The intergenerational effectsof forced separation on the social and emotionalwellbeing of Aboriginal children and young people

Family Matters 2006 No. 75 Australian Institute of Family Studies 10

also described the immediate and subsequent effectson individuals who were forcibly removed, institu-tionalised, denied contact with their Aboriginalityand in some cases traumatised and abused. Thereport also includes references to entire communitiesbeing forcibly relocated away from traditional lands ofspecial cultural and spiritual significance.

It is now generally accepted that both forced sepa-ration and forced relocation have had devastatingconsequences in terms of social and cultural dislo-cation and have impacted on the health andwellbeing of subsequent generations. However, untilrecently there has been little or no empirical data to scientifically document the nature and extent of these intergenerational effects. The recent West-ern Australian Aboriginal Child Health Survey

he 1997 Report of the National Inquiryinto the Separation of Aboriginal and Tor-res Strait Islander Children From TheirFamilies, Bringing them home (HREOC,1997), has documented the past laws,

practices and policies which resulted in the separa-tion of Aboriginal and Torres Strait Islander childrenfrom their families by “compulsion, duress or undueinfluence” (HREOC, 1997, p5). Separation took threegeneral forms: putting children into government runinstitutions; the adoption of children by white fami-lies; and the fostering of children into white families.This occurred across the country from the late 1800suntil well into the 1960s. Over this period, as many as100,000 Aboriginal children are believed to have beenforcibly separated, or ‘taken away’, from their fami-lies. Submissions to the Bringing them home Inquiry

TT

The intergenerational effectsof forced separation on the social and emotionalwellbeing of Aboriginal children and young people

It is now generally accepted that both forced separation and forced relocation

have had devastating consequences in terms of social and cultural dislocation

and have impacted on the health and wellbeing of subsequent generations.

However, until recently there has been little or no empirical data to scientifically

document the nature and extent of these intergenerational effects. In this

paper, the authors seek to address this gap in the research.

Sven R. Silburn, Stephen R. Zubrick, David M. Lawrence, Francis G. Mitrou,John A. De Maio, Eve Blair, Adele Cox, Robin B. Dalby, Judith A. Griffin,

Glenn Pearson, and Colleen Hayward

11Australian Institute of Family Studies Family Matters 2006 No. 75

(WAACHS), a large-scale epidemiological survey ofthe health and wellbeing of 5,289 Western Aus-tralian Aboriginal and Torres Strait Islanderchildren, has enabled reliable population estimatesto be made of the number of WA Aboriginal childrenand young people currently living in householdswhere one or more parents/carers and/or grandpar-ents were forcibly separated from family or forciblyrelocated away from traditional lands (Zubrick et al., 2005a). More importantly, the surveymethodology has allowed systematic comparisonsto be made of the associated health and wellbeingoutcomes for the survey children and their par-ents/carers in households affected by forcedseparation or forced relocation in contrast to out-comes observed in households not affected byexperiences of forced separation or relocation.

The Western Australian Aboriginal Child Health SurveyThe Western Australian Aboriginal Child HealthSurvey (WAACHS) was undertaken from 2000 to2002 by the Telethon Institute for Child HealthResearch, in Perth, Western Australia (Silburn etal., 2006; Zubrick et al., 2004; Zubrick et al., 2006;Zubrick et al., 2005b). The survey comprised astate-wide representative sample of around one insix families with Aboriginal children living in West-ern Australia and was designed to build anepidemiological knowledge base from which pre-ventive strategies can be developed to promote andmaintain healthy development and the social, emo-tional, academic, and vocational wellbeing of youngpeople.

Family Matters 2006 No. 75 Australian Institute of Family Studies 12

Survey content. The survey was designed to place aslow a burden on respondents as possible while at the same time acknowledging that sufficient timemust be spent in gaining access, understanding and agood level of rapport with respondents. The inter-views took considerable time, and multiple visits wereoften necessary to ensure complete data and to min-imise respondent fatigue. Interviews were budgetedfor a three-hour time periodper household in which nomore than 90 minutes wouldbe used in formal data collec-tion. Questionnaire contentcovered child and youth devel-opment; health and wellbeing;functional impairment and disability; use and access tohealth, education and socialservices; and a selected num-ber of questions about diet.These data were collectedfrom interviews with the par-ents/carers in the householdwho were the most knowledge-able about the survey children.In addition to the informationcollected on children, separateinterviews were undertakenwith up to two parents/carersper child to gather informationabout the demographic andsocial characteristics of thehousehold and family and toask quest ions about thedwelling, neighbourhood andcommunity. Consent wasobtained from parents/carers and youngpeople to collect separate health and wellbe-ing information from young people aged 12–17 years.

