Mentoring programs for Indigenous youth at risk

There is a range of risk factors that may make young people of any ethnicity more likely to engage in antisocial behaviours. These factors include the young person’s own attitudes; relationships within the family; and growing up in communities where there is widespread violence, alcohol and other substance abuse, poverty, poor health and poor-quality housing. Indigenous young people face the additional challenges ofdispossession, discontinuity of culture and intergenerational trauma.A strong connection to culture—coupled with high self-esteem, a strong sense of autonomy, and with living in cohesive, functioning families and communities—can be protective factors that result in Indigenous young people choosing productive life pathways.Mentoring is a relationship intervention strategy that can assist in building some of these protective factors. A growing body of research demonstrates that mentoring can have powerful and lasting positive effects in improving behavioural, academic and vocational outcomes for at-risk youth and, to a more limited extent, in reducing contact with juvenile justice systems.In an Indigenous context, mentoring is a particularly promising initiative because it fits well with Indigenous teaching and learning styles and can help to build strong collective ties within a community.Mentoring programs can involve adult or peer mentors and can be implemented in a range of ways, such as one-on-one or in groups.Although positive results can be achieved with single-intervention mentoring for at risk youth, integrating mentoring into broader programs produces a greater level of positive change.The way the mentoring program is run and the nature of the relationship between mentor and mentee are crucial in determining the outcomes of youth mentoring programs.

• A strong connection to culture-coupled with high self-esteem, a strong sense of autonomy, and with living in cohesive, functioning families and communities-can be protective factors that result in Indigenous young people choosing productive life pathways.
• Mentoring is a relationship intervention strategy that can assist in building some of these protective factors.
A growing body of research demonstrates that mentoring can have powerful and lasting positive effects in improving behavioural, academic and vocational outcomes for at-risk youth and, to a more limited extent, in reducing contact with juvenile justice systems.
• In an Indigenous context, mentoring is a particularly promising initiative because it fits well with Indigenous teaching and learning styles and can help to build strong collective ties within a community.
• Mentoring programs can involve adult or peer mentors and can be implemented in a range of ways, such as one-on-one or in groups.
• Although positive results can be achieved with single-intervention mentoring for at risk youth, integrating mentoring into broader programs produces a greater level of positive change.
• The way the mentoring program is run and the nature of the relationship between mentor and mentee are crucial in determining the outcomes of youth mentoring programs.

What works
At the level of program design and implementation, what works includes the following: • Starting mentoring before Indigenous young people exhibit antisocial or criminal behaviour.This is the most effective strategy, although starting mentoring after they have begun engaging in these behaviours can still achieve positive results.
• Obtaining the input of the local Indigenous community in the design and delivery of mentoring and related programs.
• Involving Elders where possible in transmitting cultural knowledge to young people through a mentoring relationship.
• Having strong partnerships between the agency running the mentoring program and other youth counselling, health and employment services in the local area.
Within the mentoring relationship, what works includes the following: • Long-term mentoring relationships of at least 12-18 months duration, based on common interests, mutual respect, genuine friendship, fun and a non-judgemental approach.
• Mentoring that continues to support the young person as they consolidate positive changes.
• Consistent, regular contact between mentor and mentee.In the initial stages, this may need to be quite intensive (up to 10-20 hours per week), depending on the young person's needs.
• Involvement of Indigenous parents in the mentoring relationship, which can improve parent-child relationships.
• Mentors who have 'been there, done that'.Mentors who have experienced similar challenges to those facing the mentee and proven their success in overcoming negative life circumstances are the most influential in achieving positive behavioural change.

What doesn't work
• Short-term mentoring (generally 6 months or less) is not generally effective.
• Mentoring is unlikely to produce any change where there is infrequent or irregular contact between mentor and mentee, where the mentor is authoritarian or judgemental, or where there is too much emphasis on expected behavioural change rather than first building a friendship.
• Too many goals can cause the mentee to become discouraged and give up.
• Mentoring by peers is not an effective replacement for mentoring by adults.

What we don't know
• The relative effectiveness of strategies such as gender-specific mentoring, matching mentor and mentee along racial or ethnic lines, and using paid versus volunteer mentors are debated in the literature and require further exploration.
• Further longitudinal research is required into whether the benefits of youth mentoring are maintained into adulthood.
• Although mentoring is repeatedly demonstrated to improve a range of outcomes for Indigenous young people that indirectly help to reduce crime, a direct link to crime reduction is not clear in the studies reviewed here.

