Attitudes towards Aboriginal and Torres Strait Islander peoples in Australia: a systematic review

ABSTRACT Objective Aboriginal and Torres Strait Islander peoples have been the longstanding targets of racism and discrimination in Australia. This paper presents the findings of a systematic literature review designed to identify and synthesise the available evidence exploring these intergroup attitudes, and the factors that correlate with them. Method Searches were conducted in Medline, Psychology and Behavioural Sciences Collection, ProQuest Psychology Database and PsycINFO databases. Eligible studies were required to measure attitudes of non-Indigenous Australians towards Aboriginal and Torres Strait Islander peoples. Results Twenty studies met the eligibility criteria (n = 2,958). The results outlined: (a) the prevalence of negative attitudes, (b) attitudes were most commonly conceptualised as modern racism, and (c) that there are a range of factors that are associated with negativity towards Aboriginal and Torres Strait Islander peoples. Conclusions The review outlined the relationship between attitudes towards Aboriginal and Torres Strait Islander peoples and demographic, ideological, individual differences, and intergroup factors. This review highlights the need for continued research in this domain to inform appropriate prejudice reduction strategies. KEY POINTS What is already known about this topic: Aboriginal and Torres Strait Islander peoples face continued adversity, prejudice, and discrimination on their own land. This heightened propensity to be targeted for prejudice is a known social determinant of poorer health and wellbeing. Understanding racial attitudes towards this group is an imperative step in understanding and combating this health disparity. What this topic adds: There is a paucity of academic research exploring negativity towards Aboriginal and Torres Strait Islander peoples The limited quantity of evidence does establish a range of correlates with negative attitudes that can be clustered into four themes – demographic factors, ideological variables, individual differences factors, and intergroup factors. More research is warranted in this domain to further establish a body of evidence on intergroup attitudes towards Aboriginal and Torres Strait Islander peoples, with a renewed focus needed on prejudice reduction techniques.


Historical accounts of prejudice
Aboriginal and Torres Strait Islander peoples 1 are the first people of Australia -a peoples comprising a vast, rich range of linguistic and cultural groups (Common Ground, 2021;Solonec, 2015). Since European settlement, Aboriginal and Torres Strait Islander peoples have faced adversity on their own land. Early evidence of this exists in historical accounts that British settlers were initially given authority to shoot unarmed Aboriginal and/or Torres Strait Islander peoples (Prentis, 2011). While such brutality has reduced over time, Aboriginal and Torres Strait Islander peoples long remained subject to disparities in social justice issues and basic human rights. For instance, there are documented voting inequalities and the forced removal of children, referred to as "the stolen generation", until as recent as 1962, and until 1970 were being paid 50% of the minimum wage. Legislations passed rectifying wage discrepancies in 1984 and "stolen wages" were acknowledged and apologised for by state governments in 2004 (Banks, 2008). While discriminatory legislation towards Aboriginal and Torres Strait Islander peoples have slowly been removed, a number of concerning social trends still exist.
According to the Australian Institute of Health and Welfare report titled "Trends in Injury Deaths, Australia", from 1990 to 1995 Aboriginal and Torres Strait Islander peoples were 16.5 times more likely to die in police custody and at present are 13 times more likely to be arrested compared to non-Indigenous Australians (AIHW, 2019). The report cites a range of stark statistics, including significantly higher rates of homelessness, unemployment, problematic drinking, and gambling among Aboriginal and Torres Strait Islander peoples compared to non-Indigenous Australians. In addition, Aboriginal and Torres Strait Islander peoples have lower completed education rates, concerningly experience 2.7 times greater psychological distress and 2.3 times higher suicide rates than non-Indigenous Australians.
These statistics unambiguously reveal disparity between Aboriginal and Torres Strait Islander peoples and non-Indigenous Australians. Researchers have strived to understand the mechanisms driving this disparity. Social psychology research has converged on the consensus that factors pertaining to the lower status of this group drive much of the negative impacts -specifically, that Aboriginal and Torres Strait Islander peoples are a minority in Australia in terms of social status and power, resulting in increased experiences of stigma, prejudice, and discrimination.

