Assessing Attachment Classification Difference Between Indigenous and Non-Indigenous Children

Attachment classification distribution is largely considered to be consistent across populations, with normative patterns indicating an approximate 60% secure and 40% insecure distribution (McLean, 2016). Some deviation to this has been observed among samples of at-risk children, including those of mothers suffering from depression and post-traumatic stress disorder (Enlow et al., 2014), maltreated children (Boris et al., 2004), those of low socioeconomic status (Cyr et al., 2010), and those removed from families and placed into care (Stoval & Dozier, 2000). While mediating factors may provide insight into such deviations in classification distribution among high-risk samples, research has also questioned whether this distribution is found among non-Westernized cultures and the universal applicability of the tools used to assess it (Van Ijzendoorn & Kroonenberg, 1988).

Culture as an influencing factor on attachment formation has garnered much interest since Mary Ainsworth (1967) began developing the Strange Situation Procedure (SSP), widely considered to be the gold standard in attachment assessment among infants and young children. Informed by her observation of separation and reunion incidents between Ugandan infants and their mothers during the 1950s, Ainsworth established a tripartite classification system of attachment (Van Ijzendoorn & Sagi-Schwartz, 2008), categorizing infant–mother attachment as either avoidant (A), secure (B), or resistant/ambivalent (C) [with Main & Solomon later adding a fourth classification—disorganized (D) in 1986]. Later, Ainsworth and Wittig (1969) sought to apply the SSP to samples of American mother–infant dyads with similar distributions being observed. These findings paved the way for further testing of the SSP by other researchers sampling across a variety of populations, including parent–infant dyads from Japan (Takahashi, 1986), Puerto Rico (Carlson & Harwood, 2003), South Africa (Tomlinson et al., 2005), and Czechoslovakia (Masopustova et al., 2023), ultimately lending support to the notion of a universal attachment distribution across cultures.

Researchers have increasingly considered the role cultural norms and values play in influencing caregiving practice and questioned the validity of current measures to assess attachment within non-Western contexts (e.g., Gaskins et al., 2017). In their review of the contextual and universal dimensions of attachment across cultures, Van Ijzendoorn and Sagi-Schwartz (2008) observed that the majority of cross-cultural attachment research has at its foundation a fundamentally Western theoretical basis and methodology elicited to observe behavior (an etic perspective), which subconsciously disregards the value of behaviors specific to cultures to gather insight (an emic perspective). Cowan and Cowan (2007) noted the paucity in research investigating cross-cultural factors that may produce variation in attachment distribution, which may partially account for a lack of culturally specific tools to approach cross-cultural attachment from an emic perspective. This research gap has consequently contributed to theories of attachment and measurement tools developed in the West being tested for their cross-cultural validity, leading to a questioning of the plausibility of applying tools based on typically Western values and principles to predominantly collectivist populations (Rothbaum et al., 2000). Such measures may lack sensitivity to identify culturally specific environmental and socioemotional factors that contribute to the development of attachment within other cultures and potentially lead to incorrect assumptions about attachment classifications between Western and non-Western populations (Kline et al., 2018). It is therefore critical to understand potential mediating factors that give rise to divergent attachment distributions when using non-Western samples within the current body of literature, which may provide the foundation for the development of culturally specific attachment measures.

Measuring Attachment Classification Across Cultures

While it is established that much cross-cultural attachment research is approached from an etic perspective, the SSP continues to be lauded as a culturally valid attachment measure yielding a generally consistent classification distribution across populations. It is considered capable of providing enough sensitivity to allow for the identification of potential cross-cultural differences (e.g., Takahashi, 1986). Van Ijzendoorn and Kroonenberg (1988) demonstrated an emergence of attachment clusters particular to certain populations as part of their meta-analysis of 1,900 attachment classifications across eight different countries (Germany; Japan; America; Israel; the Netherlands; Great Britain; Sweden; and China). It was identified that although secure classifications were most prevalent across all populations (congruent with typical SSP observations), avoidant classifications were more evident among Western European countries and resistant/ambivalent classifications more characteristic in Israel and Japan. Explanations for such variations were limited to possible disparity in socioeconomic status of participants between studies and propagation of standard parenting concepts originating from the West, with little exploration of potential culturally specific moderating factors.