Introduction
This Resource Sheet examines evidence for the effectiveness of mentoring programs in helping to set Indigenous young people at risk of engaging in antisocial and risky behaviours on healthier life pathways.
Mentoring is a relationship intervention strategy that research is showing can have powerful and lasting positive impacts on behavioural, academic and vocational outcomes for at-risk youth.Costello and Thomson (2011:1) describe youth mentoring as follows: Youth mentoring is, according to the Australian Youth Mentoring Network, defined as 'a structured and trusting relationship that brings young people together with caring individuals who offer guidance, support and encouragement'.The goal of youth mentoring is to enhance social engagement and thereby minimise negative behaviours through growth in social and developmental behaviours.
There are two types of mentoring style found in the literature-natural and planned.Among Indigenous Australians, the natural or informal form of mentoring is often spontaneous through the Elders' traditional role of sharing the wisdom, the knowledge and the spirit, which can draw Aboriginal people back to traditional ways.Elders play an extremely important role in Aboriginal families as role models, care providers and educators (Walker 1993).
This Resource Sheet focuses on the planned or formal form of mentoring, which often includes Elders as part of these programs.It does not, however, cover the following formal forms of mentoring: • a detailed analysis of mentoring, which occurs within sporting and other programs.(This is covered, where relevant, in a forthcoming Resource Sheet titled Supporting healthy communities through sports and recreation programs.) • mentoring embedded within broader youth diversionary or justice programs • mentoring within cadetship or other vocational education programs.
There is a strong body of literature on the types of youth mentoring programs and the dynamics of successful programs and mentoring relationships.This Resource Sheet draws on evidence from 45 studies.Over half were Australian studies, with additional evidence from research in other colonised nations such as New Zealand, Canada and the United States.Two-thirds of the studies were Indigenous-specific.A range of methodologies was used including evaluations, critical descriptions of programs, meta-analyses and research syntheses.

Risk and protective factors
There is a range of factors that increase the risk of young people engaging in negative behaviours, although other factors appear to reduce this risk.The literature suggests these risk and protective factors are relevant whether young people live in discrete remote communities or in larger urban centres.

Risk factors
Factors that may place young people at risk of disengaging from school, employment and community are often multiple and complex in nature (Blechman 1992) and include: • a dysfunctional social context, where young people are exposed to high rates of violence, substance abuse and welfare dependency, and to a strong peer culture of negative behaviours, as well as recording low levels of employment and educational achievement (Allard 2010;Allard et al. 2007;Costello & Thomson 2011;Eversole et al. 2004;HRSCATSIA 2011) • a negative family situation, where there is family breakdown, violence, neglect or psychological distress (Delfabbro & Day 2003;Eversole et al. 2004).Growing up in a sole-parent family has also been linked with a higher risk of antisocial behaviour where other known risk factors are present, such as poverty and unemployment (Allard 2010) • poor health, which can limit life choices (Allard et al. 2007;Dawes 2011) • low quality or crowded housing, or homelessness (Allard et al. 2010) • being male-male youth, statistically, tend to be more prone to risky behaviours (Allard 2010) • a lack of structured activities in the local community leading to boredom (HRSCATSIA 2011).The Closing the Gap Clearinghouse has prepared two additional Resource Sheets titled Supporting healthy communities through arts programs and Supporting healthy communities through sports and recreation programs that address a range of effective programs that help to alleviate boredom and build inclusive communities.
In addition to these general risk factors, Indigenous young people can face the ramifications of the dislocation resulting from colonisation, along with the consequences of the Stolen Generations.The effects of forced removal from homelands, of past generations of children being removed from their parents, as well as the economic and social dispossession that resulted from colonisation, have left a legacy of: • intergenerational trauma (Atkinson 2002;Colquhoun & Dockery 2012;Hollis et al. 2011) • people raised in institutions or abusive settings who now lack the skills as adults to parent their own children in a nurturing manner (Allard et al. 2010;Crooks et al. 2009a;Delfabbro & Day 2003) • parents who were raised in abusive settings now abusing their children (Allard 2010;Crooks et al. 2009a;Delfabbro & Day 2003) • dislocation from kinship networks, leaving families with minimal supports (Allard et al. 2007;Costello & Thomson 2011).
Although not all Indigenous communities and people struggle with the risk factors described above, statistics on community safety and crime continue to demonstrate a range of negative outcomes for many Indigenous young people growing up in negative environments.High levels of substance abuse and physical violence, poor-quality housing, poor health outcomes, and low levels of educational attainment and employment continue to plague some communities, adversely affecting younger residents and leading many to engage in risky and antisocial behaviour (AIHW 2011;Stacey 2004:3).As a result, Indigenous young people experience far higher rates of contact with the juvenile justice system (AIHW 2011).
In the case of Indigenous offenders leaving detention, when compared to non-Indigenous offenders, they also face additional challenges, including: alienation in the predominantly non-Indigenous Australian justice system, the relatively young age of Indigenous offenders, the isolation of young people due to their geographical separation from family and cultural group, lower levels of functional literacy, specific health needs, the high profile of Indigenous individuals upon release, the social status of offending among peers, and the difficult circumstances encountered upon release (Allard et al. 2007:vi).For further information on reintegrating recently released offenders, particularly in terms of employment, see <www.aihw.gov.au/closingthegap/documents/resource_sheets/ctgc-rs11.pdf>.

