between each factor and participants' attachment organisation

Participants associated with a specific factor were expected to share a similar attachment organisation. Comparisons
were therefore made between the 19 participants loading significantly onto one of the six factors, and their scores on the
Attachment Style Questionnaire (ASQ). Table 4 shows that there they mapped exactly onto the original ‘Secure’,
‘Avoidant’ and ‘Ambivalent’ attachment categories.

4.2. An illustration of three factors and their corresponding internal representations and attachment styles

To illustrate how the internal representations typifying the three attachment styles are connected, Factors 3
(Avoidant—Distant Irritation) and 4 (Ambivalent—Unresolved Yearning) are compared to Factor 5 (Secure—
Resolved Yearning) by drawing on the Q-sort and the AAI analyses.

Table 3
Correlation matrix of scores between factors
Factor
1 2 3 4 5 6
1 1.00
2 0.10 1.00
3 − 0.36 0.08 1.00
4 0.18 0.29 − 0.02 1.00
5 0.42 0.39 − 0.14 0.34 1.00
6 − 0.37 0.20 0.21 0.17 0.04 1.00
186 D. Shemmings / Journal of Aging Studies 20 (2006) 177–191

Table 4
Participants' factor loadings and the ASQ solutions
Participant's loading on factor Participant's attachment organisation (ASQ)
Confident Resolution
3 Secure
9 Secure
10 Secure
15 Secure
17 Secure

Dutiful Loyalty
6 Avoidant
20 Avoidant
22 Avoidant
24 Avoidant

Distant Irritation
21 Avoidant
23 Avoidant

Unresolved Yearning
1 Ambivalent
4 Ambivalent
14 Ambivalent

Resolved Yearning
2 Secure
7 Secure
16 Secure

Entangled Resentment
12 Ambivalent
19 Ambivalent

4.2.1. Avoidant filial attachment

It will be remembered that the avoidant attachment style is characterised by defensive exclusion, through the denial
and repression of distressing memories and emotional experience. Hence, avoidantly attached adult children's internal
representations of their relationship to their parent indicated restricted emotional availability, as the following Q-sort
summary analysis and AAI excerpts indicate.

4.2.1.1. Factor 3 Q-sort: distant irritation. This variation of the avoidant attachment organisation indicated
significant levels of annoyance with the parent. The state of the relationship seemed, superficially, to have been
accepted by the adult child, who appeared resigned to there being little or no likelihood of change. There were also
signs, on the surface, that the participant did not want any change to occur: there is ‘very little contact … and that's the
way I prefer it’. Moreover, they indicated that they disliked being in their parent's company because the statement ‘The
thought of being without my parent does not make me sad’ was often placed towards the extreme right (i.e. ‘True for
Me’) in the grid. Nevertheless, there were also indications in the configuration of the statements that powerful feelings
were being suppressed (and this was strongly confirmed in the interviews). On the surface, the relationship was
mentally represented in a rather cold, nondeclarative manner but, at a deeper level, there were clear impressions of
uncomfortable feelings being repressed or disguised.

4.2.1.2. Interview exemplar: distant irritation

‘For me there's too much water gone under the bridge and I can't feel … I know I love her because she's ‘my
mother’ … but it's only that sort of love. I would no longer drive hundreds of miles, even if she was on her
deathbed’.
 D. Shemmings / Journal of Aging Studies 20 (2006) 177–191 187

What if she said she wanted to see you?

‘I wouldn't do it; no, I wouldn't do it’.
Can you say why you wouldn't see her?
‘Cos I'm not gonna ease her conscience, she's gonna have take that with her in the end. She's got to sort that out’.

4.2.2. Ambivalent filial attachment

In contrast, ambivalently attached participants tended to be overwhelmed by intrusive affect and painful memories,
and this emerged strongly in their Q-sorts and AAI transcripts.