The SSP has also proved able to identify unexpected attachment classifications across cultures where mother–infant behaviors are not dissimilar to the typically Western parenting styles from which the SSP was modeled. In studying attachment behaviors of the Dogon ethnic group of West Africa, True et al. (2001) identified a 67% secure/0% avoidant/8% resistant and 25% disorganized distribution among mothers and their infants. As with Van Ijzendoorn and Kroonenberg’s sample, secure attachment was observed to be the dominant classification; however, a high percentage of disorganized attachment was also observed among this group. Such classifications within a population whose parenting style was characterized by prolonged periods of immediate proximity and responsiveness appears incongruent with a key component of attachment theory relating to sensitivity and development of secure attachment (Neckoway et al., 2007). True et al. (2001), however, suggested that these disorganized classifications were likely due to the level of stress placed upon infants by the unfamiliar and inconsistent separation and reunion protocols of the SSP, which were not typically experienced by infant–mother dyads in the Dogon ethnic group.

As well as inter-cultural deviations from expected distributions, the SSP has produced intracultural variations in attachment behavior within populations of the same religious and collectivist backgrounds. Sagi et al. (1994), in their application of the SSP with two groups of Jewish infants residing within various Israeli kibbutz, identified higher instances of insecure type classifications (52%) among infants residing in communal sleeping arrangements with multiple non-parental caregivers than those being cared for within their homes by primary caregivers (35% insecure). Such classifications may be expected given the benefit to be gained through increased proximity of an infant to their primary caregiver, and Ainsworth (1967) too demonstrated, as part of her Ugandan observations, that the presence of multiple caregivers alone was not a predictor of attachment security, but rather the presence of a mother–infant bond. Interestingly, Sagi et al. did not identify any significant differences in daytime interaction between both groups of infants and their mothers, perhaps indicating that insecure attachment classifications observed in one point of time may not entirely reflect a general or accurate level of attachment security. Such findings reinforce the importance of emic-informed measures to account for those differences and provide a clearer explanation to avoid incorrect assumptions being made about the attachment relationship between a child and their caregiver.

Attachment in Indigenous Cultures

Attachment research within Indigenous cultures has received less focus than other collectivist and Western cultures, resulting in a chasm in knowledge as to the fidelity of current measurement tools utilized within Indigenous cultures. To inform culturally valid measurement tools, it is crucial to consider whether classic attachment concepts are observed within and can be applied to Indigenous cultures. For example, the sensitivity hypothesis surmises that secure attachment is the result of sensitive caregiving (Ainsworth et al., 1978^, 2015), in which infants with caregivers responsive to their needs are hypothesized to be more securely attached as their needs are adequately and appropriately met. What denotes responsive parenting, however, is shown to vary interculturally. New Zealand Māori mothers strongly connected to traditional Maoridom culture, for instance, typically co-sleep with infants to alleviate their distress upon separation, thus subsequently responding to their need for comfort during sleeping periods (Jones et al., 2017). Such parenting practice can be seen as congruent with Ainsworth’s sensitivity hypothesis; however, some unresponsive Western practices, such as allowing crying infants to self-soothe through separation during sleep (Beeghly & Tronick, 2011), challenge this hypothesis. The rationale behind such practice allows for the development of self-regulation skills (Weinraub et al., 2012) and future independence, hallmarks of individualistic culture characteristic of countries including contemporary New Zealand, in contrast to interdependence favored in more collectivist populations. However, how such practice among Māori mothers translates into infant attachment classification remains unknown, owing to the dearth in such research. The sensitivity hypothesis may also not fully account for the presence of multiple attachment systems in some Indigenous cultures, particularly where primary caregiving is shared among several figures (Meehan et al., 2013). Meehan et al. (2013) observed that infants among the Indigenous Aka foragers received responsive caregiving, including breastfeeding, from multiple group members while maintaining a stable bond with their birth-mother. While a Western model of caregiving may not typically favor responsiveness from multiple caregivers, Keller (2021) found that exposure to multiple attachment systems provides children with greater social, emotional, and physical attention, and stimulation compared to those with a single attachment relationship. Furthermore, the sensitivity hypothesis may overlook that, in some Indigenous cultures, the concept of attachment itself may extend beyond the immediate family to include ancestors, land, and water elements essential to Indigenous identity and wellbeing (Johnson-Jennings et al., 2020). Although this may be perceived as an abstract form of responsive parenting when viewed through a Western lens, it represents a critical developmental foundation within Indigenous cultures.