Protective factors
The literature suggests that having a range of protective factors in place supports individuals' healthy development by providing a buffer against risk factors (Barwick 2004:11;Allard et al 2007).At least one study also indicates that having protective factors in place supports Indigenous people to improve their lives (Crooks et al. 2009b).
For individuals, protective factors can include the combination of strong cultural identity and high self-esteem, a strong sense of autonomy and subjective wellbeing (Colquhoun & Dockery 2012;Tomyn et al. 2013).
At a community level, the literature suggests that where an Indigenous community has been able to maintain links with cultural practices and knowledge-and where the community is cohesive-there is a much lower risk of substance abuse in general, emotional or behavioural difficulties among children in particular, and of suicide or suicide ideation among young people (Colquhoun & Dockery 2012).However, where these protective factors are not present, young people and their communities can struggle to maintain some quality of life.These struggles contribute to trauma, poverty, poor health and social exclusion (Colquhoun & Dockery 2012).

Types of formal mentoring programs
There are several types of formal mentoring programs (Farruggia et al. 2010:8).These are described below.
• One-to-one mentoring: where one mentor works with one mentee.This is the most common form of mentoring (see also Blechman 1992).
• Group mentoring: where one mentor works with a group of mentees.This is presented in mentoring literature from Canada and the United States as being a highly culturally appropriate form of mentoring due to the communal nature of those cultures in these two countries (Banister & Begoray 2006;Klinck et al. 2005).
Although the Australian Indigenous literature is largely silent on this point, it is possible that group mentoring would be effective in the Australian context as Indigenous cultures here also tend to be communal in nature.
• A mixture of one-on-one and group mentoring.
• Team mentoring: where a team of mentors works with a group of young people, typically in ratios of no more than 4 young people per 1 adult.
• Peer mentoring: where a young person mentors someone of a similar age.This is thought to be effective only in a limited set of circumstances (discussed later in this paper under the heading 'What doesn't work').
Mentoring can be conducted as a single intervention, or as one component of a broader program for at-risk youth (Barwick 2004;CaFCA 2010).Although either approach can produce positive changes, research suggests that integrating mentoring into broader programs produces a greater level of positive change (Barwick 2004;Farruggia et al. 2010;Hollis et al. 2011;Klinck et al. 2005).
Other activities and resources that enhance the mentoring experience include: • adventure camps and cultural walks, particularly where they reconnect Indigenous young people with traditional practices and knowledge (Barwick 2004;CaFCA 2010;Hollis et al. 2011).See also <http://www.nalag.org.au/pubs/Finding_a_Path_to_Healing_Paper.pdf> for information about healing camps.
• community projects, such as maintaining a community swimming pool or constructing an equipment storage shed at the local sportsground (Barwick 2004;Hollis et al. 2011).These initiatives help to build a stronger sense of belonging for Indigenous young people and provide an opportunity for reciprocity, which is an important social skill in Indigenous communal cultures (Joseph 2007;Sinclair & Pooyak 2007) • referral pathways to ensure Indigenous young people are fully supported; for example, for services such as counselling, health, employment and education (Stacey 2004).
These activities and resources can provide additional opportunities to connect with significant adult role models and chances to learn and practice new positive behaviours, as well as to reinforce new ideas and behaviours learned through the mentoring process (Barwick 2004;Farruggia et al. 2010;Hollis et al. 2011;Klinck et al. 2005).
Mentoring programs may or may not target a specific gender or cultural group.In the context of Indigenous youth mentoring, some studies cite a range of benefits to culturally targeted programs-such as reconnecting young people with their culture, imbibing a greater sense of respect for Elders (Stacey 2004) and building cultural pride (Higgins 2005).However, it should be noted that other studies suggest that targeting one culture can contribute to ongoing stigmatisation of that group by others (Farruggia et al. 2011).