4.2.2.1. Factor 4 Q-sort: unresolved yearning. The overwhelming impression was of significantly unresolved
feelings (unlike the more resolved yearning emerging in Secure Factor 5) . The highest positive score was over two
standard deviations away from the mean and referred directly to unresolved feelings about the past: ‘When I was a child
I never knew whether or not my mother loved me’ (z = 2.236). Feelings of being unloved as a child, however, seemed
not to produce an openly hostile response; on the contrary, unlike with Factor 3 (Distant Irritation), participants here
indicated that the thought of being without their parent ‘makes them sad’. A desire for things to be different was also
consistently expressed. Coupled with the experience of anticipatory loss – ‘The thought of being without my mother
makes me sad’ – a sense of the adult child desperately longing for the relationship to improve, but with little realistic
hope that it would, was strongly represented.

4.2.2.2. Interview exemplar: unresolved yearning

‘My mother never played with us as children. She always, you know, ‘mingled in’, but she didn't do the actual
‘playing thing’… What I would like to happen now is my mother to be happy with me, her to be happy with my
partner, and I want a … just a normal bonding, I just don't want any more insecure feelings about each other … I
want to phone her and tell her, ‘Mum, this is what I feel and you may or may not be able to help me maybe with it
but this is what I feel’ and I want her to listen to what I am saying and I want her to understand what I am feeling …
I would love her to say, ‘I am so proud of you and I want you to go forward in love – and you have done it so far –
but just be careful; don't do something silly’ or things like that. I suppose mothers do that, but …. she never has
with me. Oh, I just wish it could be a normal mother–daughter bond … it never will be and it hurts so much at
times’ (becomes tearful).

4.2.3. Secure filial attachment

Finally, the Q-sorts and AAI tapes of securely attached adult children indicated far more potential for emotional
availability than their insecurely attached counterparts. This was not always the result of necessarily having had
problem-free childhood experiences and memories. Securely attached adults reflect upon the questions in the AAI
using ‘an acceptably coherent and collaborative narrative, whether experiences are reported as having been favourable
or unfavourable’ (Hesse, 1999, p.397). Factor 5 (Resolved Yearning) is chosen to illustrate this facility to integrate a
less-than-perfect past into a more coherent model of the present and future with their parent.

4.2.3.1. Factor 5 Q-sort: resolved yearning. The Q-sorts for Factor 5 indicated that participants wanted to get closer
to their parent but they seemed realistically adjusted to their assessment that this would not happen. A similar yearning
to that evident in Factor 4 (Unresolved Yearning) was observed, but here it was considerably less ‘unresolved’. The
highest positive statement was ‘I'd love my mother to get closer to me, but I don't think that will ever happen and that
makes me feel sad’ (z = 2.313). Care needs to be taken, however, not to interpret this statement in isolation, as this
would lead to the unwarranted conclusion that the participant was preoccupied and enmeshed in their relationship with
their parent. A careful reading of proximally placed statements leads to a different interpretation. For example,
participants loading on Factor 5 enjoyed being with their parent; they indicated that they always felt loved; and they felt
that the relationship with their parent was not characterised with arguments, nor maintained merely out of a sense of
duty. Read in the context of other Q-statements – in particular, the very low positioning of ‘There seem to be quite a lot
of things that are left unresolved between my mother and myself’ – the statement, ‘I'd love my mother to get closer to
188 D. Shemmings / Journal of Aging Studies 20 (2006) 177–191

me but I don't think that will ever happen and that makes me feel sad’ is capable of being interpreted as a form of
yearning for a more rewarding relationship, which nevertheless the adult child suspects will not materialise (and which
they seem not to be unduly worried about).
They enjoyed the time spent with their parent and knew that s/he would be missed. For them, the relationship had
not deteriorated over the years and they did not remain in contact merely out of duty.