Social determinants of health and minority stress theory
The Meyer (2003) minority stress model posits that minority group members are frequently subjected to unique and excess social stressors, which negatively impact health outcomes. Research supporting minority stress theory largely focusses on differences between mental and physical health outcomes of sexual minority groups (i.e., individuals with nonheterosexual attractions or engaging in nonheterosexual sexual behaviours; Herek & McLemore, 2013). However, the minority stress paradigm has since been extended to ethnic minority groups. For instance, Shangani et al. (2020) found that African American and Latino sexual minority individuals experienced greater self-reported stigma compared to their white counterparts which was linked to significant disparities in mental and physical health. Valentín-Cortés et al. (2020) extended this finding evidencing that undocumented Latinx immigrants to the USA experience heightened stress, leading to adverse mental health outcomes. These results have also been replicated for non-immigrant ethnic minorities (Hayes et al., 2011).
Minority stress theory therefore predicts that Aboriginal and Torres Strait Islander peoples are subject to unique stressors based in prejudice that only they experience (i.e., prejudices that they are targeted for on the basis of their social group, and thus aren't experienced by non-Indigenous Australians), which consequently drives the negative impact on their health outcomes. While minority stress has not been directly researched in the context of Aboriginal and Torres Strait Islander peoples, there is a large body of evidence illustrating their subjection to racism (or negative attitudes based on the race of the attitude-target, Paradies et al., 2015). The study of racism as a social determinant of health (i.e., non-medical factors which influence health outcomes) has been heavily researched, and a metaanalysis of 293 studies synthesising data from over 300,000 participants from the USA found that racism towards ethnic minorities was associated with their poorer mental health across a range of outcomes, including depression, anxiety, and psychological distress as well as poorer general and physical health (Paradies et al., 2015).
The impacts of racism on various First People populations around the world have documented the links between self-reported racism and poor mental health, physical health, and increased substance use (Paradies, 2016). In Australia, a study on Aboriginal and Torres Strait Islander youths revealed that 32% of participants who were aged 16-20 self-reported experiences of racism which was associated with poor social and emotional wellbeing outcomes and mental health diagnoses (Priest, Paradies, Gunthorpe et al., 2011). In a study including slightly older participants (12-26 years), 52.3% reported being subjected to racism which was also correlated with poor mental and general health outcomes (Priest, Paradies, Stewart et al., 2011). Taken together, relevant theoretical models and related evidence suggest that racism is a meaningful social determinant of poor health and given that Aboriginal and Torres Strait Islander peoples report high prevalence levels of racism, it is likely a contributing factor to disproportionate health outcomes. Understanding racial attitudes towards this group is an imperative step in understanding and combating this health disparity.

Understanding attitudes in Australia
Racism towards Aboriginal and Torres Strait Islander peoples can be understood using the ABC tripartite model of attitudes (A for affective, B for behavioural, and C for cognitive; De Montlibert et al., 1961;see also, Eagly & Chaiken, 1998). The affective component, in this case, is the emotional response to Aboriginal and Torres Strait Islander peoples (Van Harreveld et al., 2015). Research has shown that racism towards Aboriginal and Torres Strait Islander peoples is related to a number of affective components. Particularly, negative affect and anger have been associated with modern and old-fashioned forms of racism (Islam & Jahjah, 2001;Leach et al., 2006). The behavioural component captures how attitudes influence discriminatory behaviours towards the target. This has been evidenced by the blatant discrimination of Aboriginal and Torres Strait Islander peoples (see above). Importantly, research evidence reveals that Aboriginal and Torres Strait Islander peoples report ongoing receipt of negative differential treatment and discrimination (Markwick et al., 2019;Ramjan et al., 2016). The cognitive component of racism relates to false beliefs and stereotypes held about the target group. False beliefs and stereotypes held by non-Indigenous Australians about Aboriginal and Torres Strait Islander peoples is prevalent and has been particularly evidenced for false beliefs regarding "special treatment" . These negative beliefs contribute to the detrimental treatment of Aboriginal and Torres Strait Islander peoples (Pedersen & Barlow, 2008;Pedersen et al., 2000).

