How an Early Caregiving Style Affects Adult Romantic Love

Lisa Dale ^[1], Department of Psychology, Staffordshire University

Abstract

This study investigates the theory that adult romantic attachment is reflective of a person's perception of an early caregiving style. Hazan and Shaver (1987) were the first researchers to investigate if adult romantic love could be conceptualised as an attachment process. Their work was heavily criticised by subsequent researchers, due to its simplicity and categorisation (Collins and Read 1990, Simpson, 1990). Griffin and Bartholomew addressed this in 1994, using a four-category questionnaire to measure adult attachment. This paper incorporates Griffin and Bartholomew's (1994) scale. Furthermore, to measure perceptions of an early caregiving style, Hazan and Shaver's (1986) parental/caregiving questionnaire was used, cited in Collins and Read (1990). To the author's knowledge, these two measures have not been used together before. 68 male and female participants from diverse socioeconomic backgrounds participated in the study, with a mean age of 34 years for women and 37 years for men.

The hypothesis was accepted: The warmth of the caregiver was positively correlated to a high secure adult attachment score. A high insecure caregiver score was positively correlated with a high insecure adult attachment score. Implications for theories of adult attachment are discussed, as are measurement limitations and other issues related to future research.

Keywords: Attachment Theory, Romantic Love, Caregiver, Secure-Attachment, Insecure-Attachment

Introduction

Bowlby's (1969) work on attachment has progressed along two fairly independent trajectories. One line of research has focused on the attachment relationship between child and parent (Bowlby 1969, 1973, 1987; Ainsworth et al., 1978). The other line of research has focused on the attachment dynamics of adult romantic relationships (Hazan and Shaver, 1987). The present study expands the work of Hazan and Shaver (1987) by using Griffin and Bartholomew's (1984) relationship framework to replace Hazan and Shaver's discrete categorical measure. Hazan and Shaver's (1986) parental caregiving questionnaire, cited in Collins and Read (1990), was used to measure perceptions of parental caregiving. A brief overview of attachment and adult romantic attachment provides a framework for the current study.

The term 'attachment' is the strong emotional bond formed between the infant and the primary caregiver (Bowlby, 1973).It involves the warm feelings that a parent and child have for each other and the comfort and support they provide for each other, which is particularly important in times of stress. Bowlby believed that attachment behaviour characterises human beings 'from the cradle to the grave' (1979: 129). It was not until the late 1980s that researchers began to integrate work on adult love relationships with early attachment theory (Hazan and Shaver, 1987). Of particular interest was how a child's early attachment with a caregiver and perceptions of early attachment can shape important beliefs about the self, and how this guides relationships in adulthood.

According to Bowlby, the child's attachment with his or her primary caregiver leads to the development of an internal working model. This internal working model is a cognitive framework comprising mental representations for understanding the world, self and others. A person's interaction with others is guided by memories and expectations from their internal model, which influence and help to evaluate their contact with others. According to Bowlby (1969) the primary caregiver acts as a prototype for future relationships via the internal working model. So, on repeated interactions with the caregiver the infant learns what to expect and behaves according to this (Hazan and Shaver, 1994). Bowlby (1969) hypothesised that sensitive and responsive caregiving leads to a secure relationship and to internal working models of the caregiver as trustworthy and helpful and of the self as deserving of the caregiver's treatment. Conversely, if the caregiver is insensitive, this can lead to an insecure attachment, and to working models of the caregiver as untrustworthy and to the self as being unworthy of care. This can lead to insecure or anxious models of attachment where the child demands more attention or withdraws from others in order to become more self-sufficient (Main, 1990). Bowlby believed that these early internal working models (IWM) guide perceptions and behaviour in later relationships (Cassidy and Shaver, 1999).