4.2.3.2. Interview exemplar: resolved yearning

‘I'll try and give you an idea how odd he is. I didn't know he had two sisters until just before my mum died and
then he was furious that I knew. I have been married for fourteen years and he has crossed the threshold of my
house only twice … and one of those times was when mum died. And, after she died, although he always used to
do the cooking when she was alive, I used to take him a roast dinner round on a Sunday. Now, for the first three
months, I was allowed into his house and I would give him his dinner and chat to him but then I observed that he
wouldn't eat it if I was there, so I thought, you know, perhaps he doesn't want me to stay. So for the next couple of
weeks I would just give it to him, briefly chat to him while I was getting his knife and fork out and then disappear’.

How do you see your relationship with dad in the future?

It will probably get easier because, if he becomes more dependant, it won't be such a problem getting the contact
and getting in, because, for example, when he was poorly, we all had a key to the door … Now, he still likes the
keys back. I only really hold out a little hope, though, that he might feel more comfortable about showing his needs
and feelings and that would be nice for us both; it would be nice but I don't hold out much hope. Still, I'll always
do what I can for him and I'm not unduly bothered by his apparent lack of appreciation’.

4.3. Some implications for adult children's emotional availability

Attachment-based research consistently indicates that individuals relate very differently when their attachment
systems are activated and this, albeit small-scale, study confirms that trend. The research also suggests strongly that the
degree of emotional availability shown by adult children towards their parents in later life may be affected by their
attachment organisation. Although the parent's attachment system was not considered in this study, its inclusion is
likely to add complexity to the relationship.
Some additional implications of differential emotional availability are now identified. The examples illustrate
how internal working models mediate the relationship between attachment security and emotional availability. The
last example above suggests how, even though the participant has a strained relationship with her father, and that she
would like to get closer to him, she is nevertheless resigned positively to the unlikelihood of such a resolution. From
the point of view of any future caregiving, however, she remains emotionally available to her father. On the other
hand, as the earlier extracts indicate, avoidantly attached adult children tend to defensively exclude feelings from
their relationship to their parent and this leaves them emotionally brittle and unavailable. Ambivalently attached
individuals are also likely to be emotionally unavailable, but for precisely the opposite reason: they quickly become
flooded by emotion and then end up preoccupied by it, which makes it difficult for them to focus on their parent's
needs.
Firstly, the strength and intensity of emotions expressed by all participants during the interviews was striking. The
relationship was usually seen by securely attached adult children as mutually supportive, whether or not the participant
loaded on ’secure’ Factor 1 (Confident Resolution) or ‘secure’ Factor 5 (Resolved Yearning). Their emotions were
generally positive, and as a result, the current relationship with their parent gave them considerable pleasure and
comfort, even if their memories of their own childhood were less optimal. This reinforces one of the basic tenets of
attachment theory: childhood events are not necessarily distressing in and of themselves; more important is the
individual's ability to make sense of them later, by understanding parental motivations and limitations in order
eventually to ‘re-story’ early experience. It is this facility that marks the difference between securely and insecurely
attached adults (Crittenden, 1996; Fonagy, 1995). Conversely, it was the intensity of the uncomfortable or distressing
feelings experienced during the interviews by insecurely attached participants, who were usually surprised by the force
and speed with which these emotions appeared, that formed a lasting impression.
 D. Shemmings / Journal of Aging Studies 20 (2006) 177–191 189

