DESIGN OPTIONS FOR THE BUILDINGANEW
LIFE INAUSTRALIALONGITUDINALSURVEY OF
HUMANITARIANMIGRANTS
June2012
MATTHEWGRAY,ADAMGRAYCARANDLOUCASNICOLAOU
RESEARCHSCHOOLOFSOCIALSCIENCES
AUSTRALIANNATIONALUNIVERSITY
ProfessorMatthewGrayandProfessorAdamGraycar are fromthe Research Schoolof Social Researchatthe AustralianNationalUniversity. DrLoucas Nicolaouis theChiefExecutiveOfficer,FederationofEthnic Communities'CouncilsofAustraliaandatthetimeofthe preparationofthispaperwas attheResearchSchool of Social Sciences, AustralianNational University.
Theauthors aregratefulforcommentsprovidedbystafffromthe DepartmentofImmigrationandCitizenship onanearlierversionofthispaper. EditorialassistancewasprovidedbyJohn Hughes. Theviewsexpressedin this paper are those of the authors and may not reflect those of the Department of Immigration and Citizenshipor the Australian Government.
1.INTRODUCTION
Australiaaccepts asubstantialnumber of humanitarianmigrants each year,withabout133,000humanitarian migrantshavingsettledin Australiaover the period 2001‐02to 2010‐11.Themajorityofmigrantsmakea successfultransitiontolifeinAustraliaandmakesignificanteconomicandsocialcontributionstoAustralia. However,as agroup,migrantshavea greater level of vulnerabilitythanthe non‐migrantpopulationandwithin themigrant populationhumanitarianmigrants are,onaverage,substantiallymorevulnerable thanother migrants.
There are a range of reasonsfor thegreater vulnerability of humanitarianmigrants whichinclude:havingbeing forcedtoleave theirhomeland;oftenhaving beenunabletoprepare for migration;oftennot speakingEnglish; havinglowlevelsofeducationincludingliteracyandnumeracy;oftenhavingexperiencedseveretrauma; having been unable to bringfinancial resources with them;andoften lackingthe family andsocial networks in Australiathatother migrantshave(Hugo, 2010,39).
Havingreliable dataonthesettlementexperiences,outcomes andserviceneeds ofhumanitarianmigrants is importantforAustralia’simmigrationpoliciesandforensuringthatpeoplewhohavebeenforcedtoleave their homelandandwhocome toAustraliaareableto resettleas well as possible.However,there is only limitedup‐to‐datedataavailableonrecenthumanitarian migrants toAustraliaandnoup‐to‐datelarge‐scale longitudinalsurveys.There isaclearneedforahigh qualitylongitudinalsurveyofrefugees andother vulnerablemigrantstoAustraliatohelpinformpublicpolicydevelopmentinthisarea(seeforexampleHugo,
2010.1
Whilepoint‐in‐time (cross‐sectional surveys)provide valuable informationonthe settlement experience and outcomesofmigrants,therearemanypolicy,servicedeliveryandresearchquestionswhichcanbe answered onlythroughlongitudinaldata(thatis,datafromsurveysthatfollowagroupofrefugeessettlingin Australia over time). AsHugo(2010,39)has argued,“thereis nearunanimityamongmigration researchers globallythat longitudinal approaches to researching migration provide a deeper understanding of the processes of migrationandsettlementthanothermethodologies.”.Forexamplelongitudinal dataisrequiredinorderto properlyunderstandthedynamicsofthesettlementexperienceandthedurationandpersistence ofvarious outcomessuchaspovertyordepression. Anotherexamplewerelongitudinaldatais particularlyvaluableisin identifyingcausalrelationshipbyallowingforunobserveddifferencesforpeopletobe takeninto accountand byallowingthe time orderingofevents tobe analysed.
Inorder to fillthis important data gap,the AustralianDepartment of ImmigrationandCitizenship(DIAC) is fundinga new longitudinalsurveyofhumanitarianmigrantstoAustralia—theBuildingaNewLifeinAustralia Longitudinal SurveyofHumanitarianMigrants.As partofthescopinganddevelopmentprocessforthesurvey, DIACcommissionedthe AustralianNational Universitytopreparethispaperoutliningdesignoptions forthe survey. Thepaper discusses the keydesignchoices,discusses the strengths andweaknesses ofdifferent design options, andoutlines a suggested designandarchitecturefor theBuildingaNew LifeinAustraliaSurvey.
DIAChasalsoproposedthatthedatacollectionforthesurveytakeplaceovertheperiod2012to2017,and thatitwill encompass thefirstfiveyears of settlementin Australia.In developingthe proposed design, DIAC indicatedanapproximatebudgetwithinwhichtheproposeddesignoutlinedhereshouldbebased.Whilethe
1Amoredetaileddiscussionoftherangeofpotentialbenefitsofhavingalongitudinalsurveyofhumanitarian migrantsareprovidedby Hugo (2010)and Smith (2010).
proposedsurveydesignhastakenaccountofthisbudget,ithasnotbeenformallycostedandtheactualcost ofimplementingastudyalongthelinesof theproposeddesignwilldependuponanumberofdetaileddesign decisions thatwill needtobemade.
Theoptimaldesignofanysurveyisdeterminedbytheissuesitisdesignedtoprovidedataabout.Thekey policy andresearchquestions identified by DIACthatthesurveyneeds toaddress are:
Thesettlementoutcomesofhumanitarianmigrantsandothervulnerablemigrants,howthey change duringthefirstfive‐yearsfollowingsettlement,andthe factorsthatinfluence settlement outcomes;
Theimpactofpre‐settlementexperiencesonsettlementoutcomes,includinghowtheyvary between on‐shore andoffshorehumanitarianmigrants;and
Theneedfor,andadequacyandappropriatenessof,settlementandothersupportservices available to humanitarian and other vulnerable migrants during their first five years in Australia.