How does mentoring produce positive changes?
Mentoring is typically viewed as a primary prevention strategy that contributes to improvements in affective or instrumental domains (Blechman 1992;Costello & Thomson 2011;Pawson 2004;Spencer & Liang 2009) through reducing risk factors and building a protective and empowering relationship (Barron-McKeagney et al. 2000).
• The affective or psychosocial domain includes emotional support, encouragement, advocating change, and developing goals/aspirations.
• The instrumental domain includes re-engagement with school, increased academic achievement, development of vocational skills, improved health (or reduction of risky behaviours, particularly in the sexual health sphere), reducing or preventing antisocial and criminal behaviours, and re-integrating released offenders back into the community.
• Changes in risk and protective factors are facilitated by providing young people with a positive role model to observe and emulate (Barwick 2004;O'Brien et al. 2009) and by breaking the links between risk factors and spiralling antisocial behaviour by providing space to think, reflect and imagine positive alternatives (Spencer & Liang 2009).
Pawson's research synthesis on mentoring describes four basic mechanisms through which mentoring helps bring about positive change (Pawson 2004:7): 1. Affective contacts: emotional support, friendship and helping the mentee to 'feel differently about themselves'.This appears to be the most common role played by mentors.
2. Direction setting: advice and guidance as the mentee sets new directions and navigates their way 'through the difficult choices confronting [them]'.
3. Coaching: building the mentee's aptitude for participating fully in society; 'encouraging, pushing and coaxing their protégés [mentees] into practical gains, skills and qualifications'.
4. Advocacy: advocating on behalf of the mentee, 'grab[bing] the mentees' hands, introducing them to this network, sponsoring them in that opportunity, using the institutional wherewithal at their disposal'.
Pawson conceptualises these mechanisms as levels of relationship because: Much more than in any other type of social programme, interpersonal relationships between stakeholders embody the intervention.They are the resource that is intended to bring about change (Pawson 2004:7 (emphasis is author's own)).
The process of effecting positive change in at-risk youth is summarised in the following diagram:

Is mentoring effective for moving at-risk Indigenous youth towards healthier pathways?
A growing body of research-both international and Australian research-demonstrates a range of benefits from mentoring for at-risk youth,-from improved self-esteem and a sense of hope through to increased engagement with school, family and community and reduced participation in risky or criminal behaviours (Farruggia et al. 2010;Hollis et al. 2011;MacCallum et al. 2005).Benefits flow to schools and the wider community in the form of increased school attendance and achievement, reduced crime and antisocial behaviour, and increased engagement with community organisations and projects (Farruggia et al. 2010;MacCallum et al. 2005).
Mentoring fits particularly well with Indigenous learning and teaching styles, and its cultural appropriateness with Indigenous cultures has been repeatedly demonstrated in Australia, as well as in New Zealand, the United States and Canada (Banister & Begoray 2006;Barwick 2004 • involvment of parents in the mentoring relationship, by building a relationship between parents and the mentor, and by offering parents access to a range of other complementary programs (Allard et al. 2007;Barron-McKeagney et al. 2000;Delfabbro & Day 2003;Farruggia et al. 2010;Klinck et al. 2005).
Both a recent research synthesis report and a smaller evaluation of a non-Indigenous Melbourne youth mentoring program demonstrate that mentoring is a highly cost-effective preventative strategy when compared to the costs of addressing antisocial or criminal behaviours, substance abuse and ongoing violence later in the lives of at-risk young people (Costello & Thomson 2011;Moodie & Fisher 2009).