Secondly, although participants loading onto ‘secure’ Factor 5 (Resolved Yearning) seemed to accept having a less
than ideal relationship with their parent (without regret or remorse), this did not diminish their feelings of empathy and
sensitivity towards them. They regularly displayed ‘flexible emotional functioning’ (Davenhill, Balfour, Rustin,
Blanchard & Tress, 2003, p.261). Insecurely attached participants, on the other hand, either oscillated during the
interview between over-involvement followed by temporary rejection (‘ambivalent’ Factors 4 and 6), or by an
insistence that their parent was, or should become, overly self-reliant (‘avoidant’ Factors 2 and 3).
Thirdly, it was clear from both the Q-sorts and the AAI transcripts that a considerable amount of unfinished business
and unresolved dilemmas dominated the accounts of participants with an insecure attachment organisation, and this is
also likely to restrict their emotional availability towards their parents. During the interviews they would tend to deny
powerful feelings, but their suppression was noticeably contra-indicated by observable aspects of their behaviour,
usually by the introduction of contradictory statements, often expressed within the same sentence but without recognition
and reflection. Such suppression is known to typify the affect regulation strategies of avoidant individuals (Fraley,
Davies, & Shaver, 1998). Ambivalent individuals, on the other hand, were often flooded by a cascade of emotional
intrusions, which they were relatively powerless to process effectively and this also was observable during the interviews.

5. Discussion

Although the present study addresses only one side of the relational dyad, stressful conditions within later life
relationships, such as conflict, actual or felt pressure to provide caregiving, and role reversal, are likely to trigger
attachment behaviour between both parties simultaneously. Consequently, as with most adult close relationships, filial
attachments are bi-directional, in the sense that either party gives and receives care and protection. Because the parent
brings his or her attachment style to the relationship, sometimes in the form of co-dependency, counter-dependency or
compulsive self-reliance, this relational dynamic raises some interesting questions from the perspective of attachment
theory, especially in terms of reciprocity. For example, how are attachment styles distributed between generational
partners: do adult children exhibit attachment behaviour reminiscent of their own childhood when in the company of
their parent and, in tandem, does the parent adopt attachment patterns redolent of their childhood? As a consequence, it
is interesting to speculate what might happen if either or both parties re-enacted early childhood experiences of
caregiving within the present relationship. And, because the developmental trajectory of the parent is at a different stage
from their adult child, how might they seek to resolve and reconcile the relationship? Ultimately, would they seek to re-
parent their child, and what might be the consequences of sons and daughters not reciprocating?
Because some adult children experience a parent's aging or diminishing independence as a form of abandonment,
secure base stability is likely to be threatened. Relational pressure could also increase if one person becomes
uncomfortable with closeness and then failed to respond to, or rejected, the other person's distress. This might also
occur if one partner seeks excessive emotional intimacy or reassurance from the other, especially if s/he then became
emotionally withdrawn or demanding were it to be refused. This situation may not be new to the relationship but it
could become magnified if experienced alongside other stressors such as illness, the need for residential care, or the
death of loved ones, close friends or other relatives.
Preserving attachments to members of the primary family group arguably increases in importance as adults grow
older (Bowlby, 1979, 1980) because one of the primary reasons for the maintenance of the bond in later life involves
the protection of the parent, whom the adult child increasingly fears, consciously or unconsciously, will be lost through
aging and death (Cicirelli, 1983, 1985). Such thoughts leave people feeling exposed and vulnerable, and they are likely
to place both relational partners under emotional pressure (Bowlby, 1969; Parkes, 1985; Parkes & Weiss, 1983). The
consequences need to be addressed in health and social care policy-making. They are also of importance to managers
and practitioners involved in elder care, where there is a growing expectation that adult children (which, in reality,
usually means daughters) will provide increased levels of emotional support and direct care, in many cases by having
their parent live with them.

References

Antonucci, T. C. (1994). Attachment in adulthood and aging. In M. B. Sperling, & W. H. Berman (Eds.), Attachment in adults: Clinical and
developmental perspectives. New York: Guilford Press.
190 D. Shemmings / Journal of Aging Studies 20 (2006) 177–191