Thetermsofreferenceforthispaperalso indicate thattheoptionofincludingvulnerablenon‐humanitarian migrants in the study should be explored, but that the priority for the survey is the experiences of humanitarianmigrants.Providingdata thatcanbe usedto conductresearchintothevulnerabilityofother groupsof immigrants, while important,is ofsecondary importance for the purposes ofthis survey.
Animportantlonger‐term questionis whether the settlement outcomes of the currentgroupof refugees differ tothoseofpreviouscohorts ofrefugeeswhosettledinAustraliaandifso,whattheimplications ofthisare.As willbeoutlinedinthispaper,Australiaisfortunateinhavingcross‐sectionalandlongitudinaldatacollected from earlier cohorts of refugees at comparable points intheir settlement processes. These earlier data collectionscanpotentiallybeusedascomparisonpointsfortheBuildingaNewLifeinAustraliaSurvey.In orderforthistobepossible,itwillbeimportanttoensurethatatleastsomeofthedataitemsusedto measure settlement outcomes are comparabletodata items usedinearlier Australiansstudies.
Thedevelopmentofthesuggesteddesignforthesurveyhasinvolvedanumberofsteps.Thedesigns of previous longitudinalsurveysofmigrantstoAustraliawerereviewed,withafocusontheirutilityfor understanding theprocess of settlement,especiallyofrefugees.Challenges involvedinsurveyingrefugees and followingthemovertimearethus discussed,andlessonslearntfrom otherlongitudinalsurveysconductedin Australiaor elsewhere are considered. Inaddition,consultations were undertaken with academicresearchers, thosewhohavehadresponsibilityfordesigningandrunningmajorlongitudinalsurveys(includingsurveys of immigrantsandsurveys of other groups),thoseinvolvedinprovidingservicestorefugees,andofficersfrom DIAC.Finally,earlierversionsofthispaper werealsopresentedtoanddiscussedattheinauguralmeetingsof the twogroups established to guide the study,theTechnical AdvisoryGroupand the SurveyReferenceGroup.
Theremainderofthispaperisthus structuredas follows.Section2providesasummaryofexistingAustralian andinternational longitudinal studies ofimmigrants andSection3 provides abriefoverview ofAustralia’s immigrationprogram. InSection4the suggesteddesignof the surveyis described. Section5discusses sample sizeconsiderationsandSection6strategiesformaximisingtheinitialresponserate andminimisingsample attrition. Thefinal sectionconcludes.
2.EXISTINGLONGITUDINALSURVEYSOFMIGRANTS
Thissectionprovidesanoverviewofselectedlongitudinal surveysofmigrantsinorder toinformthe designof theBuildingaNewLifein AustraliaSurvey.Theoverviewfocusesprimarilyuponlongitudinalsurveysof migrants, giventhat theseare mostrelevantto thedesignof theBuildingaNew LifeinAustralia Survey.
TheLongitudinalSurveyofImmigrantstoAustralia(LSIA)isthemostrelevantstudytothediscussionshere. LSIAcollected dataonthesettlementexperiencesofthreecohortsofimmigrantsandhasprovidedAustralia withhighqualitylongitudinal data onmigrantsfora numberofyears.ThefirstLSIA(LSIA1)surveyedmigrants enteringAustraliabetweenSeptember1993andAugust 1995.Itinvolvedthreewavesofinterviewsoverthe firstthree and ahalfyears after arrivinginAustralia. LSIA2surveyed migrants enteringAustraliabetween September1999andAugust 2000andinvolvedtwowavesofinterviewsoverthefirst18monthsafterarrival. ThefinalisLSIA3,whichisasurveyofmigrantsenteringAustraliabetweenDecember2004andMarch2005
who were interviewedtwiceover the first 18months after arrival.2
Both the LSIA1 andLSIA2included humanitarianmigrants,with 831interviewedat wave 1of LSIA1and558 interviewedatwave1ofLSIA2. However, theLSIA3samplewas selectedonlyfrom theFamilyandSkill immigrationstreamsandthusdoesnotincludeanyhumanitarianmigrants.TheLSIA1andLSIA2havebeen usedextensivelytoanalysethesettlementexperiencesandoutcomesofhumanitarianmigrants(e.g.Cobb‐ ClarkandKhoo,2006;VandenHeuvelandWooden,1999;Richardson,RobertsonandIlsey,2001;Richardson et al., 2004).
WhilebothLSIA1andLSIA2haveprovidedvaluableinsightintotheexperiencesofhumanitarianmigrants, theydohave—asisthecasewithallsurveys—limitationsthatarisefromtheirdesign.First,the sampleof humanitarianmigrantsisrelativelysmall,whichlimitstheextent towhichtheexperiencesofdifferentgroups can be compared. Second, LSIA1andLSIA2are relativelyshort panels, covering only 3.5years inthe case of LSIA1and1.5yearsinthecaseofLSIA2.Thismeansthattheycanonlybeusedtoexaminetheinitial settlementexperienceandcannotbe usedtoexaminesettlementoutcomesoverthemedium tolonger‐term. Third,thesampleforLSIA1andLSIA2didnotincludeonshoremigrants.Fourth,thesamplingframefrom whichthesamplewasselectedhadmissingcontactinformationforasubstantialproportionofthesample, whichraises concerns abouttherepresentativenessofthesample.Hugo(2011)notes thataddress detailsare morelikelyto beaccurateformigrantswhousegovernmentservices,andthusthesampleislikelytoover‐ represent thisgroup.3
Additionally,DIACalsorunstheContinuousSurveyofAustralia’sMigrants(CSAM),atwowavelongitudinal surveywhichinvolvesparticipants beinginterviewedtwiceovera six‐monthperiod. CSAMis designedto providetimelyinformationonlabourmarket outcomesfor recentimmigrantsfromtheFamilyandSkillStream (Smith,2010,p.24).