Measures of prejudice
Measures of prejudice have been forced to evolve with the changing nature of how permissible it is to express prejudice, based on a range of factors, including broad cultural shifts in attitude acceptability, and variables that are specific to the attitude-target's social group and the context in which the attitudes are expressed. In response to the changing nature of attitudes, theorists began to re-conceptualise prejudice, considering two distinct forms -classical or old-fashioned prejudice and modern prejudice (Coenders et al., 2001;Meertens & Pettigrew, 1997). Classical prejudice is blatant, overt and charged with negative racial stereotypes (Pedersen & Walker, 1997). Conversely, modern prejudice is discrete, subtle and often expressed through the denial of discrimination and the belief that minority groups expect more rights than the majority. This distinction lead to the McConahay (1986) development of the modern and old-fashioned racism scales and the subtle and blatant prejudice scales by Pettigrew and Meertens (1995).
Expression of prejudice is increasingly likely to be met with criticism and therefore scales have been developed to detect nuanced differences between modern and old-fashioned prejudices (Anderson, 2018). This distinction has also been made in the exploration of prejudice towards Aboriginal and Torres Strait Islander peoples. Other minority groups observe subtle prejudice to be more prevalent than blatant prejudice (e.g., asylum seeker prejudice, Anderson, 2018; gay men and lesbian women; Morrison & Morrison, 2003). Aboriginal and Torres Strait Islander peoples are similarly subjected to more subtle than blatant prejudice, however blatant racism is still highly prevalent (Barlow et al., 2009). This suggests it is more socially acceptable to express prejudice towards this group and in-turn may be greater impacted by minority stress factors.

Aims and overview
This paper presents the findings of a systematic literature review that explores the attitudes of non-Indigenous Australians towards Aboriginal and Torres Strait Islander peoples. We aim to identify and synthesise research exploring these attitudes and their correlates, with the hope that establishing meaningful patterns of prejudice-relevant correlates might help inform prejudice reduction strategies.

Search strategy
On 15 April 2021 Medline, Psychology and Behavioural Sciences Collection, ProQuest Psychology Database, and PsycINFO databases were searched for published, peer-reviewed articles. The Cochrane methodology guided the development of the protocol and search strategy for this systematic literature review which is presented in lines with relevant sections of the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines (Page et al., 2021). The strategy for the search involved using a series of relevant search terms which were combined using the Boolean search operator "OR" to form the concept of Attitudes and Aboriginal and Torres Strait Islander peoples. The two concepts were then combined using the Boolean search operator "AND". The strategy was applied to each database, and the results of the search were exported as EndNote TM libraries. The libraries were then merged into a single library, where duplicate records were removed (using a combination of automated and manual removal processes). The final library was uploaded to Covidence, which is designed for the management of the screening and extraction phases of a systematic review. No limits were added to the searches. An example search was: (DE "Attitudes") OR (DE "Social Perception") OR (DE "Judgment") OR (DE "Prejudice") OR (Attitude* OR belief* OR opinion* OR "social perception*" OR judg* or prejudice*) AND (Aborigin* OR Koor* OR "First Nation*" OR "First People*" OR "Torres Strait Islander*").
A request for unpublished data was made to the members of three professional bodies (i.e., The Australian Sociological Association, The Society of Australasian Social Psychologists, and the European Association of Social Psychology) as well as through academic avenues on Twitter. No additional data was obtained through these methods.

Screening process
The titles and abstracts of each record were screened for articles in English, containing attitudes towards Aboriginal and Torres Strait Islander peoples. Screening was a two-phase process. In the first phase, the records were double screened with high levels of interrater reliability (Cohen's κ = .96). Articles that appeared relevant based on the content of the title and abstract progressed to the next phase, while those deemed irrelevant were excluded. With any disagreements, both screeners made a joint decision as to whether or not to include the article. In the second phase, the records were downloaded as full-text versions of the article, and assessed based on the below criteria with perfect levels of interrater reliability (Cohen's κ = 1.00). Those meeting all inclusion criteria were included in the final corpus presented in the results section of this paper. Those that did not were excluded.

Inclusion and exclusion criteria
To be eligible for inclusion in the systematic literature review, the articles needed to meet the following criteria: they must a) be available in English, b) contain quantitative original data, c) contain a measure of attitudes or prejudice of non-Indigenous Australians towards Aboriginal and Torres Strait Islander peoples, and d) not have been exposed to an experimental manipulation.