Pilot, dress rehearsal and main survey. Prior tothe dress rehearsal and main survey, a pilot surveywas carried out in September and October 1999 anda full dress rehearsal was undertaken in April 2000.These permitted extensive modifications to surveycontent and process prior to the main survey.

The main survey commenced in May 2000 and wascompleted in August 2002. Dwellings were selectedfor screening using an area-based clustered multi-stage sample design. Census collection districts(CDs) were selected with probabilities proportionalto the number of Aboriginal or Torres Strait Islanderchildren living in the CD. There were 761 CDsselected containing 166,290 dwellings. Of these,139,000 dwellings were approached to determine ifresidents were eligible to participate in the survey.Using this method, a random sample of 2,386 fami-lies with 6,209 eligible children was identifiedthroughout metropolitan, rural and remote regions ofWestern Australia. A total of 1,999 of these families(84 per cent) with 5,513 eligible children consentedto participate in the survey. Interviewers gathereduseable data on 5,289 (96 per cent) of these partici-pating children. In addition to the data gathered onchildren, data were also gathered on families from:2,113 (95 per cent) participating carers identified as

The assessment of the feasibility, design and scope ofthe Aboriginal Child Health Survey was undertakenbetween 1996 and 1999. Survey methodology andinstrumentation were developed in consultation withAboriginal leaders, key Aboriginal bodies andthrough extensive community consultationsthroughout the state. Efforts were made to ensurethat the data collected were both scientifically rele-vant and pertinent to government information needsand policy initiatives. To do this, reference groupswere convened during 1997–1998 with representa-tion from various State and Australian Governmentdepartments and community agencies that had aninterest in the outcome of the survey findings.

Indigenous control. The survey operated withapproval of institutional ethics committees meetingthe requirements of the National Health and Med-ical Research Council of Australia. In addition, allphases of the survey and its development, design,and implementation were under the direction of theWestern Australian Aboriginal Child Health SurveySteering Committee. Established in 1997, theSteering Committee has the responsibility to con-trol and maintain: cultural integrity of surveymethods and processes; community engagementand consultation processes; employment opportu-nities for Aboriginal people; data access issues andcommunication of the findings to the Aboriginaland general community; appropriate and respectfulrelations within the study team, with participantsand communities, with stakeholders and fundingagencies and with the governments of the day.

Area sampling and scope of survey. Western Aus-tralia comprises over one third of the continentallandmass of Australia. The northwest and centre ofthe state includes large tracts of desert and some ofthe most remote and sparsely populated areas in theworld. The more populated southwest of the stateincludes extensive agricultural and forested areaswith numerous small population centres. Over twothirds of the state’s total population and one third ofthe Aboriginal and Torres Strait Islander populationresides in the metropolitan area of Perth. The surveywas based on an area sample of dwellings.

Families in selected dwellings who reported thatthere were Aboriginal or Torres Strait Islander chil-dren or teenagers living at this address who are agedbetween 0 and 18 years, were eligible to be in thesurvey. Children living within group homes, institu-tions and non-private dwellings were not in the scopeof the survey. However, where a selected householdhad a child temporarily living away from home (e.g.in a boarding school or hostel), these children wereincluded in the scope of the survey. Once the author-ity for the survey and the nature of the survey wasexplained to a responsible adult (usually the carer(s)or head of the household), and consent to participatewas obtained, Indigenous status was determined foreach person who was reported to ‘usually’ live in thedwelling by asking, ‘Does (the person) consider him/herself to be of Aboriginal or Torres Strait Islanderdescent?’ Data were gathered on all Aboriginal andTorres Strait Islander children under the age of 18 ineach of the participating households.