Whilethere aremanysmallerlongitudinal studiesofmigrantsandindeedofhumanitarianmigrants,their findings cannotusually be extrapolatedwithany confidenceto the broader experiencesof immigrantsbecause of either:small samples sizes;limitedcountryof originrange;or onlyincludingimmigrants settlingina particulargeographicregion.Selectedexamples of suchstudies are:
2AnoverviewofLSIA1andLSIA2isprovidedbyCobb‐Clark(2001).InformationonLSIA3hasbeenobtainedfrom variousdocuments on theDIACwebsite.
3 Forexample,forLSIA1,ofthe10,141inscopeprimaryapplicantarrivalsinitiallyselectedfromtheSettlement
Database,usable addresseswerefound for only4,178(Hugo,2011).
TheAMESLongitudinalStudy(2008–10)followed245recently‐arrivedmigrantsandrefugees enrolledintheAdultMigrantEducationPrograminMelbourne.Wave1wasconductedover theperiodSeptember2008‐March2009,withwave2conductedsixmonthsafterstartingAMEP andwave 3some 18months aftercommencingAMEP.
SettleMEN which followed 233 recently‐arrived (2004–2008) men (over 18) from refugee backgroundslivinginurbanandregionalareasofSouthEastQueenslandbetween2008and
2010.Therewasabaseline surveyandthreefollow‐upsurveysatsix monthintervals.The surveywasconductedbytheLaTrobeRefugeeResearchCentre,withtheaimtodocument healthandsettlement experiences.
GoodStartssurveyedrecently‐arrivedyouthwithrefugeebackgrounds.Thesurveyfollowed
120recently‐arrivedyoungpeoplewithrefugeebackgroundssettlinginMelbournebetween
2004and2008.Participants hadbeeninAustraliafor anaverageof5.6monthsandwere recruitedthroughthreeMelbourneEnglishLanguage Schools.Datacollectioninvolved participants keeping‘settlementjournals’,aswellasactivitiesinclassrooms,homesandpublic libraries,andconversationsinpersonor overthephone.Quantitativedatawasgatheredover three years,and qualitative dataover four years.Thissurveywasconductedby theLaTrobe RefugeeResearch Centre.
TheLongitudinalStudyoftheHealthandWellBeingofRefugeeChildrenovertheirFirstTwo Yearsof SettlementisbeingrunbytheSydneyChildren’s HospitalDepartmentofCommunity ChildHealth. ItbeganinDecember2010andwillrununtilDecember2012.Thesurvey populationis refugeechildren(aged0–16years)arrivinginAustraliaonpermanent humanitarianvisas andsettlingwithinthe SouthEasternSydneycatchmentareaandthe Illawarra /Shoalhaven local health networks. The first data was collected within sixweeks of their arrival,withfollow‐up collections occurringsix‐twelvemonths post‐arrival andthenat yearlyintervals.
Other countriesthathaveundertakenlarge‐scalerepresentativesurveysofimmigrantsincludeCanada, Germany,New Zealand,the UnitedKingdomandtheUnitedStatesofAmerica.Asummaryofselectedmajor longitudinalstudiesofimmigrants (includingLSIA)isprovidedinTable1.Informationincludedinthetable encompasses:
The time periodover which data wascollected;
The length of time between the first andlastwave;
The length of time after settlement atwave 1;
The methodofdatacollection;
Whetherhumanitarian migrants are includedin the sample;
The number ofwaves;
The wave 1 responserate;
The wave 1 sample size;
The refugee sample size at wave 1; and
Thesampleretentionrates(i.e.,theproportionofthewave1samplere‐interviewedateach subsequent wave).
Thereareseveralobservations thatcanbemadeaboutthestudiesincludedinTable1 thatarerelevanttothe design of the BuildingaNewLifeinAustraliaSurvey.
Forallbutoneofthelongitudinalsurveysofimmigrantsidentified,thefirstwaveofdata collectionis undertaken around six monthsafter arriving inthecountry4
Thestudieshavedifferedinthelengthoftimebetweeninterviews(seeTable1),althoughthey havetended towards more frequent interviews;
Theinitialresponseratesforthestudies(allofwhichusedadministrativedataasthesampling frame)varied from49% to 70%.Thisiscomparable tothe initialresponserateachievedbythe Longitudinal Studyof AustralianChildren (LSAC)whichusedthe Medicare databaseasits samplingframe.5Theinitialresponserate isgenerallyhigherforface‐to‐facethan postal surveys;and
Sampleretentionratesvariedgreatlyandwerelowerformail‐outsurveysthanface‐to‐face surveys.
Thesampleattritionrate(theinverseoftheretentionrate)doesnotdropafterwave2.This differsfromtheexperience ofmostlongitudinalstudies ofother groupsandsuggeststhat significantresources should be putintosample retention.
