The Controversy of Infant Attachment Styles Changing after One Year
In the field of developmental psychology, attachment from early in a person’s life is significantly impactful to other aspects in their later life. Attachment that is typically established between a child and their parent can indicate how the child’s personality will develop, their ability to create relationships and socialise appropriately as well as regulate emotions. The theoretical origins of attachment were coined together by psychologists, most famously Bowlby and Ainsworth, who drew upon ideas from ethology, information processing, cybernetics and psychoanalysts. Furthermore, studies from Bowlby and Ainsworth were prompted by key concepts discovered by Freud and Vygotsky, both highly regarded for their work in developmental psychology (Bretherton, 1992). Bowlby is a renowned psychologist for his discovery of the attachment theory. In 1973, he put forward an idea that humans are naturally born with an embedded behavioural pattern that aid relationships through a sequence of interactions with the environment (Bowlby, 1973, as cited in Gomez, 1997). Bowlby had proposed two hypotheses in that a) the history of interaction between an infant and caregiver greatly affects the quality of infant-caregiver attachment relationships, and b) the quality of infant attachment become the basis of individual personality differences in later life (Sroufe, 2005). Sroufe (2005) also suggests attachment is essential to the development of a person and infant-parent relationships is the core. Infant attachment styles can be tested with multiple methods, however the most acclaimed task to categorise infants’ attachment was devised by Mary Ainsworth (Ainsworth et al., 1978, as cited in Bretherton, 1992). The task used in her report research has been referred to as the Strange Situation Task (SST). It has since been utilized in further studies to identify infant attachment types. The SST involves a parent, usually mother, and their infant together in a room. Whilst playing with toys, a stranger is introduced. The mother leaves shortly after and the child properly encounters the stranger. The response elicited by the child when their parent leaves and returns is a strong indicator of their attachment style. Through this method, the child would be classified in one of three types: secure, insecure anxious-ambivalent or insecure anxious-avoidant. In further studies by Ainsworth and her colleagues, a fourth classification of attachment type was added, called disorganised or disoriented attachment (Sroufe, 2005). With all this in mind, it has been strongly debated within the developmental field of psychology, that infant attachment styles, as measured by the Strange Situation Task, does not change after one year; thus, later impacting other aspects of life. The purpose of this essay is to analyse both sides of the argument to decipher whether or not attachment styles change after a year.
Each side of the argument advocates for attachment styles staying consistent or becoming inconsistent after the first year of life, given that the SST was used to identify attachment as an infant and repeated in some way after. One may argue that attachment styles change due to inevitable growth while another may gather data to conclude it does not change. Instead, attachment becomes the foundation for other aspects to build upon. Both sides use the driving theory of attachment by Bowlby to back up their perspective. As stated previously, the strange situation task by Ainsworth et al. (1978) is a pinnacle basis of understanding attachment, through its categorisation of infant attachment styles. This task, along with further knowledge of theoretical origins, can be used to decide whether attachment styles change after the first year of life or stays consistent. To accurately make a decision and gain proper insight to the debate, two longitudinal studies have been chosen to review and compare results found from the SST and the relationship of attachment after one year. A study by Becker-Stoll et al. (2008) discovered no evidence for the idea that attachment styles stay the same but to oppose this, a study by Sutton and Fox (1994) found that it does stay consistent.
To support the impression that attachment style does indeed change, Becker-Stoll et al. (2008) found no evidence for attachment stability in their longitudinal study. The fifteen-year follow-up experiment followed children aged one to sixteen. The experiment was completed to identify if infant attachment was related to adolescents’ interaction with their mother. Participants in the study consisted of families from all social class however majority belonged in lower-middle class. Fifty-one infants aged 12 or 18 months were tested using the SST to identify their attachment to their mother and father. At six years of age, forty children were then recontacted for a reunion test, similar to the SST performed when they were infants. Both assessment methods of attachment placed children in one of three categories. These included the first three attachment types; secure, insecure-avoidant and insecure-ambivalent. Forty-three adolescents out of the original fifty-one, took part in a four-hour university visit as part three of the study. They were interviewed using the Adult Attachment Interview (AAI) to note any life events that could have influenced a fluctuation in attachment style. Becker-Stoll et al. (2008) also acknowledges other works that find continuity and discontinuity of attachment development however in their study, they reported no significant evidence was manifested to support continuity. Thus, Becker-Stoll et al. (2008) concluded infant attachment styles do change due to the extensive amount of negative experienced faced by the individual and how these experiences shape other characteristics. It was highlighted as well that other sources identify more instability is inclined when a greater number or greater effect of negative life events occur. This is supported by the amount of risk factors associated to the individual. By identifying these risk factors and other factors contributing to the change in attachment, it is a strength in the study to appropriately back up their conclusion with other research. However, Becker-Stoll et al. (2008) also identified limitations, such as their concern over the small and varying sample size. For the purpose of this essay, it is a limit that only attachment to their mother was focussed on after the initial test, even though fathers were used during the first stage. This is understandable for the study as they were concentrating on adolescent-mother relationships. Overall, Becker-Stoll et al. (2008) had been able to justly demonstrate changes in attachment styles in infants to adolescents.