Ainsworth et al.(1978) tested some of Bowlby's ideas using a laboratory test called the 'strange situation'. Infants and their caregivers were separated and then reunited after a short period of time, over a series of episodes. From this experiment, Ainsworth and her colleagues came up with a three-category model of attachment styles, secure, anxious and avoidant. Although Ainsworth et al. (1978) found that the majority of infants had a secure attachment style, the variety of attachment styles observed show the difference in how the attachment system is organised, a central part of which is the child's perception of whether the caregiver will be available and responsive when needed (Collins and Read, 1990).

In the early 1980s several researchers began to use the ideas of Bowlby (1973) and Ainsworth et al. (1978) to understand the nature of adult loneliness and love. Some researchers found that many lonely adults report difficult childhood relationships with parents, suggesting that attachment history, or the way parental caregiving is perceived later on in life, can have an influence on adult loneliness (Rubenstein and Shaver, 1982). Additionally, some researchers had observed a high degree of changeability in the way people approach and respond to love relationships including intense anxiety and avoidance to commit (Hendrick and Hendrick, 1986; Lee, 1988). Despite this research no researcher had, at that time, proposed a theory, which explained individual differences in adult attachment, particularity romantic love (Fraley and Shaver, 2000).

In order to address this, Hazan and Shaver (1987) investigated whether romantic love could be conceptualised as an attachment process. Hazan and Shaver (1987) developed a 'forced choice' self-report measure of adult attachment, which consisted of three paragraphs written to capture the main features of Ainsworth's three attachment styles.

For example, the secure classification was

'I find it relatively easy to get close to others and am comfortable depending on them and having them depend on me. I don't worry about being abandoned or about someone getting too close to me.'

The anxious classification was

'I find that others are reluctant to get as close as I would like. I often worry that my partner doesn't really love me or won't want to stay with me. I want to merge completely with another person, and this desire sometimes scares people away.'

The avoidant classification was

'I am somewhat uncomfortable being close to others, I find it difficult to trust them completely, difficult to allow myself to depend on them. I am nervous when anyone gets too close, and often, love partners want me to be more intimate than I feel comfortable being.'

(Hazan and Shaver 1987)

Subjects were asked to choose which paragraph best described their feelings in close relationships.

Hazan and Shaver (1987) observed that romantic relationships are characterised by similar dynamics to those described by Bowlby (1973) and Ainsworth et al. (1978). For example, adults feel more secure when their partner is nearby, accessible, and responsive. The partner, in these circumstances, may be used as a 'secure base' from which to explore the environment. When an individual is feeling distressed, sick or threatened, the relationship partner is used as a source of protection for safety and comfort. Hazan and Shaver (1987) also found that anxious lovers had emotional highs and lows, were jealous and pre-occupied with a partner, and that avoidant adults tended to shy away from intimacy in fear of becoming hurt.

Hazan and Shaver (1987) also found that adult attachment was related to self-reports of perceptions of early parent-child relationships. For example, secure adults reported their parents to be respectful and more accepting, anxious adults remember their parents being intermittently available and avoidant adults remember their parents as critical and distant (Cassidy and Shaver, 1999). However, it is important to note that adult self-reports do not necessarily relate to observed parent-child relationships (Feeney and Rayne, 1994) or to attachment itself, as they are perceptions of the adult about the care given to them and may not be objective.

Although Hazan and Shaver's (1987) research was an important step in attachment theory, there have been a number of criticisms of the methodology. For example although the three-category model was a reasonable version of the infant attachment styles and many researchers continue using it due to its simplicity, there are limitations when using this to measure complex adult romantic love. For example, each description contains statements about more than one aspect of relationships (i.e. the 'secure' description includes both being comfortable with closeness and being able to depend on others). Respondents must therefore accept an entire description that may not fully reflect their feelings (Collins and Read, 1990; Simpson, 1990).