Principles for effective mentoring of at-risk Indigenous youth
A large body of research on youth mentoring, including several research syntheses and systematic reviews, suggests a range of principles for implementing effective youth mentoring programs.Some are effective with at-risk youth of any ethnicity, although others are particularly salient for working with Indigenous young people.
Several general principles for implementing mentoring programs emerged from the literature.These are described below: • Work with young people who are at high risk of engaging in risky behaviours: Youth mentoring produces the most positive results with those in high-risk categories; that is, those experiencing multiple risk factors for antisocial behaviours (Allard et al. 2007;Farruggia et al. 2011).
• Early intervention is crucial: Mentoring programs for at-risk Indigenous youth should commence when they are aged 10-14, or even 8-10 (Allard 2010; Banister & Begoray 2006;Delfabbro & Day 2003;Stacey 2004).Given that Indigenous young people who engage with the juvenile justice system tend to be younger than non-Indigenous young people (AIHW 2011), early intervention is crucially important.The literature also suggests that mentoring should commence before anti-social behaviour starts (Costello & Thomson 2011).
• Programs should be grounded in theory and empirical evidence and evaluated regularly: The studies clearly demonstrate that programs grounded in theory and empirical evidence are the most effective (Allard et al. 2007;Costello & Thomson 2011;Farruggia et al. 2011;White 2007).Likewise, established programs that integrate regular, independent evaluation also appear to achieve more positive outcomes in young people's lives (Farruggia et al. 2011).However, studies also suggest that the local Indigenous community and Elders should be involved in the design, implementation and evaluation processes to ensure mentoring is relevant to the needs of Indigenous young people and to the cultural context in which it is being used (Fogarty & Schwab 2012;Joseph 2007;Klinck et al. 2005;Schwab 2006;Sinclair & Pooyak 2007).
Effective programs include the following elements: • Embedding of mentoring in broader programs and provision of structured activities: The need for structured mentoring along with other activities as part of a broader program is a consistent theme in the mentoring literature (Allard et al. 2007;Costello & Thomson 2011;Delfabbro & Day 2003;Farruggia et al. 2010Farruggia et al. , 2011;;Pawson 2004).
• Strong partnerships with other service agencies, with mentoring program targeted to a specific local area: Partnerships with other agencies allow for more efficient referral of at-risk youth to the mentoring program, as well as for the ready referral of mentees to other services as needed (CaFCA 2010; Crooks et al. 2009a;Delaney & Milne 2002;Richards et al. 2011).
• Provision of a wide range of activities and options for engaging young people: Young people have a variety of interests, so where possible, structured activities should provide the broadest possible set of options to maintain their engagement (Barwick 2004).This is important within the times that mentor and mentee spend together, and as an adjunct to mentoring.Adjunct program activities that enhance the outcomes of mentoring relationships and processes with Indigenous young people include adventure camps, fishing, hunting for bush tucker, yarning and instruction in traditional cultural knowledge and practices (CaFCA 2009(CaFCA , 2010;;Schwab 2006).
• Involvement of community members in design and implementation of program: Where local Indigenous community members have genuine input into tailoring the program to the local context, and where they have opportunities to participate as mentors, ownership of program processes and outcomes is more likely.This translates into not only more young people (and often, more adults) participating as mentees but also local cultural protocols being correctly applied, such as how program staff interact with the local Indigenous community and how young people interact with more senior community members (CaFCA 2009; Klinck et al. 2005;Schwab 2006).
• Involvement of Elders in mentoring: Where possible, the involvement of local Elders as mentors and in adjunct activities enhances the cultural connections of the young people and has been demonstrated to improve the level of respectful relationships with local community leaders (Schwab 2006).
• Mentors who have 'been there, done that': Mentors who have faced similar complex life challenges and succeeded in breaking out and living fulfilling lives are the most influential in achieving positive behavioural change (Pawson 2004).This is particularly important for Indigenous youth given the additional intergenerational challenges of dispossession and cultural discontinuity their families may have faced.
• Effective recruitment and screening of mentors: Mentors need to be carefully screened to protect vulnerable children from potential exploitation.At the same time, screening processes for Indigenous mentors need to acknowledge that, due to ongoing effects of Indigenous history, some potentially effective mentors may not fit conventional screening criteria.Indigenous Australians who have recovered from addictions or discontinued former criminal activities may have much to share with at-risk Indigenous youth and add to the number of available mentors (Klinck et al. 2005:120).
• Paid agency staff member to coordinate the program and support mentors: A paid coordinator can play a pivotal role in building a successful mentoring program (Barron-McKeagney et al. 2000;Stacey 2004).Such a coordinator can support mentors and parents of youth, follow up with parents any issues that may require parental action, and promote the program to other agencies and the broader community.
• Qualities of mentoring program coordinator: Qualities such as a calm demeanour, a long-term commitment to the local community and living an exemplary life as a role model are important if the coordinator is to have meaningful influence in the lives of mentees, their parents and mentors (Barron-McKeagney et al. 2000;Stacey 2004).The same studies just cited also showed that coordinators who were respected members of the community had greater influence.
• Ongoing training and support for mentors: This is important in ensuring that mentors have the necessary skills and, at times, the resilience to stay the course.Mentoring is rarely a linear process, with setbacks and relationship breakdowns common (Pawson 2004).A mentor needs support to rebuild the relationship where it breaks down and to deal with the complex challenges faced by at-risk youth (Pawson 2004).
Where non-Indigenous mentors work with Indigenous young people, locally-specific cultural training is important (Schwab 2006;Sinclair & Pooyak 2007).Following training, mentors should have the option to opt out.This approach represents a valuable way of screening out those who may not be the most appropriate mentors (CaFCA 2010).A respected member of the local community was employed as a community coordinator to set up and coordinate the recruitment, training and support of mentors, as well as to implement the mentoring program.
The coordinator spent several months promoting the program in local community settings (for example, church meetings) to build local ownership, to reassure parents that the program would be beneficial, and to recruit local community members as mentors.Mentors were members of the Chicano community who had successfully established themselves in careers as well as in broader society.Training for mentors was adapted from a range of different mentoring programs to suit the local context, and mentors were provided with ongoing supervision by the program coordinator.
Strong partnerships were also formed with other local community service agencies, such as a resource and information centre for Chicanos.This assisted with referrals of young people into the mentoring program, as well as referrals of mentees to services outside the mentoring program as needed.
Most mentoring relationships continued for at least 12 months and mentoring activities were combined with a range of other activities that inner-city children might not otherwise have had the opportunity to experience (for example, riding in an aeroplane, or visiting a farm).These activities were intended to provide structured opportunities for the mentor and mentee to have fun together, as well as to learn new skills.
In addition to overseeing volunteer mentors working with young people, the coordinator ran both educational and social programs for the mentees' parents, which provided opportunities to support them and build a range of skills.
Through their participation '... [parents] found support, gained resources, expanded their commitment to parenting, increased involvement in community life, and gained a sense of empowerment from seeing the results of their personal efforts' (p.52).
The program evaluation indicated that mentors, mentees and their parents almost unanimously felt that the program was beneficial and enjoyable.Outcomes include 'gains in social skills and decreases in problem behaviours' among mentees (p.54), as well as increased resources and support for parents (p.52).