Data extraction and synthesis
Data extracted from studies included the author, year, sample size, participant demographics, measures of attitudes, and a summary of the findings including the direction and strength of the relationships between attitude scored including any prejudicerelevant correlates reported. All extracted articles, with a summary of their extracted information, are presented in Table 1.
To present results in a meaningful way, thematic grouping techniques were used to aggregate similar variables into clusters. This would allow an overview of the higher-order relationships between attitudes and relevant factors (a conceptually similar process to the qualitative analysis techniques to synthesis text-based data proposed by Braun & Clarke, 2006). For example, the correlates of religious affiliation, religious dimensions, political ideology, orientation and policy support are distinct variables, but for the purposes of meaningful synthesis were aggregated to form the cluster of "ideological factors", based on the notion that these are all individual differences factors that are sets of beliefs that are held for reasons that are not purely epistemic (Honderich, 1995).

Study characteristics
The results of the search strategy identified 20 articles (N = 2,958) for inclusion in the systematic review from a total of 1,435 studies (for full details of the screening process, see Figure 1). Each included study was conducted in Australia and almost half (n = 9) utilised a sample comprised solely of students. The corpus had a wide range of publication dates , with five of the articles published prior to the year 2000, and 16 of the studies were published prior to 2010, highlighting the lack of recent research. Nine studies measured modern attitudes, while three measured both modern and old-fashioned prejudice. Table 1 presents the extracted data included in the review. Based on the thematic clustering techniques described above, the range of correlates presented in the articles identified by the search were classified under four subheadings: demographic factors, ideological factors, individual difference factors and  (Fraser & Islam, 1997).

Primary students Preschool Racial Attitude
Measure II Majority group participants rated the in-group more positively than the two out-groups, with the indigenous out-group being rated least positively.
Ethnic minority participants rated the in-group and the ethnic majority outgroup equally positively, while the Aboriginal out-group was rated least positively.
GRD and RWA predicts both forms of prejudice, but modern prejudice more strongly. intergroup factors. The synthesis of the results are presented below according to these clusters.

Demographic factors
There were 12 effects across 7 studies that explored how attitudes towards Aboriginal and Torres Strait Islander peoples related to variables within the demographic factors cluster. Four studies observed the relationship between age and attitudes with conflicting findings. Walker (1994) found attitudes to be negatively correlated with age. Pedersen et al. (2000) found this also to be the case for old-fashioned prejudice but not modern prejudice. In contrast, two recent studies found that age was not related to modern racism towards Aboriginal and Torres Strait Islander peoples (Barlow et al., 2010;Skinner et al., 2013). These results suggest that age may not relate to modern forms of prejudice.
From six studies, a strong pattern emerged that males report more negative attitudes compared to females (Barlow et al., 2009;Hill & Murray, 2020;Pedersen & Walker, 1997;Walker, 1994). However, two studies reported no relationship between gender and attitudes (Heaven & St. Quintin, 2003;Skinner et al., 2013). This could be explained by Skinner et al. (2013)  Five studies reported the relationship between levels of education and attitudes -all showed that higher education is related to decreases in negative attitudes. Pedersen et al. (2000) demonstrated that both modern and old-fashioned racism was negatively correlated with education. Earlier studies by Pedersen and Walker (1997) and Walker (1994) found the same to be true. These findings are supported by more recent studies from Skinner et al. (2013) and Hill and Murray (2020) who found that people with a bachelor's degree or higher held more positive attitudes than those with lower levels of educational attainment.
Of note, the significant correlations between attitudes towards Aboriginal and Torres Strait Islander peoples and demographic variables were relatively weak to moderate in strength.

Ideological factors
There were nine effects across eight studies that explored how negative attitudes towards Aboriginal and Torres Strait Islander peoples related to variables within the ideological factors cluster.
Multiple studies observed the relationship between attitudes and their political orientation and support for political policies. Three studies observed negative attitudes to be positively correlated with socialdominance orientation and right-wing authoritarianism (Feather & McKee, 2008;Pedersen & Walker, 1997;Walker, 1994). Heaven and St. Quintin (2003) observed the same pattern of findings following the experimental induction of either national or personal identity.
Two studies evidenced support for a government apology to Aboriginal and Torres Strait Islander peoples was negatively correlated with modern (Barlow et al., 2010) and symbolic (Leach et al., 2007) racism (using well-accepted beliefs to justify the advantaged position of the dominant culture). Both symbolic and blatant racism were positively correlated with relational orientation, voting one nation, and supporting Pauline Hanson, but negatively correlated with support for acknowledgement of native landowners (Fraser & Islam, 2000;Mabo & Wik).
Two studies observed the relationship of religious factors on attitudes. Religious affiliation to the Church of England held significantly lower attitudes compared to that of Catholics, Christians, and those reporting no religious affiliation (Walker, 1994). Religious dimensions of fundamentalism and intrinsic religiosity were, respectively, negatively and positively related to attitudes (Mavor & Gallois, 2008).
Correlations between attitudes towards Aboriginal and Torres Strait Islander peoples and political ideology varied from moderate to strong. Consistent were the strong correlations observed between attitudes and opposition to apology and support for One Nation and Pauline Hanson. Moderate correlations were observed for religious factors suggesting that overall, ideological factors have a greater relationship with attitudes towards Aboriginal and Torres Strait Islander peoples than demographic factors.