Fraley and Waller (1998) recommended that a shift was needed towards a gradual assessment of attachment style, which produces multiple scores. In the early 1990s a four-category model was proposed (Bartholomew and Griffin, 1991) and was developed as a continuous measurement of adult attachment. Bartholomew and Horowitz (1991) developed a four-category model (Secure, Preoccupied, Fearful-Avoidant and Dismissing-Avoidant) to assess adult attachment. Based on this, Griffin and Bartholomew (1994) went on to develop the Relationship Scale Questionnaire. This measure is a 30-item adult attachment questionnaire that investigates relationship quality and adult attachment and is based on a continuous measure.

The aim of the current article is to focus on how subjects remember their early caregiving experience and to identify if there is a relationship between perceptions of early caregiving and current adult romantic attachment. Recognising the criticism made by researchers regarding the simplicity of Hazan and Shaver's (1987) three-category model, Griffin and Bartholomew's (1994) 'Relationship Scale Questionnaire' will be used to measure adult attachment.

Given the link between early caregiving and adult self-reports of parental caregiving, Hazan and Shaver's (1986) Caregiving questionnaire, cited in Collins and Read (1990), was used to elicit current perceptions of parental caregiving from the participants in this study.

According to Cassidy and Shaver (1999), self reports are an appropriate instrument to use in assessing adult attachment for three reasons:

According to Bowlby (1969), attachment plays a pivotal role in a person's life, and adults are able to provide researchers with information about these experiences;
The majority of adults have had enough close relationships to be in a position to reflect how they behave in those relationships;
Conscious and unconscious processes operate in the same direction to achieve a goal.

To the author's knowledge, these two self-report measures have not been used together to assess the influence of perceptions of early caregiving on adult attachment. In addition, it is important to examine the influence of perceptions of parental care-giving on adult attachment because it is not always possible to conduct lengthy adult attachment interviews or to observe longitudinally the effects of parent-child attachment. However, for relationship therapists as well as those working with adolescent services, it might be useful to have data on both perceptions of caregiving and relationship measures. If there is a correlation between the two questionnaires it could be extremely useful in diagnosing possible sources of relationship problems based on early caregiving.

In summary, the main aim of the current study is to validate earlier findings that adult perceptions of early caregiving (using self-report measures) are related to specific types of adult attachment relationships. The hypothesis is: whether a warm early caregiving score is correlated with secure attachment, and whether an insecure early caregiving score is related to an insecure adult attachment. These broad categories were chosen as it was not possible to recruit enough participants to all four categories of RSQ (as this was a time-limited undergraduate project). Nevertheless, it is important to establish whether these measures yield similar results to those found by studies using attachment measures, as they may be much quicker to use in clinical settings.

Method

Design

A correlational design was employed. Griffin and Bartholomew's (1994) 'Relationship Scale Questionnaire' was used to measure adult attachment. In relation to perceptions of early caregiving Hazan and Shaver's (1986) 'Descriptions of Caregiving Style' was used, cited in Collins and Read (1990); see Appendix 1. The adult attachment scores had two levels: secure and insecure (adult) attachment. The early caregiving scores also had two levels, warm/secure score and insecure score. Participants' scores on the questionnaires were analysed to determine whether there was a correlation between perceptions of early caregiving and adult attachment, based on attachment theory.

Participants

A power analysis (correlational design) indicated 68 participants were required to achieve a medium effect size (Cohen, 1988, cited in Clark-Carter, 2009: 308; see Appendix 2). 68 participants were recruited and all participants took part in all aspects of the study. Opportunist sampling was employed. In order to gain a wide age range, participants were recruited from the study body at Staffordshire University (both undergraduate and postgraduate), professionals in the local area and mothers from a local parent-and-baby group. Participants were from a range of socioeconomic backgrounds and occupations. There were 49 women and 19 men in the final sample. The male age range was 19-54 years (mean age of 34 years); the female age range was 19-61 years (mean age of 37 years).

Materials

All measures were administered in the form of a single composite questionnaire (including both Hazan and Shaver's caregiving questionnaire and Griffin and Bartholomew's relationship scales questionnaire). The first section explained the background information to the study, including the aim and signed consent.