Another fundamental principle within attachment theory, the secure-base hypothesis (Bowlby, 1988), suggests that infants and children who are securely attached display a greater readiness to venture away from their caregivers to explore and engage in play. This stands in contrast to those who exhibit insecure attachment patterns (Straus, 2017). Bowlby believed that this inclination results from the formation of an internal working model (IWM) of attachment, allowing infants and children to maintain an internal representation of their caregiver even when physically distant. This internal model accordingly provides reassurance that their caregiver will be accessible upon their return.

The interpretation of the secure-base hypothesis in collectivist cultures, like that of the Australian Aboriginal and Torres Strait Islander peoples, presents a unique challenge. In some Indigenous communities, there is an observed tendency to discourage infants from exploring away from their caregiver(s) until they reach 2 years of age (Malin et al., 1997). This raises questions about how the secure-base hypothesis applies in such contexts. From a Western viewpoint, an infant’s hesitance to stray away from their caregiver might be interpreted as a sign of insecure attachment. However, Yeo’s (2003) research, informed through discussion with Aboriginal and Torres Strait Islander community members, indicates that such parenting practice can be implemented as a means to strengthen community connection, thereby fostering a positive Aboriginal identity and sense of security. This may be considered a more valuable component of attachment formation, serving to support the survival and continuity of Australian Indigenous culture. In contrast, Canadian Inuit infants and small children are afforded greater freedom to explore and engage in a diverse range of activities away from their caregivers, with their autonomy encouraged based on their individual capabilities (Pauktuutit Inuit Women of Canada, 1990). While within a Canadian Inuit community this is regarded as vital for progression through developmental stages, a Western perspective might perceive such freedom as indicative of a more permissive parenting style with less control, potentially raising concerns about a child’s safety.

Attachment security is widely believed to have a lasting impact on an infant’s social competence throughout their life, as suggested by the competency hypothesis (Weinfield et al., 2008). However, it is important to note that the concept of social competence is culturally relative and may depend on the extent to which a culture values independence versus interdependence. Many Indigenous cultures, for instance, prioritize values such as early self-reliance, fostering interdependence, displaying restraint of negative emotions (especially toward Elders), and emphasizing loyalty to community and spiritual connectedness. These values differ from the Western approach, which encourages self-exploratory behavior, self-expression, emotional regulation, and resilience in children (Ryan, 2011). In the case of Australian Aboriginal and Torres Strait Islander children, these culturally rooted behaviors may be mistakenly assessed as signs of insecure attachment, stemming from parenting practices viewed negatively in Western countries (Yeo, 2003). Such limitations recognized within attachment theory therefore call for a cautious approach when considering the use of standardized attachment tools in Indigenous communities (e.g., Neckoway et al., 2007) or indeed any other non-Western culture.

Ramifications of Incompatible Attachment Measures in Indigenous Cultures

Conducting attachment assessments that fail to align with cultural norms, or do not consider diverse caregiving systems, can have profound consequences on the lives of children from culturally diverse backgrounds, particularly if they reflect inaccurate classifications. Attachment assessments are increasingly utilized in family court proceedings to guide decisions related to the removal of a child from families due to welfare issues (Yeo, 2003), the assessment of risk posed by caregivers (Choate et al., 2019), and decisions relating to custody (Malin et al., 1997).