Individual difference factors
There were 10 effects across 8 studies that explored how negative attitudes towards Aboriginal and Torres Strait Islander peoples related to variables within the individual difference factors cluster.
Two studies explored the link between attitudes and values. Those holding a stronger psychological sense of national or global community reported more positive attitudes (Hill & Murray, 2020). Modern racist attitudes were positively correlated with power and security and negatively correlated to universalism and benevolence (Feather & McKee, 2008). No correlations between e2039043-8 attitudes and achievement, hedonism, self-direction, or conformity were observed (Feather & McKee, 2008).
Personality factors of self-transcendence and anger were, respectively, negatively and positively related to symbolic racism towards Aboriginal and Torres Strait Islander peoples (Hill & Murray, 2020;Leach et al., 2007). While Islam and Jahjah (2001) found the best predictor of prejudice towards this group was negative affect.
Personal beliefs about Aboriginal and Torres Strait Islander peoples was shown to directly relate to attitudes towards the group (Islam & Jahjah, 2001). Ecker et al. (2014) found that people used race-related information to justify inferential reasoning when it was consistent with their attitudes. This is also consistent with the findings of Haslam and Wilson (2000) that personal beliefs are predictive of attitudes towards Aboriginal and Torres Strait Islander peoples. Particularly, false beliefs have been shown to relate to negative attitudes (Pedersen & Barlow, 2008).
Overall, individual difference factors shared mostly moderate correlations with attitudes towards Aboriginal and Torres Strait Islander peoples.

Intergroup factors
There were eight effects across eight studies that explored how negative attitudes towards Aboriginal and Torres Strait Islander peoples related to variables within the intergroup factors cluster.
The majority of studies used measures of modern or old-fashioned racism as the measure of attitudes. Three studies compared the prevalence of both, each finding significantly higher levels of modern racism than oldfashioned racism (Hill & Augoustinos, 2001;Pedersen & Walker, 1997;Pedersen et al., 2000).
Three studies researched the attitudes of children and adolescents. Largely, the research focussed on attitudes towards Aboriginal and Torres Strait Islander peoples compared to attitudes towards other minority groups. Haslam and Wilson (2000) found that prejudice existed towards Aboriginal and Torres Strait Islander peoples by adolescents, and that personal beliefs were more predictive of prejudice than the extent of knowledge held about the group. Black-Gutman and Hickson (1996) demonstrated that Euro-Australian children aged 5-12 years held negative attitudes towards Aboriginal and Torres Strait Islander peoples, relative to both their own group and Asian-Australians. Similarly, Griffiths and Nesdale (2006) observed children aged 5-12 years in the majority group of Anglo-Australians rate the in-group more positively than the two outgroups in Pacific Islanders and Aboriginal and Torres Strait Islander peoples, with the latter being rated the least favourably. Minority group children rated the majority group and ingroup equally while out-group Aboriginal and Torres Strait Islander peoples were rated least favourably. In-group and majority group preferences for neighbours were also observed with Aboriginal and Torres Strait Islander peoples being least preferred. In contrast, a study of adults comparing attitudes to other minority groups saw more positive attitudes towards Aboriginal and Torres Strait Islander peoples compared to Asian and Arab populations (Islam & Jahjah, 2001).
One study examined the role of ethnocentrism on attitudes towards Aboriginal and Torres Strait Islander peoples. Participants with high ethnocentricity reported more negative prejudice than participants with low ethnocentricity (O 'Driscoll & Feather, 1985).
Overall, intergroup factors provided an overview of trends in attitudes towards Aboriginal and Torres Strait Islander peoples, highlighting the prevalence of modern racism compare to old-fashioned racism and detailing the negative views of non-Indigenous children.