13Australian Institute of Family Studies Family Matters 2006 No. 75

the persons who knew the most about the individualsurvey child; 1,040 (83 per cent) other participatingcarers of the survey children wherever this was pos-sible and wherever they were present in thehousehold; 1,073 (73 per cent) participating youngpeople aged 12 to 17 years and the school principaland teacher(s) of 2,379 surveyed children in 410Western Australian schools.

years was selected in the survey. As a result of thisselection hierarchy, the data for individual childrenin the survey sample violate one of the basicassumptions of traditional regression modelling:that the observations are independent. As a result,multi–level, or hierarchical, modelling was used toaccount for the hierarchical structure of the surveydata. This entailed a modification of the method of

Non-response and refusal characteristics. Non-response characteristics and methods for theiradjustment are described extensively elsewhere.With respect to the 387 families who refused partic-ipation in the survey, analyses showed that familieswith older children were more likely to refuse par-ticipation. In the Perth region, household size andsocioeconomic status were also significant predic-tors, with families refusing to participate more likelyto live in large households and live in more disad-vantaged areas. Within the South West and theMidwest and Goldfields regions, the only significantassociation was with older age of child. In the Kim-berley and Pilbara, household size was also asignificant factor with non-respondents more likelyto come from large families. Because of these find-ings, age, region and household size are factors thathave been incorporated into the weighting design.

Analyses. Unlike data collected from a simple ran-dom sample, the survey children are clusteredwithin families and communities. The sample wasselected in three stages: census collection districts(CDs), families and children. CDs were selectedwith probabilities of inclusion in the survey propor-tional to the number of Aboriginal and Torres StraitIslander children living in the CD. A list of all eligi-ble families in each selected CD was prepared, andfamilies were selected at random from this list.Once families had been selected, each Aboriginaland Torres Strait Islander child under the age of 18

Probability Weighted Iterative Generalized LeastSquares (see Pfeffermann et al, 1998) for both con-tinuous and binary response variables. This methodallows children’s health and wellbeing to bedescribed in terms of not only child level factors,but family and community level factors as well. Fulldetails of the WAACHS statistical methods may befound elsewhere (Silburn et al., 2006; Zubrick et al.,2004; Zubrick et al., 2006; Zubrick et al., 2005b).

Experience of forced separation and forced relocation

Households affected by forced separation. Thesurvey asked primary and secondary carers of Abo-riginal and/or Torres Strait Islander origin whetherthey had been ‘taken away’ from their natural family by ‘a mission, the government or welfareagency’. Respondents were not asked to identifywhich of these entities took them, where or whenthey were taken or under what circumstances thistook place. The only information collected waswhether they were taken away. Around 12.3 percent of primary carers and 12.3 per cent of second-ary carers) reported they had been subject to suchseparation. Carers were given the option of not pro-viding answers to questions relating to forcedseparations and relocations and 5.0 per cent of pri-mary carers and 3.8 per cent of secondary carerschose not to answer these questions.

It is now generally accepted that both forced separation and forced relocation have had devastating consequences in terms of social

and cultural dislocation and have impacted on the health and wellbeing of subsequent generations.

Family Matters 2006 No. 75 Australian Institute of Family Studies 14

Mental health and wellbeing of Aboriginal parents/carersThe impact that forced separations may have had onthe social and emotional wellbeing of Aboriginal car-ers of Aboriginal children was investigated byexamining the association between forced separa-tions from natural family and carer reports of mentalhealth and wellbeing. This was done by comparingthe proportions of carers who reported the followingitems: a) problems caused by overuse of alcohol inthe households; b) problems caused by gambling inthe household; c) cigarette smoking, d) whether theprimary carer has a partner; e) whether the primarycarer was ever arrested or charged with an offence; f)were social support networks available to the pri-mary carer; g) had any children of the primary careever been placed in foster care; and h) levels of finan-cial strain in the household.