In Australia, child protection systems consider professional judgment to assess a child’s vulnerability to harm (Harnett & Featherstone, 2020) and a caregiver’s ability and willingness to protect their child (Department of Children, Youth Justice and Multicultural Affairs, 2021). These judgments include assessing risk, often based on observations of family functioning, and an assessment of attachment between a child and their caregiver (Department of Child Safety, Seniors and Disability Services, 2023). Attachment assessments, therefore, play a pivotal role in Australian court proceedings, influencing child removal from, and reunification with, family (Yeo, 2003).

Similar attachment measures, rooted in attachment theory, have been used to inform risk assessments affecting the permanent placement of children in Canadian court proceedings, including post-removal from families of origin (Choate et al., 2019). Crichlow (2002) expressed concern about the preference of Canadian family courts to utilize a predominantly Western approach to the child-welfare system fundamentally incongruent with traditional Aboriginal family systems, with Carriere and Richardson (2009) adding that the practice of applying culturally uninformed assessments to Indigenous families carries a risk in failing to identify appropriate and safe parenting within the context of traditional Aboriginal parenting.

Globally, Indigenous children are disproportionately represented in child protection systems compared to their non-Indigenous peers (Harnett & Featherstone, 2020^; Yeo, 2003) and face a higher likelihood of being placed in out-of-home care (Krakour et al., 2018). In Australia, despite constituting only 6% of the total population, Aboriginal and Torres Strait Islander children (aged 0–17 years) account for a significant 43.7% of the out-of-home care population (Australian Government, 2023). Similarly, in New Zealand, Māori and Pacific Islander children account for 64% of those placed in out-of-home care, a stark over-representation compared to the predominantly Pākehā (NZ European) population (Atwool, 2016). In North America, Native American children, constituting only 1% of the overall child population, represent 2% of those in foster care (Lawler et al., 2012), while in Canada, despite comprising just 7.7% of all children under 15 years, Indigenous children make up 53.8% of those in foster care placements (Statistics Canada, 2022).

While it is not implied that the removal of a child from their family solely occurs because of culturally incongruent attachment measures, there are evident limitations in applying attachment theory to Indigenous cultures, which casts doubt about the reliance on Western-developed attachment measures to inform critical decisions concerning Indigenous children. Consideration and knowledge of attachment concepts in Indigenous cultures are therefore critical to comprehend family structures and correctly interpret the attachment process relating to those communities (Lindstedt et al., 2017).

This systematic review aimed to provide a comprehensive and unbiased examination of relevant studies to answer the following questions:

1. Is there a reported difference in attachment distribution between Indigenous and non-Indigenous children?

2. What attachment measures are used with Indigenous samples?

3. Are attachment measures modified to consider specific cultural factors?

It was anticipated that findings could draw further attention to the problematic use of standardized assessments of attachment cross-culturally, which may disadvantage Indigenous populations, particularly children, and contribute to the growing body of evidence advocating for alternative means of assessing attachment in Indigenous populations.

Method

The Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRIMSA; Page et al., 2021) was used to guide this systematic review to identify relevant English language, peer-reviewed, research articles that used attachment measures as per the inclusion criteria outlined below. Qualitative and quantitative research articles, of any year, were included.

Study Selection

The initial search, conducted between March and May 2022, yielded 5,980 potentially suitable studies, from which 2,251 duplicates were removed using Endnote software. Title and abstract screening of the remaining 3,729 studies, and 1,227 records identified from the citation search, were conducted against the inclusion criteria by S.-L.B.T., from which 204 studies from the initial search and 184 studies from the citation search were subject to full-text screening by S.-L.B.T., and partial screening by PhD candidate A.L. S.-L.B.T. and A.L. conducted independent screening of the 204 initial studies (three of which could not be retrieved), and removed 198 studies assessed as ineligible (Figure 1 shows the study selection process). Next, S.-L.B.T. completed full text screening of the 184 citation studies (two of which could not be retrieved) and removed 171 records, resulting in a combined total of 14 eligible studies being included within the review. A re-run of the search completed in October 2023 resulted in one additional study, bringing the total number of eligible studies included in the review to 15.