4Theonlysurveywhichconductedthefirstwave ofinterviewssubstantiallyearlierthanthefirstsixmonthsinthe countrywastheSurvey ofNew Refugees: ALongitudinalStudyofRefugeeIntegrationintheUKwhichcollected dataalmostimmediatelyaftersettlementintheUK. Thissurveyhadagoodinitialresponserate(70%)butthena verylow retention ratewith just 17%of thewave1samplereinterviewedatwave4.
5IntheLSACsurveythefinalwave1samplerepresented53%ofallfamilieswhoweresentaletterbyMedicare.
Therewasa10%non‐contactrate(addressdetailsoutofdateorifonlyapostofficeboxaddresswasavailable). Ifnon‐contactablefamiliesareexcludedtheresponserateachieved atwave1ofLSACisabout60%(Grayand Smart,2008).
Gray,GraycarNicolaou
TABLE1. SELECTEDEXAMPLESOFLONGITUDINALSTUDIESOFIMMIGRANTS
RetentionrateTime period / Time betwee n1standlast
wave
(months
) / Lengthtime after settlement atW1 (months) / Methodof data collection / Include d refuge es / Number ofwaves / W1 response rate / W1 sample / Refugee samplesize (atwave1) / W1‐
W2 / W1‐
W3 / W1‐
W4
LongitudinalSurveyof
ImmigrantstoAustralia1 / 1994‐99 / 36 / 6 / F2F / Yes / 3 / 59% / 5,192 / 831 / 86% / 72% / ‐
LongitudinalSurveyof
ImmigrantstoAustralia2 / 2000‐02 / 12 / 6 / F2F / Yes / 2 / 59% / 3,124 / 558 / 85% / ‐ / ‐
LongitudinalSurveyof
ImmigrantstoAustralia3 / 2005‐06 / 12 / 6 / W1‐Mailout
surveyW2–
telephone / No / 2 / 49% / 9,865 / 0 / 53%* / ‐ / ‐
LongitudinalSurveyof
ImmigrantstoCanada / 2001‐05 / 42 / 6 / F2F / Yes / 3 / 60% / 12,040 / 2,269 / 79% / 64% / ‐
SurveyofNewRefugees:A LongitudinalStudyofRefugee IntegrationintheUK / 2005‐09 / 21 / 1‐week / Mailout survey / Yes / 4 / 70% / 5,678 / 5,678 / 32% / 22% / 17%
TheLongitudinalImmigration
Survey:NewZealand / 2005‐09 / 30 / 6 / F2F / No / 3 / 66% / 7,137 / 0 / 85% / 72% / ‐
USNewImmigrantSurvey / 2003‐08 / 60 / F2F,someuse ofphonein W2 / Yes / 2 / 69% / 8,573 / ** / 46% / ‐ / ‐
Notes:F2Findicatesthatthedatawascollectedviaanintervieweradministeredinterviewinperson(face‐to‐face).
Wherethesurveyinvolvesinterviewingmorethanonepersonperhousehold,thesamplesizegiveninthetableisthenumberofhouseholdsratherthanthenumberofrespondents. Forthehouseholdbasedsurveystheattritionrateiscalculatedfortheprimaryrespondent.
TheUSNewImmigrantSurveyalsoincludedasampleofimmigrantswithchild‐of‐U.S.‐citizen visaswhoareunder18yearsofageandadoptedorphansunderfiveyearsofage.The childsamplewas810andinterviewswereconductedwithsponsor‐parentsofthesampledchild.Theinformationprovidedinthetableisfortheadultsample.
Allofthestudiesusedadministrativedataasthesamplingframe.
*TheretentionrateofLSIA3islowduetoadeliberatedecisiononlytofollow‐uphalfofthefamilystreamcohort.
**MigrationpathwaystotheUSarecomplicatedanditisnoteasytoderiveafigureforthenumberofrefugeesintheNISsample.Thereappeartobe175respondentswhoare classifiedasusingtherefugee/asyleemigrationpathway(HayesandHill,2008).
Sources:Informationonstudiesisfromacombinationofdocumentsontherespectivestudywebsites,publishedarticlesandinsomecasespersonalcommunicationswiththoseresponsiblefor thestudy
June2012Page7
3.OVERVIEWOFAUSTRALIA’SIMMIGRATIONPROGRAM
Indevelopingmethodologicaloptionsfora longitudinalsurveyofrefugees,itisimportanttohaveaclear understandingofAustralia’simmigrationprogram.Thefirstpartofthissectionprovidesabriefoverviewof theprogram,withafocusonaspectsthat arerelevanttothedesignsof thesurvey.6Whiletheprimaryfocusis onhumanitarianmigrants, thetermsof reference forthisprojectindicatethat DIAC isalsointerestedin the surveyprovidingdataon thevulnerability ofothermigrants. Thesecondpartofthis sectionprovidesbasic statisticalinformationaboutthenumbersofmigrantsarrivingunderthedifferentmigrationprograms,and thenfor humanitarianmigrantsinformationisprovidedoncountryofbirth,theiragedistributionand geographicdistribution across Australiain the initial post‐migrationperiod.
InAustraliathere are two pathways to permanent migration: (1)the HumanitarianProgram for refugees and those in humanitarian need; and (2) the Migration Program for family and skilled immigrants. The HumanitarianProgramconsistsoftwocomponents.TheoffshorecomponentisforpeoplenotinAustralia whoseektomigratetoAustraliaundertheHumanitarianProgram,whiletheonshore componentisforpeople who claim asylum once theyhavearrivedinAustralia(either by boator byair).