The second section asked participants to read three paragraphs, describing the caregiving characteristics of three particular styles, after Hazan and Shaver (1986). One described a warm/responsive parent; an example of this includes 'She/he was generally warm and responsive, she/he was good at knowing when to be supportive and when to let me operate on my own'. A second paragraph described a cold/rejecting parent, an example of this includes 'She/he was fairly cold and distant, or rejecting, not very responsive, I wasn't her/his highest priority'. Finally an anxious/ambivalent parent was described as 'she/he was noticeably inconsistent in her/his reactions to me, sometimes warm and sometimes not'. Respondents were then asked to rate on a 5-point scale (1= strongly disagree to 5= strongly agree) the extent to which they agreed with each statement while they were growing up.

The second questionnaire was the relationship scales questionnaire(Griffin and Bartholomew, 1994). This 30-item questionnaire consists of phrases from the paragraph descriptions of Hazan and Shaver's (1987) adult attachment questionnaire and Bartholomew and Horowitz's (1991) categorical measures, as well as three items developed by Collins and Read (1990). Respondents were asked to read the statements and rate the extent to which it described their feelings about their current romantic relationship, on a 5-point Likert scale (1=not at all like me to 5= very like me). The questions made up the following categories: secure, pre-occupied (anxious), dismissing-avoidant and fearful-avoidant. Five statements contributed to the secure and dismissing attachment patterns and four statements contributed to the fearful and preoccupied attachment patterns. Taking the mean of the four or five items representing each attachment prototype derives scores for each attachment pattern. Examples from the secure questions include: 'I find it easy to get emotionally close to others' and 'I am comfortable depending on other people'; examples of the pre-occupied questions were 'I am comfortable without close emotional relationships' and 'I worry that others don't value me as much as I value them'. Examples of the dismissing questions include: 'It is very important for me to feel independent' and 'I prefer not to have other people depend on me' and finally, examples of the fearful questions: 'I find it difficult to depend on other people' and 'I worry that I will be hurt if I allow myself to become too close to others'. At the end of the session, participants were debriefed with an explanation of why the study was conducted.

Procedure

Standard ethical procedures were adhered to. Participants were asked if they were currently in a relationship to ensure respondents were referring to the current relationship they were in. If participants were in a relationship they were asked to sign a consent form and to record their age and gender. They were notified of their right to withdraw, confidentially and anonymity. Participants were issued with a unique number. This number identified them should they wish to withdraw at a later stage. Once the consent form was signed, participants were given instructions and questionnaires. There were no time constraints but participants were told that the questionnaire would take approximately 15 minutes to complete. Participants were asked to think about one parent (their main caregiver) when completing the parental caregiver questionnaire (Hazan and Shaver, 1986, cited in Collins and Read 1990). Participants were then asked to think about their current relationship and complete the relationship scales questionnaire.

When assigning participants to the attachment styles, for the purpose of the correlations, the following method was employed. For the parental caregiving style questionnaire (Hazan and Shaver, 1986, cited in Collins and Read, 1990), participants rated the three statements as warm/secure parent/caregiver, anxious/ambivalent or cold caregiver. For each participant a score was given for the warm caregiver; an average was taken from the anxious/ambivalent or cold caregiver score and was called an 'insecure caregiver score'.

When scoring the Relationships Scale Questionnaire, items were grouped using the four attachment styles described by and Griffin and Bartholomew (1994). The mean score for each of the four categories of question (secure, anxious, dismissing-avoidant and fearful-avoidant) was used to assign participants to either the 'securely' attached category or 'insecurely' attached category. To be categorised in the 'securely attached' category, participants' secure score had to be higher than the insecure score (the insecure score was the combined score of the anxious, dismissing-avoidant or fearful-avoidant category). To be 'insecurely attached' participants' mean score had to be higher in the insecure category. These broad categories were chosen as it was not possible to recruit enough participants to all four categories of RSQ (as this was a time-limited undergraduate project).