Open multimedia modal

Data Extraction

Data extraction of the 15 studies was completed by S.-L.B.T. and S.M., using an Excel spreadsheet to record relevant data for each, including date and country in which the study took place, target population type (e.g., Australian Aboriginal and/or Torres Strait Islander, NZ Māori, etc.) and size, comparator population type (e.g., Anglo-Australian, NZ European, etc.) and size, attachment intervention measure applied, distinct attachment classifications for target and comparison populations, and indication that cultural factors had been considered during study design (see Table 1). Given inconsistency across the studies in how attachment classifications were reported (i.e., some studies reported secure/insecure classifications only, whereas others reported secure/ambivalent/resistant insecure/disorganized attachment), insecure types were aggregated and classified as one category during data extraction. In addition, some studies included more than one measure of attachment over a defined timespan; however, this was often associated with participant attrition. As it was not possible to identify the cultural background of participants who left such studies, only the first-time attachment measure was reported on for this review. Furthermore, only those participants who had a recorded attachment score within the results section of each study were included within the recorded sample sizes. Any discrepancies were resolved between S.-L.B.T. and S.M. throughout the data extraction process.

Open multimedia modal

Results

Discussion

During the review period, literature on attachment classification in Indigenous populations was limited, particularly in studies focused on Australian, New Zealand, and Canadian Indigenous populations. While a small number of attachment-related studies included Native American participants, their representation was substantially smaller than participants from comparison demographics, and there were no Indigenous-specific classifications reported among the findings. Such underrepresentation is not surprising, given the estimate that 90% of sampling pools in psychological research consist of American and European participants (Nielsen et al., 2017). The deficit in Indigenous representation across attachment-related research is problematic and exposes a significant research gap which hinders our understanding of the cross-cultural validity of attachment measures, and risks reinforcing an assumption that cultural contextual factors do not necessarily influence attachment behavior and classification. Such a notion may therefore be the result of low Indigenous participation and limited representation across attachment research, rather than empirical evidence.

Several issues may help to explain this deficit. For instance, although notably absent in attachment research, Indigenous populations and communities are over-researched across other fields, particularly health (Bainbridge et al., 2015). Bainbridge et al. (2015) noted that Indigenous communities often question the value of research they participate in, which tends toward further exposure of Indigenous disadvantage and a failure to translate findings into beneficial and tangible outcomes. Such repetitious experiences can result in a hesitancy of Indigenous communities to participate in research, particularly if facilitated by institutions perceived to be incongruent with, or unwilling to incorporate, Indigenous values and context (Kelley et al., 2013). This may be further complicated by a distrust toward non-Indigenous researchers who may portray cultural inaccuracies or fail to distinguish intracultural nuances, researchers perceived to conduct poorly designed research, or those who fail to consult with Indigenous communities as part of their design process (Thomas et al., 2014), as was evident across the studies within this review. Some researchers may further fail to safeguard Indigenous Cultural Intellectual Property (Arts Law Centre of Australia, 2024) or adhere to First Nations Principles of Ownership, Control, Access, and Possession (First Nations Information Governance Centre, 2024). Research led by and within Indigenous communities, with the active support of the community and involvement of Indigenous scholars, may go some distance in addressing this underrepresentation in research and additionally provide valuable insights into how findings can be meaningfully interpreted within these communities.