Belsky, J. (2002). Developmental origins of attachement styles. Attachment and Human Development, 4, 166−170.
Berlin, L. J., & Cassidy, J. (1999). Relations among relationships: Contributions from attachment theory and research. In J. Cassidy, & P. R. Shaver
(Eds.), Handbook of attachment: Theory, research and clinical applications (pp. 688−712). NY: Guilford Press.
Bowlby, J. (1969). Attachment and loss. Attachment, Vol. 1. NY: Basic Books.
Bowlby, J. (1973). Attachment and loss. Separation, Vol. 2. NY: Basic Books.
Bowlby, J. (1979). The making and breaking of affectional bonds. London: Tavistock.
Bowlby, J. (1980). Attachment and loss. Loss, Vol. 1. NY: Basic Books.
Bradley, M., & Cafferty, T. P. (2001). Attachment among older adults: Current issues and directions for future research. Attachment and Human
Development, 3, 200−221.
Bretherton, I., & Mulholland, K. A. (1999). Internal working models in attachment relationships: A construct revisited. In J. Cassidy, & P. R. Shaver
(Eds.), Handbook of attachment: Theory, research and clinical applications (pp. 89−114). NY: Guilford Press.
Brown, S. R. (1980). Political subjectivity: Applications of Q methodology in political science. New Haven, CT: Yale University Press.
Cassidy, J., & Shaver, P. R. (1999). Handbook of attachment: Theory, research and clinical applications. NY: Guilford Press.
Cicirelli, V. G. (1981). Helping elderly parents: Role of adult children. Boston: Auburn House.
Cicirelli, V. G. (1983). Adult children's attachment and helping behaviours to elderly parents: A path model. Journal of Marriage and the Family, 45,
825–822.
Cicirelli, V. G. (1985). Divorce as a factor influencing adult children's helping behaviour to elderly parents: A path model. Journal of Marriage and
the Family, 47, 815−825.
Cicirelli, V. G. (1993). Attachment and obligation as daughters' motives for caregiving behaviour and subsequent effect on subjective burden.
Psychology and Aging, 8, 144−155.
Connidis, I. A., & McMullin, J. A. (2002). Sociological ambivalence and family ties: A critical perspective. Journal of Marriage and the Family, 64,
558−567.
Consedine, N. S., & Magai, C. (2003). Attachment and emotion experience in later life: The view from emotions theory. Attachment and Human
Development, 5, 165−187.
Craik, K. (1943). The nature of explanation. Cambridge: Cambridge University Press.
Crawford, L. M., Bond, J. R., & Balshaw, R. F. (1994). Factors affecting sons' and daughters' caregiving to older parents. Canadian Journal on
Aging, 13, 454−469.
Crittenden, P.M. (1996) The Modified Adult Attachment Interview, unpublished manuscript.
Davenhill, R., Balfour, A., Rustin, M., Blanchard, M., & Tress, K. (2003). Looking into later life: Psychodynamic observation and old age.
Psychoanalytic Psychotherapy, 17, 253−266.
Diamond, D., & Blatt, S. (1994). Internal working models and the representational world in attachment and psychoanalytic theories. In M. B.