Survey indicators of health and wellbeing. Theassociation between forced separation and theabove health and wellbeing indicators was analysedusing logistic regression modelling to account for arange of likely confounding factors. Using thismethod it was found that after accounting for age,sex and level of relative geographic isolation, carerswho had been forcibly separated from their naturalfamilies were:

1.95 times more likely to have been arrested orcharged with an offence;

1.61 times more likely to report the overuse ofalcohol caused problems in the household;

2.10 times more likely to report that betting orgambling caused problems in the household; and

less than half as likely to have social support inthe form of someone they can ‘yarn’ to aboutproblems.1

No significant associations were found betweenforced separation and smoking status or financialstrain.

Independent measures of mental health serviceuse. The WAACHS survey methodology includedwritten consent for access to the survey participants’

Aboriginal carers were also asked whether either oftheir parents had been forcibly separated from theirnatural family by a mission, the government or wel-fare agency. Some 20.3 per cent of the mothers ofprimary carers (e.g. grandmothers of the surveychildren) had been forcibly separated. In contrast,12.6 per cent of the fathers of primary carers (e.g.grandfathers of the survey children) had beenforcibly separated. Some 16.1 per cent of secondarycarers reported their mothers had been forciblyseparated and 11.0 per cent reported their fatherswere separated from their natural family.

Of the 29,800 Aboriginal and Torres Strait Islanderchildren and young people living in Western Aus-tralia, 35.3 per cent were found to be living inhouseholds where a carer or a carer’s parent (e.g.grandparent) was reported to have been forcibly sep-arated from their natural family. While the proportionof households affected by forced separation did notvary significantly by level of relative isolation (LORI),some differences were observed between ATSICregions. This variation is shown in Figure 1 belowwhere it can be seen that the Broome ATSIC regionhad the highest proportion of children in familiesaffected by forced separation (53.0 per cent) in con-trast to other regions such as South Hedland (27.3 percent) and Kununurra (26.1 per cent).

Households affected by forced relocation. Primaryand secondary carers were also asked if either they ortheir parents had been forcibly relocated from an areathat was their traditional country or homeland.Around 23.8 per cent of children were living in house-holds that had been affected by such relocation. Infigure 1 below, this percentage varied by ATSIC region,ranging from 41.8 per cent in the Broome ATSICregion to 14.0 per in the Geraldton ATSIC region.

Households affected by forced separation and/orforced relocation. Around 40.9 per cent of childrenwere living in households where at least one primaryor secondary carer had been affected by forced sepa-ration from their natural family or forced relocationfrom traditional country or homeland. The proportionof children thus affected varied across the state with arange from 57.5 per cent in the Broome ATSIC regionto 32.1 per cent in the Geraldton ATSIC region.

Aboriginal children aged 4-17 years – proportion living in households with experience of forced intergenerational separation or relocation*

0

20

40

60

80

Warburton Kununurra DerbySouthHeadland

Broome GeraldtonKalgoorlie Narrogin Perth

Relocated from traditional homeland

Separated from family

AISIC region

Figure 1

Per c

ent

Note: * Whisker bar shows the 95%ile Confidence Interval.

15Australian Institute of Family Studies Family Matters 2006 No. 75

hospital and health system records. This enabled anexamination of children and carers’ use of WA MentalHealth Services by linking survey responses withtheir administrative health records. Bivariate analysisof these data by forced separation found that a higherproportion of primary carers who were forcibly sepa-rated from their natural family had had contact withWestern Australian Mental Health Services (29.5 percent) in contrast to 21.3 per cent among primary car-ers who had not been forcibly separated.2 Furtheranalysis by logistic regression modelling taking intoaccount possible confounding factors confirmed thatafter adjusting for age, sex and level of relative geo-graphic isolation, those carers who had been forciblyseparated from their natural family were 1.50 timesmore likely to have had contact with Mental HealthServices in Western Australia.

Emotional and behavioural difficulties in childrenAssessment of emotional and behavioural prob-lems. To assess Aboriginal children’s emotional andbehavioural difficulties, a modified version of Good-man’s Strength and Difficulties Questionnaire (SDQ)was used. Details of the cultural adaptation, pilotingand evaluation of its reliability and consistency bymeans of confirmatory factor analysis are available inthe technical report accompanying Volume 2 of theWA Aboriginal Child Health Survey (De Maio et al.,2005). This measure of Aboriginal children’s emo-tional or behavioural difficulties was analysed byforced separation of their primary carer from theirnatural family. This revealed that of the childrenwhose primary carer was forcibly separated fromtheir natural family by a mission, the government orwelfare agency, nearly one third (32.7 per cent) wereat high risk of clinically significant emotional orbehavioural difficulties. This proportion is signifi-cantly higher than that found in children looked afterby primary carers who had not been forcibly sepa-rated from their natural family (21.8 per cent) (SeeFigure 2). In comparison, 15.0 per cent of non-Abo-riginal Western Australian children aged 4–17 yearswere found to be at high risk of clinically significantemotional or behavioural difficulties.