Summary of evidence
This review aimed to identify and document non-Indigenous Australian attitudes towards Aboriginal and Torres Strait Islander peoples and determine the correlates of these attitudes. The systematic search identified 20 relevant studies (N total = 2,958) which met our eligibility criteria. Four clusters of variables emerged from this review. The data were not reported sufficiently to allow a statistical synthesis of the effect sizes, but qualitatively it seems that ideological and individual difference factors most strongly relate to attitudes towards Aboriginal and Torres Strait Islander peoples, followed by intergroup factors. Demographic factors produced the least consistent patterns of findings. In addition, the literature clearly outlined that modern racism is more prevalent than old-fashioned racism. This finding is consistent with self-reported experiences of racism by Aboriginal and Torres Strait Islander peoples (Markwick et al., 2019), and with literatures exploring other attitude objects (e.g., Muslims: Anderson & Antalikova, 2014; asylum seekers: Anderson, 2018).
Together, there were relatively stable patterns of relationships between attitudes and variables in the demographic cluster. The review indicated that age was negatively related to attitudes towards Aboriginal and Torres Strait Islander peoples for data collected prior to 2000 with adult samplesparticularly for old-fashioned prejudice and not modern prejudice -although this relationship seems to be dissipating with time since more recent studies identified no relationship between age and attitudes. The review showed that higher education levels related to less prejudice towards Aboriginal and Torres Strait Islander people, and overwhelmingly showed that non-Indigenous males report more negative attitudes than non-Indigenous females. These findings are relatively unsurprising, since these factors are known to correlate with other forms of prejudice in Australia (e.g., Cowling et al., 2019;Paradies et al., 2015).
Each study including a measure of political ideology yielded significant correlations with attitudes towards Aboriginal and Torres Strait Islander peoples. Attitudes were negatively correlated with social dominance orientation and right-wing authoritarianism, as expected based on a heavy body of literature documenting that these variables correlate with general and specific prejudices (and a theoretical postulation explaining why this combination of variables strongly predicts prejudice, see, Duckitt & Sibley, 2010 dualprocess model of prejudice). Modern and symbolic racism were negatively related to support for apology and acknowledgement of native landowners, and positively to voting for right-wing political party one nation and support for Pauline Hanson whose views centre on the notion that Aboriginal and Torres Strait Islander peoples received "special treatment" that nonindigenous Australians are not able to access. This also aligns with evidence that negative attitudes towards other groups predicts political policy support (e.g., Hartley et al., 2019 shows that negative attitudes towards refugees predicts support for punitive policies about their treatment).
Unsurprisingly, the studies observing religion and attitudes, consistently revealed that religion was related to attitudes regardless of how religion was operationalised (i.e., as a categorical affiliation or as a continuous measure of religiosity, Anderson, 2015[also Deslandes & Anderson, 2018] for a discussion). As expected, those with a strong sense of national and global community exhibited more positive attitudes. Suggesting that when people are connected to their community, they hold more positive attitudes to alternative cultures.
Negative affect was shown to be a significant predictor of prejudice while self-transcendence and anger were, respectively, negatively and positively related to attitudes towards Aboriginal and Torres Strait Islander peoples. Personal beliefs were also predictive of attitudes towards Aboriginal and Torres Strait Islander peoples and it was evidenced that people are likely to use personal beliefs to support inferences made regarding Aboriginal and Torres Strait Islander peoples. Such false beliefs were shown to relate to negative attitudes suggesting education informing a better understanding is required to address negative attitudes held.

Addressing negative attitudes
Means of addressing prejudice towards Aboriginal and Torres Strait Islander peoples has been researched at individual and systemic levels. Effective means of reducing individual-level racism provide accurate information regarding the nature of racism and highlight personal accountability (Chapman et al., 2014;Paradies et al., 2015). Turoy-Smith et al. (2013) demonstrated that increased quality of contact decreased prejudice towards Aboriginal and Torres Strait Islander peoples, which increased support for legislation and willingness to act. Effective means of addressing systemic racism and reducing prejudice improves awareness of the nature of racism and educates on negative attitudes and their impacts (Durey, 2010;Goold Oam & Usher, 2006;Pedersen & Barlow, 2008). This literature review has outlined the changes in prejudice towards Aboriginal and Torres Strait Islander peoples and should be used to inform prejudice reduction strategies that are driven by socialpsychological theories around intergroup contact, and psychoeducation.