Characteristics of effective mentoring relationships
The characteristics of effective mentoring relationships are described below: • High-quality relationships are of central importance: The literature is rich in descriptions of high-quality mentoring relationships.It is important that relationships are non-judgmental, affirming, empowering, inspiring; and built upon trust, mutual respect and dialogue (Barwick 2004;Crooks et al. 2009a;Lemmon 2005;MacCallum et al. 2005).A large number of studies also talk about the importance of having fun together, enjoying each other's company, and of settings goals that become progressively more challenging (that is, to allow 'graded mastery' of new challenges) (Barwick 2004;Hollis et al. 2011).Marlene Burchill talked about the importance in an Indigenous context of 'yarning up not down' (Burchill 2004, cited in Higgins 2005:8)that is, talking together about solutions to challenges, rather than coming in with outside solutions.Finally, it is important that mentors have realistic but high expectations for what at-risk young people can achieve (Barwick 2004;Higgins 2005;Hollis et al. 2011).
• Long-term relationships are maintained: Perhaps the strongest finding from research studies is the importance of committing to and forming long-term mentoring relationships (Barwick 2004;Costello & Thomson 2011;Delfabbro & Day 2003;Farruggia et al. 2010Farruggia et al. , 2011;;Moodie & Fisher 2009;Pawson 2004;Schwab 2006;Sinclair & Pooyak 2007;Spencer & Liang 2009).Meaningful contact needs to be maintained for at least 12-18 months, with effectiveness and influence increasing the longer the relationship is maintained (Costello & Thomson 2011;Farruggia et al. 2010).Evidence suggests that short-term mentoring relationships of less than 6 months may actually disadvantage young people, as they tend to reinforce a sense of loss and disappointment already experienced by the mentee so many times in their relationships with adults (Dawes & Dawes 2005).Studies also demonstrated the importance of staying with the mentee after the 'at-risk' period, and continuing to support them through a phase where positive changes are consolidated (Pawson 2004;Stacey 2004).Contact needs to be regular, consistent and sufficiently frequent to form a strong relationship (Barron-McKeagney et al. 2000;Delfabbro & Day 2003;Stacey 2004).In the early stages of the relationship with those who are already engaging in risky or criminal behaviours, contact may need to be intense (20 hours or more per week).Where this is the case, paid mentors with a low caseload are crucial for managing the intensity of contact (Stacey 2004).
• Parents are involved in the mentoring relationship and additional support for parents is provided: This was another particularly strong finding across multiple studies (Barron-McKeagney et al. 2000;Delfabbro & Day 2003;Stacey 2004).Program evaluations repeatedly demonstrated that where parents were involved in the mentoring process, and where they were provided with additional mentoring or training (for example, in parenting skills), parent-child relationships improved considerably.Additional training and support for the parents also appeared to significantly reduce the stress levels in their lives and improve daily functioning (Barron-McKeagney et al. 2000;Stacey 2004).
• The focus is on a small number of goals: Where the mentee has a large number of goals, the challenges may be too great and the mentee may therefore disengage from the mentoring process.Several international Indigenous-specific programs found that setting a smaller number of goals with the mentee resulted in more sustainable change (Farruggia et al. 2010).Smaller goals can later be built upon, as larger issues are tackled.