Results

Table 1 shows the number of participants in each category for both Hazan and Shaver's (1986) caregiver questionnaire, cited in Collins and Read (1990) and Griffin and Bartholomew's (1994) relationship scales questionnaire.

Early attachment to caregiver	N	Adult Attachment Score	N
Secure (warm)	53	Secure	35
Insecure (cold / ambivalent)	15	Insecure (preoccupied, dismissing, fearful avoidant)	33

Table 1: Participants in each category for the caregiver questionnaire and the relationship scales questionnaire

Correlations for Analysis 1 (warm early caregiver score and a high secure adult attachment score)

Preliminary analysis was performed to ensure all z scores were within the +3/-3 range, with no outliers. Data was checked and was above interval level and scores were independent. Skewness and kurtosis were within the +1/-1 range, data was therefore normally distributed and parametric assumptions were met. The relationship between perceived warm parent/caregiver (as measured by Hazan and Shaver's (1986) parent/caregiver questionnaire, cited in Collins and Read 1990) and perceived secure adult attachment (as measured by Griffin and Bartholomew's (1994) relationship scales questionnaire) was determined using a Pearson's product moment correlation coefficient. The results of the analysis revealed a moderately large, linear, positive correlation between the two variables, r₍₆₆₎= .340, p < 0.002.

Correlations for Analysis 2 (insecure early caregiver score and a high insecure adult attachment score)

Preliminary analysis was performed to ensure all z scores were within the +3/-3 range, with no outliers. Data was checked and was above interval level and scores were independent. However skewness and kurtosis was not within the +1/-1 range; data was therefore not normally distributed for the insecure caregiver questionnaire, due to the majority of respondents scoring low on the insecure category. The relationship between perceived insecure parent/caregiver (as measured by Hazan and Shaver's 1986 parent/caregiver questionnaire, cited in Collins and Read (1990) and perceived insecure adult attachment (as measured by Griffin and Bartholomew's 1994 relationship scales questionnaire) was investigated using Spearman's Rho correlation coefficient. The results of this analysis showed a moderately large, linear, positive correlation between the two variables rho= .304, n = 68, p < 0.006.

Discussion

As predicted, warm caregiving was positively correlated with a high secure adult attachment (r₍₆₆₎= .340, p < 0.002) and high insecure caregiver scores were positively correlated with a high insecure adult attachment score (rho= .304, n = 68, p < 0.006).This demonstrates that subjects who perceived their relationship with their parents as warm and accepting were more confident they could depend on others, were comfortable getting emotionally close to others and were less anxious about being abandoned or unloved, as measured by the relationship scales questionnaire. In contrast, those who remembered their caregiving as 'insecure' were more likely to worry that others did not value them, or would prefer to feel independent and avoid closeness with others. This is consistent with Hazan and Shaver's (1987) results that adults who were secure in their romantic relationships, were more likely to remember their early caregiving as being warm and responsive (Feeney and Noller, 1990; Collins and Read, 1990; Mikulincer et al., 2009; Levy et al., 1998).

Overall, the results of this study support the theory that perceptions of early caregiving relationships do have an impact on adult romantic attachments; however, there are some contextual elements to this conclusion, which deserve further explanation.

Firstly, the measures for the perception of an early caregiver were self reports, rather than behavioural observations, therefore the current study focuses on how people 'think' they were raised which could be different to what actually happened. Obviously, due to time constraints it was not possible to do this over a longitudinal study; however, this needs to be considered for future research.