Sperling, & W. H. Berman (Eds.), Attachment in adults: Clinical and developmental perspectives (pp. 72−98). NY: Guilford Press.
Feeney, J. A. (1995). Adult attachment and emotional control. Personal Relationships, 2, 143−159.
Feeney, J. A. (1998). Adult attachment and relationship-centered anxiety: Responses to physical and emotional distancing. In J. A. Simpson, & W. S.
Rholes (Eds.), Attachment theory and close relationships (pp. 189−220). NY: Guilford Press.
Feeney, J. A., & Noller, P. (1996). Adult attachment. London: Sage.
Feeney, J. A., Noller, P., & Hanrahan, M. (1994). Assessing adult attachment: Developments in the conceptualisation of security and insecurity.
In M. B. Sperling, & W. H. Berman (Eds.), Attachment in adults: Clinical and developmental perspectives (pp. 128−152). New York:
Guilford Press.
Finley, N. J., Roberts, M. D., & Banahan, B. F. (1988). Motivators and inhibitors of attitudes of filial obligation towards aging parents. Gerontologist,
28, 73−78.
Fonagy, P. (1995). Mental representations from an intergenerational cognitive science perspective. Infant Mental Health Journal, 15, 57−68.
Fraley, R. C., Davis, K. E., & Shaver, P. R. (1998). Dismissing-avoidance and the defensive organisation of emotion, cognition and behaviour. In J. A.
Simpson, & W. S. Rholes (Eds.), Attachment theory and close relationships (pp. 249−279). New York: Guilford Press.
Fraley, R. C., & Shaver, P. R. (1997). Adult Attachment and the suppression of unwanted thoughts. Journal of Personality and Social Psychology, 79,
816−826.
Fulton, G., Madden, C., & Miniciello, V. (1996). The social construction of anticipatory grief. Journal of Social Science and Medicine, 43,
1349−1358.
George, C., Kaplan. N. & Main, M. (1985) The Adult Attachment Interview, unpublished manuscript, Department of Psychology, University of
California at Berkeley.
Hamon, R. R., & Blieszner, R. (1990). Filial responsibility expectations among adult child–older parent pairs. Journal of Gerontology. Psychological
Sciences, 45, 110−112.
Hayes, N. (2000). Doing psychological research. Buckingham: Open University Press.
Hazan, C., & Shaver, P. R. (1987). Romantic love conceptualized as an attachment process. Journal of Personality and Social Psychology, 52,
511−524.
Hesse, E. (1999). The adult attachment interview: Historical and current perspectives. In J. Cassidy, & P. R. Shaver (Eds.), Handbook of attachment:
Theory, research and clinical applications (pp. 395−433). NY: Guilford Press.
Howe, D., Shemmings, D., & Feast, J. (2001). Age at placement and adopted people's experience of being adopted. Child and Family Social Work, 6,
337−350.
Krause, N. (2004). Lifetime trauma, emotional support, and life satisfaction among older adults. Gerontologist, 44, 615−623.
Krause, A. M., & Haverkamp, B. E. (1996). Attachment in adult child–older parent relationships: Research, theory and practice. Journal of
Counseling and Development, 75, 83−91.
 D. Shemmings / Journal of Aging Studies 20 (2006) 177–191 191