Box 2: Promising practice: mentoring program in Adelaide, South Australia
The Panyappi Indigenous Youth Mentoring Program was set up in Adelaide in 2001 in response to growing numbers of at-risk Indigenous young people congregating in the inner city and engaging in antisocial behaviours.Panyappi focused on specific geographical areas in order to build strong partnerships with other local service agencies.
Mentors were provided with formal training and informal supervision.Many gained formal qualifications in youth work as a result.
The program targeted young people aged 10-15.These young people had a history of poor school attendance and educational achievement, substance abuse, unstable living environments and experiences of abuse.Some had already begun engaging in offending behaviours.
The auspicing agency for the program employed full-time mentors who had a low caseload and, initially, some mentors had responsibility for only one young person.This allowed them to engage with the young person intensively, building trust within a relationship that was formalised but voluntary.Other programs were also offered by the auspicing agency, including a genealogy program to help these young people and their families to reconnect with their heritage.
Being an Indigenous-specific program, attention was given to cultural appropriateness.One specific outworking of this was that the program coordinator also engaged with families.Stacey ( 2004) suggests that: Rebuilding and strengthening family connections is an important aspect of culturally appropriate practice in mentoring-this contrasts with mainstream mentoring programs that do not engage with families as standard practice (p.4).
The staff often provided informal mentoring of parents, linking them to a range of services to help address the results of historical abuses and ensure support was available to address the complex needs of the young person and the family.This additional support to the parents as well as to their young people contributed to decreased stress in parenting, increased support for the youth mentoring and improved family relationships.Other examples of cultural appropriateness included activities designed to build a stronger sense of cultural identity (p.7) and the program's design around respecting 'cultural values and beliefs' (p.26).
Stacey's evaluation showed that the young people generally valued their relationship with their mentor.
This was not always straightforward to achieve, particularly when there was a history of rejection, poor treatment or losses for young people, and it was important to allow time for this trust to develop (p.73).
The program achieved 'positive shift [s] in the young persons' behaviour and attitude regarding offending', a decrease in their contact with the juvenile justice system, as well as the development of a sense of self-discovery and selfdetermination in the young people and their families (pp.74-5).
Family members who agreed to be involved with other support agencies to address personal or family issues started to improve their skills in responding constructively to their young people, and became more convinced that their young person was making substantial changes.They reported experiencing a decrease in stress over time...(p.75).

Ineffective mentoring practices
Research suggests that there are several practices, described below, that are ineffective in achieving positive changes for at-risk youth in a mentoring program: • Short-term mentoring can actually do harm to at-risk youth: -Some evidence suggests that short-term mentoring programs (6 months or less) may actually disadvantage at-risk youth as they can reinforce or compound the sense of loss and disappointment frequently linked with other youth-adult relationships (Dawes & Dawes 2005).This sense of loss can be particularly acutely felt where the relationship has ended poorly or suddenly.
• A poor relationship can work against change in the mentee's life: -Where there is infrequent or irregular contact between mentor and mentee, where the mentor is authoritarian or judgemental, or where there is too much emphasis on expected behavioural change rather than first building a friendship, mentoring is unlikely to produce any change (Allard et al. 2007;Farruggia et al. 2010).
• Too many goals can lead the mentee to become discouraged and give up (Farruggia et al. 2010).
• Peer mentoring is not an effective replacement for an adult-mentee relationship: -It is often assumed that peer mentoring will be at least as effective as adult-to-youth mentoring, due to the ability of peers to more readily build relationships.Where there is an expectation that peers will act as a proxy for an adult in the mentoring relationship or where peers are expected to convey adult values and norms, peer mentoring appears to be an ineffective strategy for improving the behaviour and attitudes of at-risk youth (Farruggia et al. 2010;Pawson 2004).However, peer mentoring does appear to be effective in managing and preventing chronic disease, and may therefore be useful for health promotion among Indigenous youth at-risk (Adams et al. 2011).

Further research needed
Overt matching of mentor and mentee based upon interest, ethnicity and gender is contentious and does not appear to make a significant difference in improving life outcomes for at-risk youth.
• Although it is important to ensure that mentors and mentees are sufficiently compatible for building a strong relationship, two meta-analyses of non-Indigenous-specific youth mentoring suggest that there is no appreciable difference between matches that are orchestrated and matches that are not (DuBois 2002;Farruggia et al. 2010;Pawson 2004).This needs further research to assess whether there are measurable benefits of ethnic matching in a context that is specific to Indigenous Australians.
• Theories of needing to match female mentees to female mentors may be based more upon cultural expectations of female needs, desires and culturally-conditioned behaviour than actual need (Spencer & Liang 2009).However, the work of Spencer and Liang was not Indigenous-specific and further research is needed to determine whether gender-based matching is beneficial in an Indigenous setting.
Culturally-specific programs can sometimes be effective and at other times counterproductive, as illustrated further in the points that follow.
• The literature around mentoring Indigenous at-risk youth suggests that sometimes it is better to have mixed (or culturally non-specific) programs (for example, Hollis et al. 2011).In contexts where Indigenous young people face social exclusion on the basis of race, having a mixed group of mentees may improve social inclusiveness and cohesion (see Box 3).
• In other contexts where Indigenous young people are disconnected from their history, culture and even family, a culturally-specific mentoring program may be a vital part of building a strong cultural identity (Barwick 2004;Hollis et al. 2011;Stacey 2004).