Another key issue affecting Indigenous recruitment into research may lie in the locations selected from which to conduct a study. The majority of studies included within this review were conducted in metropolitan areas, utilizing local participant samples that might not typically represent Indigenous communities. In Australia, as is often the case for other countries with Indigenous populations, a significant portion of Aboriginal and Torres Strait Islander communities reside within regional or remote areas, with those in urban areas often being dispersed (Australian Bureau of Statistics, 2021). Sampling exclusively from Indigenous populations in metropolitan areas may therefore lead to findings that do not fully represent the attachment systems, family structures, or kinship networks prevalent in rural or remote Indigenous communities (Australian Institute of Family Studies, 1993). To address this issue, cross-cultural attachment research should incorporate regions with higher Indigenous populations, in addition to metropolitan areas, to ensure more accurate and inclusive representation.

Although this review expanded its scope to encompass various categorical-based attachment measures, 12 studies adopted the SSP in their methodologies. Six of these studies showed a normative 60% secure and 40% insecure distribution consistent with previous research using this measure, whereas the remaining six using the SSP observed lower rates of secure and higher rates of insecure attachments. While this variability cannot be attributed to the non-significant representation of Indigenous participants, it is important that assumptions are not made about the generalizability of such classifications to Indigenous populations or participants of other cultural backgrounds included within these studies without further investigation into the causes of such variance. This is also evident for other attachment measures used in the remaining studies (k = 3), such as the AAI, Story Stem Task, and Global Rating Scale, which too yielded diverse distribution results.

The review further highlighted a notable limitation in adjustments made to methodologies in consideration of cultural contextual factors, which may have contributed to variance in observed attachment distribution. Given the widely recognized impact of culture on behavior, its incorporation into research design is critical to avoid invalid behavioral observations and outcomes (Awad et al., 2016). While none of the reviewed studies directly investigated the link between cultural factors and attachment, one study (Kerns et al., 2011) made minor adjustment to a measure to ensure cultural relevance to American participants, and two others (Pasco Fearon & Belsky, 2004^; Spieker & Bensley, 1994) included naturalistic observations to examine behavior within the home environment. However, no adjustments were made to any study to ensure measures aligned to Indigenous contexts. The findings across these studies may therefore be attributable to methodologies that overlook key cultural concepts or a failure to adapt for participants from varying cultural backgrounds. Such gaps in attachment research design might be addressed by considering and incorporating Indigenous caregiving models, which offer valuable insights into attachment concepts that contribute to healthy Indigenous child development (Morelli et al., 2018). The Indigenous Connectedness Framework (Ullrich, 2019), for example, highlights the importance of fostering connections between an infant and their environment through establishing relationships with family, community, intergenerational ties, spirituality, and land to support healthy development. Unlike classical attachment theories, this framework has at its core multiple attachments that contribute to an Indigenous concept of wellbeing.

Similarly, the Social and Emotional Wellbeing (SEWB) model (Healing Foundation, 2020), used by Australian Aboriginal and Torres Strait Islander peoples, views kinship and intergenerational connections as an attachment network spanning generations. As in the studies reviewed, culturally informed models appear largely unacknowledged by cross-cultural attachment researchers, despite their potential to inform research methodologies and subsequent measures and broaden our understanding of attachment across diverse cultural contexts.

These observations reinforce the importance of developing foundational knowledge of attachment systems across diverse cultures, which could be achieved through naturalistic and observation-based methods (e.g., Morelli et al., 2018). Dedicated Indigenous attachment research is essential not only to validate Indigenous attachment systems but also to move beyond assumptions of universality and improve inclusivity in research (Kline et al., 2018). Emic-based methods may further support the creation of culturally valid tools by identifying and accounting for culturally specific caregiving practices. For example, Spieker and Bensley’s (1994) findings suggest that infant attachment security may be influenced by grandmother support, which varies with living arrangement and by race. However, there is insufficient understanding of how race and cultural factors influence attachment security and the mechanisms through which this occurs. This gap leaves a lack of culturally grounded content on which to build relevant measures, which may prove beneficial in determining what insecure attachment looks like in other cultures and how to appropriately respond.