Secondly, the scores for cold and ambivalent parent/caregiver were combined to create an insecure early caregiver attachment. The pre-occupied, dismissing-avoidant and fearful-avoidant scores were also combined to create an insecure adult attachment. Although this decision was taken due to the low numbers of participants in the named categories, it does come with caution. For example, a dismissing-avoidant person will have very different traits to those of a preoccupied person (a preoccupied person is reluctant to get close to others and worries a partner will leave them, whereas a dismissing avoidant person is not comfortable relying on others and maintains a defensive sense of self-reliance and independence). For the purpose of future research more participants are required in order to gain a higher sample in each of the 'insecure' categories.

Another limitation to the study could be due to potential relationships that have recently formed. In the questionnaires, participants were not asked how long they had been in their current relationship. This could have an effect on the final scores as, if the relationship was very recent, as Bowlby states, this could not be categorised as an attachment relationship, due to the very nature of time (Bowlby, 1982). Any research looking into the area of early caregiving and adult romantic love should explicitly state when a romantic relationship becomes an attachment relationship. Hazan and Shaver (1987) did not address the possibility that some partners were attached and some were not. They also did not offer a method for making this distinction. For the purpose of future research, it is important to distinguish how long a respondent was in a romantic relationship for.

Future research must track individuals longitudinally, to clearly establish whether attachment styles remain stable across several relationships. It would also be advantageous in a future study to ask participants if they were parents themselves. This may have an impact on a participant's recollection of the parenting they received themselves.

Conclusion

The present study provides support for the hypothesis that positive interactions in early family life contribute to positive interactions in romantic relationships and therefore secure attachment. Negative interactions in early family life contribute to insecure representations in adult romantic relationships. This information is important in understanding both the initial development of and subsequent changes in attachment security. With adequate training and support, front line professionals such as Health Visitors and various early years staff can support parents by encouraging strong attachments from the outset, as well as encouraging parents to attend local parenting programs. This will ensure parents have the confidence to apply positive parenting skills and offer encouragement and positive attachments to their children.

Acknowledgements

This research was completed as an undergraduate dissertation at Staffordshire University. I gratefully recognise the following people, without whose support this research would not be possible. Many thanks to my supervisor, Dr Mani Das Gupta, for support and guidance throughout this research and also the 68 participants who volunteered their time and willingness to take part in this research.

Appendix 1 - Materials Used

Participant Information and Consent Sheet

Hi, my name is Lisa Dale. Thank you for taking the time to take part in this study.

The title of my third-year project is: 'does early caregiving have an effect on our relationships in adulthood?' This project investigates whether there is a link between early caregiving to parents and later adult relationships.

This experiment will last approximately 15 minutes, and is only a one-time attendance/completion session.

What is the purpose of this study?

The purpose of this study is to find out if there is a correlation between how a participant views their adult relationships and their own relationships with a parent/caregiver. If you only knew one of your parents you can still take part. It is important to think of one parent only (main caregiver) when completing the questionnaire. In relation to adult relationships (the relationship scales questionnaire), you can only take part in the study if you are currently in a relationship, please think of your current relationship as you are completing the questionnaire.

My project supervisor and I will only see the questionnaires and answers you provide. You are represented in this experiment by a number, which will be indicated at the top of your questionnaires. Therefore, you will remain completely anonymous in this experiment.

This experiment is based on voluntary participation. If you feel uncomfortable at anytime you are free to leave, without giving reason and withdraw your results, no questions asked. If you do have any queries please do not hesitate to contact me.

Now that you have read the above information and, if you would still like to participate in the study, please sign and date on the line below. This signature is to confirm that you have read and understand the terms of this study and are happy to proceed.

Participant Number: …………………………………………………

Age: ……………………………………………………………………………

Gender: ………………………………………………………………………

Signed: ………………………………………………………………………

Date:……………………………………………………………………………

(N.B. This document will not be kept with your results of the experiment.)

The relationship scales questionnaire (Griffin and Bartholomew, 1994)

Instructions:

Please read each of the following statements and rate the extent to which you believe each statement best describes your feelings about your current relationship.

1. I find it difficult to depend on other people.