The relationship between forced separation of the pri-mary carer from their natural family and SDQ scoresof the children in their care was also examined bylooking at the scores on a continuous scale. As shownin Figure 3, the proportion of children whose primarycarer had been forcibly separated from their naturalfamily increased steadily with the increasing totalSDQ score of the child. The rate of increase was great-est for children whose SDQ scores were above 22.

Once again, logistic regression modelling was usedto analyse the likelihood of Aboriginal childrenexperiencing emotional or behavioural difficultiesafter accounting for a number of factors. The modeladjusted for: a) age group of child (4-7, 8-11, 12-14,15-17 years); b) level of relative isolation; c) sex ofchild; and d) birth mother status of primary carer(i.e. natural mother/non-natural mother). Thisanalysis showed that independently of these fac-tors, children whose primary carer had been

forcibly separated from their natural family by amission, government or welfare agency were 2.34times more likely to be at high risk of clinically sig-nificant emotional or behavioural difficulties thanchildren whose carers were not forcibly separated.

Inter-generational effects of forced separationNo significant findings were made with respect torisk of clinically significant emotional or behaviouraldifficulties in children having a grandparent only (i.e.primary carer’s father or mother) who had beenforcibly separated from their natural family by a mis-sion, the government or welfare agency. However,although not statistically significant, the data weresuggestive of an inter-generational impact on thechild, particularly in cases where the primary carer’smother was forcibly separated. Among children forwhom the primary carer’s mother had been forciblyseparated from her natural family, 27.2 per cent wereat high risk of clinically significant emotional orbehavioural difficulties, compared with 22.3 per centof children for whom the primary carer’s mother wasnot forcibly separated from her natural family.

Logistic regression modelling found that afteraccounting for age, sex and level of relative isolation,

Children aged 4-17 years – risk of clinically significant emotional or behavioural difficulties, by whether their primary caregiver was forcibly separated

0

20

40

60

80

Forcibly separatedNot separated

Moderate

High

Low

Carer forcibly separated from natural family

Figure 2

Per c

ent

Children aged 4–17 years – proportion of children whosecarer was forcibly separated, by total SDQ score of the child

0

10

20

30

40

30 4020100

ModerateLow High

SDQ Total Score

Figure 3Pe

r cen

t

Aboriginal people from their natural families. Whilethe survey findings are confined to the effects of pastchild removal policies on the Western Australian Abo-riginal population, they help to inform aspects of thenational discussion which has followed the release ofHuman Rights and Equal Opportunity Commission’sBringing them home report (HREOC, 1997).

Much of this discussion has centred on the report’ssummary finding that “somewhere between one-in-three and one-in-ten Aboriginal children had been separated from their familiesbetween 1910 and 1970”. The“one-in-three” estimate has beenwidely criticised on the groundsthat it over-generalised the findingfrom a number of local studies inMelbourne, the Kimberley region ofWestern Australia and the Bourkeregion of New South Wales (Mann,2001). The lower estimate of ‘one-in-ten’ was based on the 1994 ABSNational Aboriginal and TorresStrait Islander Survey whichreported that 10.1 per cent of thoseaged 25-44 years, and 10.6 per centof those older than 44 years hadbeen separated from their naturalfamily by missions or governmentor welfare agency (ABS, 1995). Thefindings from the WAACHS areconsistent with the NATSISS datain showing that a much higher proportion of child separationoccurred within Western Australiathan occurred nationally.