Limitations of the literature
There were several constraints to this literature, based on limitations in the individual studies that were eligible for synthesis. There were issues with a limited amount of available evidence to synthesise. For instance, only three studies assessed the attitudes of children/adolescents towards Aboriginal and Torres Strait Islander peoples (Black-Gutman & Hickson, 1996;Griffiths & Nesdale, 2006;Haslam & Wilson, 2000). Nearly half of the studies (n = 9) contained samples comprised entirely of university students, limiting the generalisability of the findings. The generalisability of the findings is limited by disproportionate sample characteristics. For instance, of the 17 studies reporting gender, 14 had more female participants than male -moreover, 8 of those studies contained greater than 70% female participants. As such, reflections on gender differences are less feasible.
In the case of the demographic factors cluster, there were missing data that typically are recorded and reported in the prejudice literature. For instance, three studies did not provide gender composition statistics (Haslam & Wilson, 2000;Leach et al., 2007;Pedersen et al., 2000) and three studies did not record age demographics (Fraser & Islam, 2000;Heaven & St. Quintin, 2003;Pedersen et al., 2000), and moreover a further 6 either only provided a range or a mean age (with no index of the variability in age). There were some issues that limited the internal validity of this literature. For instance, three studies failed to use validated measures of attitudes (Islam & Jahjah, 2001;O'Driscoll & Feather, 1985;Pedersen & Barlow, 2008), and 16 of the studies predate 2010 (and use outdated terminology)..

Limitations of the current review
There are also limitations to the systematic review presented in this paper. The main limitation is the paucity of relevant articles assessing attitudes towards Aboriginal and Torres Strait Islander peoples. Of the limited articles, few reported the same correlates of attitudes, making interpretation of results difficult. Many articles poorly reported their findings, frequently omitting effect sizes and confidence internals. This ruled out any possibility of a quantitative meta-synthesis of the findings, and disqualified interpretations of the size or magnitude of the reported effects. Finally, there is no possibility of drawing inferences of causality since all extracted data was correlational.

Suggestions for future research
The current review highlights the growing prevalence of modern racism, while simultaneously making salient the paucity of evidence that is available on this topic. This combination is problematic. In addition, none of the 20 studies used a measure of implicit attitudes (i.e., measures of non-conscious bias), which is problematic since explicit attitudes (assessed with self-report measures) are known to correlate with socially desirable responding concerns (Anderson, 2019) and have been argued by theorists as providing only part of the "picture" of social attitudes (Nosek, 2007). Taken together, the findings of this review suggest that there is a need to focus on: (a) an increase in the quantity of available evidence on attitudes towards Aboriginal and Torres Strait Islander peoples, (b) an increase in the range of correlates that are assessed, (c) evidence from measures that are not reliant on self-report, and (d) a focus on evidence for the efficacy of prejudice intervention strategies.

Conclusion
This paper presents the first systematic review and synthesis of the available evidence of attitudes (and their correlates) of non-Indigenous Australians towards Aboriginal and Torres Strait Islander peoples and highlighted attitude themes to which they have been and are currently subjected. Of most significance were the findings that Aboriginal and Torres Strait Islander peoples are more subjected to modern forms of prejudice than old-fashioned, males are more likely to demonstrate negative attitudes towards this group than females, and that children held also negative evaluations. This review should assist and inform government and education prejudice reduction strategy and psycho-education efforts. Note 1. A range of terms exist to describe the first peoples of Australia. "First Australians", "First Nations People", and Indigenous Australians, are often used, and each have proponents of the term, but also they're criticised for many reasons, not the least of which is that it infers a homogenous group. Given that they group is varied in languages and culture, the consensus from the first peoples themselves is that the preferred term is Aboriginal and Torres Strait Islander peoples (Common Ground, 2021;Solonec, 2015). Finally, we use the plural "peoples" to acknowledge the distinctiveness and multiplicity of the many groups of people who fall under this umbrella term.

Disclosure statement
The authors would like to disclose their positionality and identities as White individuals residing in Australia, a land that was colonised by the British Empire. We acknowledge that this occurred at great cost to Australia's First Nations people, and recognise our position as settlers on this stolen land.

Data availability statement
Data sharing is not applicable to this article as no new data were created or analysed in this study.