TheHumanitarianProgram isforrefugeesandothers inrefugee‐likesituationsandhastwoimportant functions.Theonshoreprotection/asylum componentfulfilsAustralia'sinternational obligationsbyoffering protectiontopeoplealreadyinAustraliawhoarefoundto berefugeesaccordingtothe RefugeesConvention.7
Theoffshore resettlementcomponentexpressesAustralia'scommitmenttorefugeeprotectionbygoing beyondtheseobligationsandofferingresettlementto peopleoverseas.Withintheoffshoreresettlement componenttherearefivevisasubclasses,fourofwhichfallwithinthe Refugeecategoryandonewithinthe SpecialHumanitarianProgram.TheRefugeecategory helpspeoplewhoaresubject topersecutionintheir homecountry,who aretypicallyoutsidetheirhomecountry,andare inneedof resettlement.The visatypes under thiscomponentareRefugee,In‐CountrySpecialHumanitarian,EmergencyRescue,andWomanatRisk. Themajorityofapplicants whoare consideredunderthis categoryare identified and referredbyUNHCR to Australiaforresettlement.
TheSpecialHumanitarianProgramhelps peopleoutsidetheirhome countrywhoaresubjecttosubstantial discriminationamountingto grossviolationofhumanrightsintheirhomecountry,andtheimmediatefamilies of persons who have been granted protection in Australia. Applications for entryunder the SHP must be supportedbyaproposer whois anAustraliancitizen,permanentresidentor eligibleNew Zealandcitizen,or an organisation thatis basedinAustralia.
TheMigrationProgramconsistsofthe Familyandthe Skillstreams.TheFamilyStream allowstheimmigration ofimmediatefamilymembersofAustralian citizens, permanent residentsoreligibleNewZealand citizens.The SkilledStream bringsmigrantstoAustraliawhomhaveskillsassessedasbeingneededbythe Australian economy.AproportionoftheFamilyStreamimmigrants aresponsoredtocome to Australiabypeoplewho haveoriginally cometoAustraliaundertheHumanitarianProgram.Whilethesegroups ofmigrants are technicallynot classifiedas beinghumanitarianmigrants,manyofthemarelikelytohaveexperiencesand characteristicsnot dissimilar to those of humanitarianmigrants.
Table2providesinformation onthenumberofmigrants comingtoAustraliabyimmigrationpathway overthe period2001‐02to2010‐11.In2010‐11,13,799visasweregrantedundertheHumanitarianProgram,ofwhich
8,971weregrantedoffshoreand4,828onshore.Oftheoffshorevisasgranted,5,998weregrantedRefugee
6 TheinformationinthesectionissourcedfromtheDIACwebsiteandCommonwealthParliamentaryLibrary
Background Notes.
7AsylumseekerswhoareinAustraliaaregrantedpermanentProtection(classXA)(Subclass866)visasiftheyare owedprotectionunderAustralianmigrationlaw,becausetheyhavebeenfoundtoberefugeesandsatisfied health, characterand securityrequirements.
visasand2,973wereSpecialHumanitarianvisas.Thenumberofpeople comingtoAustraliaundertheFamily andSkillstreamsismuchlargerthanundertheHumanitarianProgram. In2010‐11 therewere113,725visas grantedunderthe Skill Stream and 54,543visas were granted under theFamily Stream.
TABLE2.NUMBEROFMIGRANTSTOAUSTRALIABYIMMIGRATIONPATHWAY,2001‐02TO2010‐11
HumanitarianProgram / FamilyandSkilledMigrationProgram / Total
immigration
Refugee / Special
Humanitarian / Onshore
Protection / Total
Humanitarian / Family / Skilled / Total
2001‐02 / 4,160 / 4,258 / 3,891 / 12,349 / 38,090 / 53,520 / 105,429
2002‐03 / 4,376 / 7,280 / 869 / 12,525 / 40,790 / 66,050 / 120,595
2003‐04 / 4,134 / 7,669 / 2,020 / 13,823 / 42,230 / 71,240 / 128,183
2004‐05 / 5,511 / 6,585 / 1,082 / 13,178 / 41,740 / 77,880 / 133,238
2005‐06 / 6,022 / 6,736 / 1,386 / 14,144 / 45,290 / 97,340 / 157,074
2006‐07 / 6,003 / 5,313 / 1,701 / 13,017 / 50,080 / 97,920 / 161,217
2007‐08 / 6,004 / 5,110 / 1,900 / 13,014 / 49,870 / 108,540 / 171,644
2008‐09 / 6,499 / 4,630 / 2,378 / 13,507 / 56,366 / 114,777 / 184,825
2009‐10 / 6,003 / 3,233 / 4,534 / 13,770 / 60,254 / 107,868 / 182,393
2010‐11 / 5,998 / 2,973 / 4,828 / 13,799 / 54,543 / 113,725 / 182,484
Sources: VariousDIAC statistical reportsand Karlsen,Phillipsand Koleth (2011).
Since 2001‐02thenumberof humanitarianarrivalshasremainedrelativelyconstant,althoughtherehasbeen somechangeinthecomposition,withthe numberscomingundertheSpecialHumanitariangroupfallingand thenumbersintheonshore protectiongrouphavingincreased(althoughthisgroupvariesfromyear‐to‐year dependinguponthenumber ofarrivals).IncontrastthenumberscomingundertheFamilyandSkilledstreams haveincreasedsubstantiallysince2001‐02.