Given the differences in removalpolicies which existed between theStates and the ways in which thesechanged in their application over time, itseems unlikely that the number of Aboriginal andTorres Strait Islander people who were separated willever be precisely ascertained from historical sources.This suggests that the current lived experience of Aboriginal people as reported in representative cross-sectional population surveys such as the WAACHS,NATSIS (1994) and the NATSISS (2002) will have tobe relied upon for the best estimate of the minimumnumber of people and families so affected. The logis-tic regression analysis in this study did take accountof a range of possible confounding factors, such as thecarer’s age, sex and the level of geographic isolation.Other potential confounders which were not consid-ered include the age of those forcibly removed at thetime of separation from their natural family, and theofficially recorded reasons for the removal (i.e.whether for reasons of abuse or neglect, or throughenforcement of the then policies of assimilation).While this is a limitation of the study, the historicalevidence from official records reported in the Bring-ing them home report does suggest that most of thechildren who were forcibly removed were removed asyoung infants, and that that prior to 1969, the greatermajority of forced removals occurred for reasons ofassimilation rather than reasons of child protection(HREOC, 1997).

those children for whom both their primary carer andtheir primary carer’s mother had been forcibly sepa-rated from their natural family were over two and ahalf times as likely (Odds Ratio = 2.62) to be at highrisk of clinically significant emotional or behaviouraldifficulties, while those children whose primary carerwas forcibly separated but the primary carer’s motherwas not separated were over twice as likely (OddsRatio = 2.33) to be at high risk of clinically significantemotional or behavioural difficulties. If only the pri-mary carer’s mother was forcibly separated from hernatural family, there was no significant difference inlikelihood of being at high risk of clinically significantemotional or behavioural difficulties (Odds Ratio1.17). These results confirm the impact of the forcedseparation of the primary carer from their naturalfamily on the risk of clinically significant emotional orbehavioural difficulties in his or her children, butshow no evidence to suggest there is any furtherimpact beyond two generations.

Inter-generational effects of forced relocation With regard to risk of clinically significant emo-tional or behavioural difficulties in children and theforced relocation of Aboriginal carers from tradi-tional country or homeland, there were no findingsof statistical significance. However, the data weresuggestive of an inter-generational impact on thechild in the case where the primary carer’s parentswere forcibly relocated.

Age of carer and reason for forced separation Since the original publication of these findings thequestion has been raised whether the parents/carersof the children surveyed in the WAACHS would havebeen old enough to have been subject to the past poli-cies of forced removal. While our analysis did accountfor carer age as a possible confounder, this promptedus to re-examine the data to ascertain whether thefindings would be modified in any way if the analysiswere restricted to carers born prior to 1966, and whowere thus at risk of being forcibly separated from theirnatural families as part of the government policies andpractices of the day. While 43 per cent of all Aborigi-nal carers surveyed were born prior to 1966, thesecarers represent 63.8 per cent of Aboriginal carerswho reported being forcibly separated. Any of thesecarers forcibly separated from their natural family,are thus likely to have been separated at a time whenthere was significant change in the practice of forcedseparation in Western Australia. It is thus of particularnote that when we repeated the logistic regressionanalyses restricted to only carers born prior to 1966and their children, no differences were found from ouroriginally reported findings (Zubrick et al., 2005a).

DiscussionThe interview question on forced separation used inthe WAACHS survey was identical to that used in the2002 ABS National Aboriginal and Torres StraitIslander Social Survey (NATSISS) (ABS, 2004). Whilethere are some differences in the methodology ofthese two surveys, the NATSISS and the WAACHSboth demonstrate the links between adverse healthand social outcomes and prior forced separation of

Family Matters 2006 No. 75 Australian Institute of Family Studies 16

In conclusion, we believe the nature of the recentdebate about the actual number of Aboriginal chil-dren and families where a parent or grandparentexperienced forced separation has displaced andexcluded from the national discussion the realitythat these experiences occurred at all – and theextent to which these past experiences continue toimpact on the lives of the current generation of Aboriginal families. It is our hope that the wideravailability of the WAACHS findings will enable amore nuanced and compassionate discussion of the

Pfeffermann, D, Skinner CJ, Holmes DJ, Goldstein H, & Rasbash J.(1998). Weighting for unequal probabilities in multi-level mod-els. Journal of the Royal Statistical Society, Series B, 60, 23-40.