Luescher, K., & Pillemer, K. (1998). Intergenerational ambivalence: A new approach to the study of parent–child relations in later life. Journal of
Marriage and the Family, 60, 413−425.
Magai, C. (1999). Affect, imagery and attachment: Working models of interpersonal affect and the socialisation of emotion. In J. Cassidy, & P. R.
Shaver (Eds.), Handbook of attachment: Theory, research and clinical applications (pp. 787−802). New York: Guilford Press.
Magai, C., Consedine, N. S., Gillespie, M., O'Neal, C., & Vilker, R. (2004). The differential roles of early emotion socialization and adult attachment
in adult emotional experience: Testing a mediator hypothesis. Attachment and Human Development, 6, 389−418.
Magai, C., Hunziker, J., Mesias, W., & Culver, L. C. (2000). Adult attachment styles and emotional biases. International Journal of Behavioural
Development, 24, 301−309.
Main, M., Kaplan, N., & Cassidy, J. (1985). Security in infancy, childhood and adulthood: A move to the level of representation. Monographs of the
Society for Research in Child Development, 50, 66−104.
McKeown, B. F., & Thomas, D. B. (1988). Q-methodology. London: Sage.
Mickelson, K. D., Kessler, K., & Shaver, P. R. (1997). Adult attachment in a nationally representative sample. Journal of Personality and Social
Psychology, 73, 1092−1106.
Miessen, B. M. L. (1993). Alzheimer's disease, the phenomenon of parent fixation and Bowlby's attachment theory. International Journal of
Geriatric Psychiatry, 8, 147−153.
Mikulincer, M. (1998). Adult attachment style and individual differences in functional versus dysfunctional experiences of anger. Journal of
Personality and Social Psychology, 74, 513−524.
Mikulincer, M., & Florian, V. (1998). The relationship between adult attachment styles and emotional and cognitive reactions to threat. In J. A.
Simpson, & W. A. Rholes (Eds.), Attachment theory and close relationships (pp. 143−165). NY: Guilford.
Mikulincer, M., Florian, V., & Weller, A. (1993). Attachment styles and fear of personal death. Journal of Personality and Social Psychology, 58,
273−280.
Myers, J. E. (1988). The mid/late life generation gap: Adult children with aging parents. Journal of Counselling and Development, Vol. 66.
(pp. 331−335). New Haven, CT: Yale University Press.
Parkes, C. M. (1985). Bereavement. British Journal of Psychiatry, 146, 11−17.
Parkes, C. M., & Weiss, R. S. (1983). Recovery from bereavement. New York: Basic Books.
Pillemer, K., & Suitor, J. J. (2002). Explaining mothers' ambivalence toward their children. Journal of Marriage and Family, 64, 602−613.
Roberts, J. E., Gotlib, I. H., & Kassel, J. D. (1996). Adult attachment security and symptoms of depression: The mediating roles of dysfunctional
attitudes and low self-esteem. Journal of Personality and Social Psychology, 70, 310−320.
Rothbard, J. C., & Shaver, P. R. (1994). Continuity of attachment across the life span. In M. B. Sperling, & W. H. Berman (Eds.), Attachment in
adults: Clinical and developmental perspectives (pp. 31−71). NY: Guilford Press.
Seale, C. (1998). Constructing death: The sociology of dying and bereavement. Cambridge: Cambridge University Press.
Seeman, T. E., & Crimmins, E. (2001). Social environment effects on health and aging: Integrating epidemiologic and demographic approaches and
perspectives. Annals of the New York Academy of Sciences, 954, 88−117.
Shaver, P. R., & Mikulincer, M. (2002). Attachment-related psychodynamics. Attachment and Human Development, 4, 133−161.
Shaver, P. R., & Mikulincer, M. (2004). Attachment in the later years. Attachment and Human Development, 6, 451−464.
Sorensen, S., Webster, J. D., & Roggman, L. O. (2002). Adult attachment and preparing to provide care for older relatives. Attachment and Human
Development, 4, 84−106.
Sperling, M. B., & Berman, W. H. (Eds.). (1994). Attachment in adults: Clinical and developmental perspectives NY: Guilford Press.
Steele, H. (2000) Seminar paper presented at a Conference on Attachment and Older People held at the University College, London on Friday 10th
March, 2000.
Stein, C. H., Wemmerus, A., Ward, M., Gaines, M., Freeberg, A. L., & Jewell, T. C. (1998). “Because they're my parents”: An intergenerational study
of felt obligation and parental caregiving. Journal of Marriage and the Family, 60, 611−622.
Thompson, R. A. (1994). Emotion regulation: A theme in search of definition. In N. A. Fox (Ed.), The development of emotion regulation: Biological
and behavioural considerationsMonographs of the Society for Research in Child Development, Vol. 59. (pp. 25−52). Serial No. 240.
Tompkins, S. S. (1993). Affect, imagery, consciousness, Vol.4. Cognition, duplication and transformation of information. New York: Springer.
Van IJzendoorn, M. H., & Bakermans-Kranenburg, M. J. (1997). Intergenerational transmission of attachment: A move to the contextual level. In L.
Atkinson, & K. J. Zucker (Eds.), Attachment and psychopathology (pp. 135−170). New York: Guilford Press.
Waters, E. (1995). The attachment Q-set. In E. Waters, B. E. Vaughan, G. Posada, & K. Kondo-Ikemura (Eds.), Caregiving, cultural and
cognitive perspectives on secure-base behaviour and working modelsMonographs of the Society for Research in Child Development, Vol. 60.
(pp. 247−254).
Zhang, F., & Labouvie-Vief, G. (2004). Stability and fluctuation in adult attachment style over a 6-year period. Attachment and Human Development,
6, 419−438.