Table3showsthecountryofbirthbycategoryofvisaforhumanitarianmigrantsin2010‐11forthe20 countries ofbirthfrom whichthe largest number ofimmigrants originated. Overall,these 20 countries accountedfor 88%ofhumanitarianmigrantswiththe fivecountriesfromwhichthemosthumanitarian migrants comefrom accountingfor58% ofhumanitarian migrants. While there are some differencesaccording totypeofhumanitarianvisa,theoverallpatternissimilar.Foroffshorerefugeesthelargestcountryof birthis IraqfollowedbyBurmaandBhutan.FortheSpecial HumanitariangroupthemaincountriesareIraq, AfghanistanandIran,andforthe onshoregroup, the mostcommoncountryof birthis Afghanistanfollowedby IranandIraq. Althoughthesemaincountriesofbirthmaydifferatthetimethesampleforthesurveyis selected,butthebroadpointthatmostofthehumanitarianmigrants arecomingfromasmallnumberof countries hashistoricallybeen the case andis likelytoremainthecase.
TABLE3.COUNTRYOFBIRTHBYCATEGORYOFVISA,HUMANITARIANARRIVALS,2010‐11
Offshore / OnshoreRefugee / Special
Humanitarian / Permanent
Protection
(Onshore) / Total
Iraq / 1,159 / 1,112 / 459 / 2.730
Afghanistan / 327 / 589 / 1,360 / 2,276
Iran / 189 / 275 / 925 / 1,389
Burma(Myanmar) / 874 / 40 / 110 / 1,024
Bhutan / 742 / 742
Sri Lanka / 73 / 212 / 355 / 640
China (excludesSARsand Taiwan) / 143 / 285 / 428
Pakistan / 14 / 163 / 226 / 403
Nepal / 364 / 2 / 34 / 400
Congo, DemocraticRepublicof / 329 / 45 / 374
Thailand / 326 / 2 / 2 / 330
Ethiopia / 208 / 74 / 19 / 301
Sudan / 88 / 141 / 1 / 230
Laos / 200 / 200
Zimbabwe / 2 / 15 / 164 / 181
Egypt / 11 / 27 / 136 / 174
Somalia / 109 / 35 / 2 / 146
Kuwait / 1 / 40 / 88 / 129
Kenya / 75 / 27 / 4 / 106
Eritrea / 89 / 2 / 12 / 103
Other / 605 / 401 / 664 / 1,670
Total / 5,785 / 3,345 / 4,846 / 13,976
Source: DIAC Administrative data.
Justoverhalf(58%)ofhumanitarianmigrantsaremale. Oftheprimaryapplicants,73%aremaleand27%are female.Bycontrast,56%ofsecondaryapplicantsarefemaleand44% male.Whenconsideringthesample frameforthesurveyitisimportanttotakeintoaccounttheagedistributionofhumanitarianarrivals,given thatitisnotenvisagedthatchildrenwillbeinterviewed.Amongsthumanitarianimmigrantswhoarrivedin
2010‐11 a littleover a quarterwere agedless than 15‐years (Table4).
TABLE4.AGEDISTRIBUTION OFHUMANITARIANIMMIGRANTS(PRIMARYAND
SECONDARYAPPLICANTS),ARRIVEDIN2010‐11
Number / Distribution(%)0‐4years / 1,200 / 9
5‐14years / 2,530 / 18
15‐24years / 3,110 / 22
25‐34years / 3,393 / 24
35‐44years / 1,998 / 14
45‐54years / 983 / 7
55‐64years / 462 / 3
65+years / 300 / 2
Total / 13,976 / 100
Source: DIAC Administrative data.
Consideration needsalsotobegiventotheminimumageforrespondents.Whilechildrenundertheageof15 yearsareclearlyout‐of‐scope asrespondentsforthissurvey,itmaybepossibletointerviewthoseaged15‐17 years.8 If15‐17yearoldsaretobeincludedinthesample,thencarefulconsiderationneedstobegivento consent issues and whether there are cultural differences regarding the appropriateness of interviewing
peopleinthis agerange. Otherlongitudinalstudieshavediffered intheminimumageofrespondents.The LongitudinalSurveyofImmigrantstoCanadaandtheLSIA1andLSIA2includemigrants(includingrefugees) aged15‐years plusat the timeofarrival.Bycontrast,theSurvey of New Refugees:ALongitudinalStudyof RefugeeIntegration intheUKis restrictedtorefugees aged18 years or older.
Table5shows the geographicareasinwhichhumanitarian arrivalsfromtheperiodNovember2010to October
2011wereliving. Immigrants arrivingundertheHumanitarianProgramareconcentratedinparticular geographicareasthataremainlyincapitalcities.Almostfouroutoffive(78%)ofhumanitarianimmigrants whoarrivedovertheperiodNovember2010toOctober2011werelivinginMelbourne(3,442),Sydney (3,020),Adelaide(1,720),Brisbane(1,273) andPerth(965)in2011.Thenumberslivingoutsideofthelarger capital citiesarequitesmall.Map1showsthecitiesortownwiththelargestnumber ofhumanitarianarrivals whoarrivedduring2010‐11. Thesefigures indicatethe totalnumber ofhumanitarianarrivalsinclusiveof children.
8Anumberofhouseholdlongitudinalstudiesincludeallhouseholdmembersincludingchildrenaspartofthe studybutonlyconductinterviewswiththoseaboveaparticularage(oftenfromtheageof15‐years). Whle longitudinalstudiesdoincludechildrenagedlessthan15‐years asrespondents,specialisedmethodsandsurvey techniquesarerequired,particularlyforchildrenwhomayhaveexperiencedverystressfulsituationandhave experiencedviolenceand trauma.