Silburn, S. R., Zubrick, S. R., De Maio, J. A., Shepherd, C.,Griffin, J. A., Mitrou, F. G., et al. (2006). The WesternAustralian Aboriginal Child Health Survey: Strengtheningthe capacity of Aboriginal children, families and communi-ties. Perth: Curtin University of Technology and TelethonInstitute for Child Health Research.

Zubrick, S. R., Lawrence, D. M., Silburn, S. R., Blair, E., Milroy,H., Wilkes, E., et al. (2004). The Western AustralianAboriginal Child Health Survey: The health of Aboriginalchildren and young people. Perth: Telethon Institute forChild Health Research

17Australian Institute of Family Studies Family Matters 2006 No. 75

A more open-hearted acknowledgement of the extent of the suffering and disadvantage which the past policies of separation inflicted

on Aboriginal Australians, would in our view, significantly further the process through which these concerns are eventually resolved.

enduring impact of past forced separations and whythese experiences remains of such deep concern forthe affected individuals and families. A more open-hearted acknowledgement of the extent of thesuffering and disadvantage which the past policies ofseparation inflicted on Aboriginal Australians, wouldin our view, significantly further the process throughwhich these concerns are eventually resolved.

Endnotes1 Odds Ratio (OR) is a statistical term which describes whether

the probability of an event is the same for two different groups.For example, an Odds Ratio of 1 implies that the event is equallylikely in both groups. Where the value of an Odds Ratio is greaterthan 1, this indicates how much more likely the event is in thefirst group. Similarly, where an Odds Ratio is less than 1, thisindicates how much less likely the event is in the first group

2 Not significant at the .05 level in the bivariate analysis.

ReferencesAustralian Bureau of Statistics. (1995). National Aboriginal and

Torres Strait Islander Survey 1994 detailed findings.Catalogue 4190.0. Canberra: Author.

Australian Bureau of Statistics. (2004). National Aboriginal andTorres Strait Islander Social Survey 2002 Western Australia.Catalogue 4714.5.55.001. Canberra: Author.

Human Rights and Equal Opportunities Commission (HREOC).(1997). Bringing them home: Report of the national inquiryinto the separation of Aboriginal and Torres Strait Islanderchildren from their families. Canberra: Author.

Mann R. (2001). In denial: The Stolen Generation and the right.The Australian Quarterly Essay, 1, 1-113.

Zubrick, S. R., Silburn, S. R., De Maio, J. A., Shepherd, C.,Griffin, J. A., Dalby, R. B., et al. (2006). The WesternAustralian Aboriginal Child Health Survey: Improving theeducational experiences of Aboriginal children and youngpeople. Perth: Curtin University of Technology and TelethonInstitute for Child Health Research.

Zubrick, S. R., Silburn, S. R., Lawrence, D. M., Mitrou, F. G.,Dalby, R. B., Blair, E. M., et al. (2005a). The WesternAustralian Aboriginal Child Health Survey: The social andemotional wellbeing of Aboriginal children and youngpeople. Perth: Curtin University of Technology and TelethonInstitute for Child Health Research.

Zubrick S.R., Silburn S.R., Lawrence D.M., Mitrou F.G., DalbyR.B., Blair E.M., et al. (2005b). The Western AustralianAboriginal Child Health Survey: Forced separation from nat-ural family, forced relocation from traditional country orhomeland, and social and emotional wellbeing of Aboriginalchildren and young people: Additional notes. Perth: CurtinUniversity of Technology and Telethon Institute for ChildHealth Research.

Sven Silburn and Stephen Zubrick are clinical psycholo-gists and co-directors of the Centre for DevelopmentalHealth, Curtin University of Technology and the TelethonInstitute for Child Health Research. David Lawrence isSenior Statistician, Centre for Developmental Health. Fran-cis Mitrou, John De Maio, Eve Blair, Adele Cox, RobinDalby and Judith Griffin are members of the WAACHS vol-ume 2 analysis and writing team at the Telethon Institutefor Child Health Research. Colleen Hayward heads theKununga Indigenous Research network at the TelethonInstitute for Child Health Research. She and Glenn Pearsonare responsible for the overall project management of theWAACHS, including the community communication anddissemination strategy.