Box 3: Promising practice: mentoring program in Coonamble, New South Wales
The Stride Foundation uses sports-based mentoring in small country towns to promote healthy youth development through the teaching of life skills, thereby increasing the young people's connection to their community, school and peers (CaFCA 2009:1).
The program was developed with the guidance of a local reference group.It was run over 2 years and targeted both Indigenous and non-Indigenous boys aged 12-18.The community regarded this approach as important in supporting existing efforts to promote an inclusive community (p.3).
Volunteer mentors had to demonstrate involvement in at least one of the local sporting clubs.They were provided with training and then given an opt-out period before committing to the program.Their task was to 'engage the young people, model good social and emotional competencies, and talk openly about their life experiences including how to cope with adversity' (p.2).Where possible, Indigenous mentors were matched to Indigenous boys 'to ensure that culturally sensitive and appropriate mentor partnerships were established' (p.3).
The mentors and mentees were given opportunities to be involved in multiple community initiatives and group activities, including a trip to Sydney to a major league game where they met the professional players.This trip appears to have been crucial in 'develop[ing] and maintain[ing] the motivation and enthusiasm of mentees and mentors throughout the lifespan of the project' (p.4).
In addition, the mentees were involved in a project that constructed a shed at the local sports facilities for storing equipment.This project helped to increase the mentees' sense of connection to the community.
An evaluation of this project demonstrated several sustainable outcomes including reduced rates of school-drop out, improved self-esteem, increased connection to the community, and evidence of skills in positive decision making.These led to positive outcomes in terms of school attendance, employment, drug and alcohol abuse, criminal and antisocial behaviour and overall health (p.6).
There is ongoing debate in research on mentoring of young people about whether paid or volunteer mentors are more influential in achieving positive changes in the lives of at-risk Indigenous youth.
• Although the literature is clear that a paid coordinator is important, it is less clear on whether paid or voluntary mentors are more effective with Indigenous young people.Some studies indicated paid mentors were more effective (Stacey 2004), although others demonstrated that volunteer mentors had greater influence in young people's lives (Pawson 2004).For example, the Panyappi Indigenous Mentoring Program in Adelaide was clearly able to produce greater positive outcomes by having paid mentors who could spend intensive time with mentees (Stacey 2004).However, Farruggia et al. (2010) suggested there is no difference in effectiveness between paid and volunteer mentors.
The effectiveness of mentoring, as a crime prevention strategy, needs closer examination.
• Mentoring appears to indirectly reduce crime rates by channelling at-risk youth away from antisocial behaviour, providing positive peer and adult role models and supporting re-engagement with school and the community.However, the evidence examined here is not able to demonstrate its usefulness directly for the task of reducing crime (Allard et al. 2007;Lemmon 2005).

Conclusion
Young Indigenous Australians in some communities clearly face multiple and complex challenges that can put them at high risk of disengaging from their communities, schools and positive life courses.Although mentoring is no panacea, a robust body of international and Australian evidence clearly demonstrates that it can be one powerful means by which a spiral into negative behaviours and outcomes can be short-circuited.To be effective, mentoring programs need to involve the local community in planning and delivering mentoring, involve parents in the mentoring relationship, and-crucially-must involve long-term, respectful, and mutually fulfilling relationships between mentors and mentees.As an integrated part of a broader youth program, mentoring can help to build protective factors that can facilitate healthy and productive lives.

Appendix
The Closing the Gap Clearinghouse Assessed Collection includes summaries of research and evaluations that provide information on what works to overcome Indigenous disadvantage across the seven Council of Australian Governments building block topics.
Table A1 contains a list of selected research and evaluations that were the key pieces of evidence used in this Resource Sheet.The major components are summarised in the Assessed collection.

Figure 1 :
Figure 1: Process of effecting positive change in at-risk youth Outcomes Risk factors reduced, protective factors increased and youth re-engage with education, employment and community Mentoring program commenced: • progress through affective, direction setting, coaching and advocacy contacts

Box 1 :
Promising practice: mentoring program in Chicano, United StatesBarron-McKeagney et al. (2000) describe (p.38) a program in a highly disadvantaged Chicano neighbourhood* that was typically 'under-served by community programs' (p.38) in the United States which provided mentoring for at-risk minority youth.This program mentored children around the age of 10 (that is, it had an early intervention approach).

Table A2
contains a list of Closing the Gap Clearinghouse issues papers and resource sheets related to this resource sheet.To view the publications, visit <http://www.aihw.gov.au/closingthegap/publications/>.

Table A2 : Related Clearinghouse resource sheets and issues papers
Strategies and practices for promoting the social and emotional wellbeing of Aboriginal and Torres Strait Islander people2013Closing the Gap ClearinghouseStrategies to minimise the incidence of suicide and suicidal behaviour 2013 Closing the Gap Clearinghouse