In addition, clear reporting of culturally specific attachment classifications was absent in this and all other studies. This reporting is essential for identifying culturally unique patterns of attachment formation and to provide future directions for Indigenous-specific attachment research.

Despite a paucity within attachment studies, research informed by emic perspectives has proven effective in generating culturally validated findings in various fields. For example, emic-based measures have broadened our understanding of children’s psychosocial development across cultures (Lansdown et al., 1996), facilitated cross-cultural diagnosis and prognosis of mental illnesses (e.g., Ahmad et al., 2000), and expanded our knowledge of culturally specific responses to key life events such as grief (Killikelly & Maercker, 2023). Moreover, they have contributed to culturally relevant service provision, such as assessing and responding to domestic violence in diverse immigrant populations (Canadian Domestic Homicide Prevention Initiative, 2008). Emic-based approaches have also led to the development of measurement tools tailored to Indigenous populations across various research fields. These tools include assessments of health and wellbeing in Canadian Indigenous populations (d’Agincourt-Canning et al., 2024), interventions to treat depression and anxiety in Native American youth (Listug-Lunde et al., 2013), self-report measures of identity and wellbeing in New Zealand Māori (e.g., Manuela & Sibley, 2013), and therapeutic outcomes for Australian Aboriginal and Torres Strait Islander peoples (e.g., Copley et al., 2021). The failure of attachment research to advance measurement tools to ensure their applicability to Indigenous populations perpetuates a limited understanding of cultural influences in attachment dynamics and classification. This issue is particularly problematic when attachment measures inform assessments to guide decisions affecting the wellbeing of Indigenous children and young people, such as in family court proceedings. These potentially inaccurate measures risk subjecting Indigenous people to unnecessary or unjust service responses, such as the removal and sustained separation of children from their families, based on incorrectly perceived instances of unsafe parenting (e.g., Choate et al., 2019). Among Indigenous populations, such practices may be perceived as a continuation of policies enforced during colonization. This risks retraumatizing families already affected by the forced removal of children during that period and potentially creating a new generation of “stolen children” (Australian Institute of Aboriginal and Torres Strait Islander Studies (2023). The paucity of research from target countries, known to have an over-representation of Indigenous children and young people in child protection systems, highlights a concerning omission.

All reviewed studies sought to investigate the relationship between attachment classification and a particular phenomenon and accordingly drew conclusions about the effect of secure and insecure attachment types on that particular phenomenon. However, there were inconsistencies between the studies as to what, if any, confounding factors were present and how they were controlled. Key factors such as parental trauma history and presence of support systems, significant life events, relevant infant medical history, sibling ordering, family stress, and environmental factors or exposure to bullying were not routinely considered as part of the designs yet were relevant factors that had the capacity to influence a participant’s performance or behavior during testing. The disregard of culturally relevant compounding factors across the studies may be on account of the relatively high ratio of participants from comparison populations to those target populations. Resultingly, these studies present a narrowed perception of the influence attachment classification had on selected behaviors and overlooked the opportunity to investigate whether cultural factors may moderate those relationships.

Conclusion

This systematic review is the first to investigate variations in attachment distribution between Indigenous and non-Indigenous populations, identify measures employed to determine those classifications and assess adaptations made to methodologies that account for Indigenous-specific cultural factors. However, the distinct lack of specific research investigating attachment formation and classification in Indigenous populations was a significant issue that hampered the review’s objective to adequately answer its research questions. As a result, robust conclusions from the synthesized literature regarding the association of specific attachment classifications with Indigenous populations were unable to be drawn. A limited variety of attachment measures were employed, as well as a lack of adjustment to account for a small number of Indigenous participants, further compounded these objectives, providing little guidance as to how best to measure Indigenous attachment. This systematic review concludes by highlighting the ongoing need of targeted and culturally informed research to explore potential underlying differences related to caregiver practice, cultural context, or family structure within Indigenous populations to aid understanding of their relationship with attachment classification, which in turn may translate to significantly broad societal, political, and contextual implications, and benefits that ensure better outcomes.