TABLE5.GEOGRAPHIC LOCATIONOFHUMANITARIANARRIVALS,2010‐11
Number / Distribution(%)Sydney / 3,020 / 22.6
Restof New SouthWales / 367 / 2.7
Melbourne / 3,442 / 25.8
Restof Victoria / 495 / 3.6
Brisbane / 1,273 / 9.5
Restof Queensland / 702 / 5.3
Adelaide / 1,720 / 12.9
Restof SouthAustralia / 186 / 1.4
Perth / 965 / 7.2
Restof Western Australia / 107 / 0.8
Tasmania,NorthernTerritoryandthe AustralianCapitalTerritory / 1,084 / 8.1
Total / 13,361 / 100.0
Note:ThisdataisfortheperiodNovember 2010‐October2011.Columntotalmaynotsumexactlyto100.0dueto
rounding errors.
Source:Cully(2011);FiguresfromDIACadministrative data.
MAP1.GEOGRAPHIC DISTRIBUTIONOFHUMANITARIANARRIVALSIN2010‐11
Note:Thenumberofhumanitarianarrivalslivinginareasoutsideofthecitiesandtownsshownonthemapis
1,172.
Source:Cully(2011);FiguresfromDIACadministrative data.
4.DESIGNOFTHESURVEY
Thetermsofreferencefor thispaperare thatthesurveyneedstoprovideinformationonthesettlement outcomesofhumanitarianmigrantsand possiblyothervulnerablemigrants.Asoutlinedinthissection,the populationof humanitarianmigrantscanbereadilydefinedonthebasisofvisatype.Itishowevermuchmore difficult to identifyother migrants orgroups of migrantswho are vulnerable.
Thissectionwillfirstconsidertheissueofincludinganon‐humanitariancohortofmigrants inthestudybefore movingon to considerthe morespecificdetailsofthe surveydesign. Thisincludes coverageofthe keyissues warranting considerationincludingsampleframe,geographicscope,datacollection methods and timingof interviews. Where appropriate,a recommendationhasbeenputforward ineachcase.
4.1NON‐HUMANITARIANMIGRANTS
Ifseekingtorecruitacohort ofnon‐humanitarianvulnerablemigrants,thereappearstobetwoapproachesto obtainingarepresentativesampleofsuchpersons.Thefirstistoidentifythecharacteristicsofmigrantsthat areknowntoincreasevulnerabilityinAustraliaandthentoconstructthesampleusingcriteriabasedupon these characteristics. The difficulty with this approach is that it requires vulnerability to be accurately identifiedatthepointofdefiningthesample.Thedataresultingfrom suchanapproachcouldbe usedto analysetheexperiencesand outcomesofa groupofmigrantswhohave characteristicswhicharebelieved to makethemvulnerable,butwouldnotallow ananalysisof theextentofvulnerabilityamongstthe groupmore broadly. Thesecondapproachisto includeallrecent migrants (humanitarianandother)inthe surveyandthen to use the surveydata to identify those whoare vulnerablemigrants.
Giventhedifficultiesoutlinedhere,thelimitedavailablebudget,andthatthepriority ofDIACforthissurveyis to understandingthe settlement experiences andoutcomes of humanitarianimmigrants,it is recommended thatageneral sampleofnon‐humanitarianvulnerablemigrantsnotberecruited.However,onegrouparriving undertheFamily stream that may provide a usefulcomparison group arethosewho are sponsored by someonewho themselvescametoAustraliaasahumanitarianmigrant. Itisthussuggestedthat thisgroup be includedinthesamplesincetheyarelikelytohavepre‐settlementexperiencesthatare similartothoseofthe humanitarianimmigrantwho sponsoredthem.Giventhesepoints,itissuggestedthat thepopulationforthe surveybemigrantswhohavebeengrantedaPermanentProtectionVisa9andFamilyReunionstreammigrants who are sponsoredbyahumanitarianmigrant.
4.2METHODOFDATACOLLECTION
Itisstronglyrecommendedthatthe BuildingaNewLifeinAustraliaSurveybeconductedlargelyviaface‐to‐ faceinterviews.Thereareanumberofdistinctanddifferentreasonssupportingthisapproach,eachofwhich is outlined below.
Aprimaryreasonisthegreater capacityto buildrapport withparticipants andthusretaintheirinvolvementin thestudyforalongerperiodoftime.Indeed,virtuallyallofthelong‐runningmajornationallongitudinal studies which havehighretentionrates ofthesampleover timehave themajority ofthedatacollection conducted face‐to‐face. It isthus atried andtestedapproach which lends itselfto this typeof study.
9Until late 2011 bridging visas were uncommon for protection visa applicants. However, policy changes announcedon 25November2011meanthat IrregularMaritimeArrivalswillnowbe releasedinto the community onbridgingvisaspriortoadecisionontheirprotectionvisaapplication.Asylumseekersgrantedbridgingvisas willbe allowedtoundertakepaidemployment,willhaveaccesstohealthservicesand maybe eligible forrefugee supportservices.Adecisionwillneedtobemadeastowhethertoincludethisgroupinthesamplingframefor thesurvey. Totheextenttowhichmanyofthosegranted aBridgingVisaarelikelytobesubsequentlygranted a Permanent ProtectionVisathenitwouldbedesirabletobegincollectingdatafromthemshortlyafterbeing released intothecommunitysincethisthesettlementprocessislikelytostartoncetheyarereleasedintothe community.
Additionally,for thisstudy, manyrecent humanitarianmigrants have limitedor noEnglishlanguageskills, limitededucationandlowlevelsofliteracy,andthus difficultiesaremorelikelytoariseusingself‐complete or telephonemethods.Finally,thesensitivenatureofthetopicswhichareexpectedtobe coveredinthe survey also lends itselftowards face‐to‐faceapproaches.