On the other hand, Sutton and Fox (1994) challenges the points raised by Becker-Stoll et al. (2008) in that attachment was used to predict children’s maturity level in their ability to engage in joined activities; consequently, showing evidence for consistency in attachment types. Sutton and Fox (1994) gathered 48 participants and used the SST at 14, 24 and 48 months. They belonged to families that were predominantly Caucasian and raised within middle to upper-middle class. The aim of their study was to test attachment security between a mother and child, mainly focussing on how children regulate emotions as they work through activities at 58 months. One of their hypotheses that aids this side of the argument was that children’s attachment style would be stable across all months. Their data somewhat supported this hypothesis in that consistency was patent at 14 to 24 months, but not significantly enough to be counted at 58 months. However, at 58 months, Sutton and Fox (1994) also categorised children into the fourth style; disorganised attachment. Therefore, some form of stability was captured after the first year of life and across the several months of testing. In addition, Sutton and Fox (1994) predicted the proportion of mothers’ acknowledgement of their child’s emotion and the child’s attachment style at 14 and 24 months made an impact to the maturity level at 58 months. An example from their research is that children classified to have an avoidant attachment style at 24 months correlated to lower maturity level on tasks and joined activities at 58 months. By concluding attachment styles are stable enough to predict maturity levels in certain tasks, it is reasonable to suggest their results show attachment styles do not change after a year. The experimenters identified a few limits in their design, most based on their sample size being too small with a narrow range of participants. They elaborate on how their overall conclusions may be hard to generalise to the wider population. On a positive outtake, a strength in their study would be the utilisation of SST, as thoroughly described in their method section. It is clear that Sutton and Fox’s (1994) experiment supports the idea that infant attachment styles stay consistent from the first year of life and onwards.
Without a doubt, both sides of the argument raise important experimental findings to support or deny the concept of infant attachment styles being consistent. The most notable reason for change to occur in the attachment as an infant is due to negative experiences faced while they mature. In a similar study to Becker-Stoll et al. (2008), Waters et al. (2000) found opposing views that correspond to Sutton and Fox (1994). Waters et al. concluded that while there is evidence for stability, the factors contributing to change in classification is inevitable; i.e. a secure attachment forms into an insecure attachment and vice versa. These negative impact ranges from parental divorces, life threatening illnesses to a family member to more personal experiences such as physical or sexual abuse. Sixty-four per cent of their participants had consistent results from the SST as infants and in the AAI as adolescents. Similarly, another study by Hamilton et al. in 2000 gathered data using both the SST and AAI to assess attachment. They discovered over three-quarters (77%) of adolescents maintained their attachment. Conversely, McConnell and Moss (2011) identified other research following similar methods of SST and AAI that indicated consistency. They reviewed literature from Aikens, Howes and Hamilton (2009) who concluded a quarter of their participants represented signs of attachment development. Furthermore, Weinfield, Scroufe and Egeland’s (2000) experiment has also been known to support the notion of attachment instability. They concluded a stability rate of 38%, ultimately suggesting no significant evidence for attachment to be the same at an infant’s age to an adult. All of these sources, though substantial in evidence for each side of the debate, are limited to its generalizability. The majority of the participants chosen are from Caucasian backgrounds, primarily fitting the WEIRD population. This has been a concern in the field of developmental psychology as it prevents an accurate overview of data.
Ultimately, the purpose of this report was to underpin data that lead to the controversial debate of whether or not infant attachment styles change after one year. It is vital to understand attachment as it has detrimental effects to other developmental aspects through the lifespan. Two articles used Ainsworth’s Strange Situation Task and were chosen to conclude opposing sides of the argument. Becker-Stoll et al. (2008) revealed there was no significant data to support continuity of attachment development. Attachment styles naturally varied depending on the amount of negative experiences encountered. This construct was reinforced by the findings from Waters et al. (2000) and Hamilton (2000). In contrast, Sutton and Fox (1994) contests against their ideas in that attachment does change. They found infant attachment styles at age 14 and 24 months to be consistent and subsequently predicted maturity at 58 months. Hence, some consistency must have been evident among infancy years to predict early childhood behaviours. Aikens, Howes and Hamilton (2009) and Weinfield, Scroufe and Egeland (2000) were reviewed in McConnell and Moss’s paper (2011) to assist these views. They contributed to the same notion of attachment styles cannot significantly develop. A major concern with each study is the lack of random sampling. This limit is seen through the chosen pool of participants being mainly Caucasians. The problem with this is that the world is diverse. The findings make it difficult to generalise to a global context. To add to this, longitudinal studies are also difficult to execute. Participants drop out through the years, decreasing the power of the study and also it is confounded to one generation. In summary, the essay has illustrated evidence for stability of infant attachment styles in the first few years of life, however to conclude consistency across the whole lifespan is an overstatement. With experiences and growth occurring naturally through development, the attachment formed as an infant is subjective to change throughout the course of their growing life.