Thismeansthatthecaseforconductedface‐to‐faceinterviewsisparticularlystrongforalongitudinalsurvey ofhumanitarianarrivalstoAustralia.However,withintheface‐to‐faceapproachthere canbeseveraldifferent administration approachesincludingpaper‐and‐pencilor Computer AssistedPersonal Interview (CAPI).Studies suchas theHouseholdIncomeandLabour DynamicsinAustralia(HILDA)surveyandtheLongitudinalStudyof AustralianChildren(LSAC)arenowusingtheCAPI approachand ithas been foundtoprovidea numberof benefits.In particulariteliminatesthe need forseparatedataentry,speeds upthedeliveryofdata,allowsthe use of dependent data inthe questionnaire(information provide inquestions is fed forwardinto subsequent questions), allowsforautomatic checkingofinconsistenciesinsurvey respondentsand makesiteasiertobuild skips into thequestionnaire.
4.3GEOGRAPHIC COVERAGEOFTHESAMPLE
Inprinciple,thebestsamplingapproachisthatallgeographicareasin whichmigrantsarelivingareincluded. Thisresultsin asamplethat providesthe mostrepresentativesurveyofthe population.However,in orderto containthecostsofcollectingthedata,mostAustralian longitudinalstudiesthatinvolveface‐to‐facedata collection,selecttheinitialsamplefromonlysomeareasofAustralia.Giventhatthisstudyalsohasalimited
budget,itisthereforerecommendedthatthesampleisclustered.10 Thecostsavingsfromclusteringthe
samplearisefromareductioninthenumberofareasthatinterviewersneedtotravel to.Itisalsocommonfor majorlongitudinal surveys to exclude veryremoteareas ofAustraliadue tothehighdata collectioncosts and thefactthatonlyasmallproportionof the Australian populationlivein theseareas. Examplesofstudiesthat havemadethis decision are the HILDAand LSAC surveys.
As outlinedin Section3, recent humanitarianarrivals tendto be clusteredwithinparticular geographic areas of Australia. This means thatinorder tocontaincosts the samplewillneedtobe restrictedtopeople livingin geographicareas inwhichlargernumbersofhumanitarian arrivalslive.Thereishoweveratrade‐off between thecostsavingsresultingfromcollectingdatafromasmallernumberof geographicareas,particularlyregional andremoteareas,andthepotentiallossofrepresentativenessofthesamplebygoingtoonlyalimited numberof areas.
Ifthe questionofwhetherthereare differencesinsettlement experiences andoutcomesbetween humanitarianarrivalswhoinitiallysettleinregionalandruralareasandthosewhosettleinmajorcitiesisa high prioritythenit willbe necessaryto includeseveral regional andruralareasinthesample. If onlyone or tworegionalandruralareas areincludedinthesamplethenitwillnotbepossibletodeterminewhetherany differencesinoutcomesare explainedbysettlinginaregionalcentreas comparedto amajorcity,orwhether anydifferencesareexplainedbyparticular featuresoftheregionalcentresthathappentohavebeenselected forthesurvey. Thesmallnumbersofhumanitarianarrivalsoutsideofthelarger capitalcities(Map1)means thatitwillbe necessarytogotoseveralregionalsitesinordertoachieveasampleof sufficientsize toallow statisticalcomparisonsbetweenhumanitarianimmigrantsinmajorcitiesandthoseinregionalareas.This wouldhoweveraddsignificantlytothesurveycosts.Therefore,given thebudgetfortheproject andthe geographicdistributionofrecenthumanitarianmigrants, we wouldrecommend trading‐offfewer sites for a larger total sample size.
10LSIA1andLSIA2coveredallofthestateandterritorycapitalsandotherselectedmajorurbanareaswhichare relativelyclosetothecapitalcities.Inaddition,inLSIA1Cairnswasincluded.Itisestimates thatLSIA1andLSIA2 representedapproximately96percentof thepopulation(Cobb‐Clark,2011).
4.4NUMBER AND TIMINGOFSURVEYWAVES
Longitudinalstudiesdifferinthelengthof timebetweeninterviews(waves). Indecidingupontheoptimal lengthoftimebetweeninterviewsthereareanumberofconsiderations.Themorerapidlythecircumstances ofthe populationchangethenthemore frequentlyinterviewsareneeded.Formoregeographicallymobile populations,morefrequentinterviewscanassistinreducingattritionduetonon‐contact.11Ontheotherhand
therearecostsassociatedwithconductinginterviewsmorefrequently. First,thereistheincreasedburdento those participatinginthe survey,whichcan leadto respondents droppingout andwhichundermines its value asalongitudinalsurvey.Second,thereisthefinancialcosttothedatacollection agencyofundertaking additionalwavesofdata collection, particularlyiftheinterviewsareconductedface‐to‐facebutalsoiftheyare
conducted over the telephone.12
Itisrecommendedthatthesurveyinvolveannualwavesofdatacollection.Thisrecommendationisbased uponthe followingconsiderations.TheinitialperiodfollowingsettlementinAustraliaistypicallyoneof significantadjustmentandchange.Itisthusimportanttocollectdata atrelativelyfrequentintervals.Another reasonforconductingannualdatacollectionisthatashorterperiod betweensurveywavesisgenerally associatedwithlowerratesofnotbeingabletolocaterespondentsbetweensurveywaves,andthuslower rates of sample attrition. This is likely to be particularly important for humanitarian migrants given the evidencethat refugeesoften havefrequent changesintheirlivingcircumstancesduringtheinitialsettlement period(BeerandFoley,2005).Itiswellknownthatthose whochangeaddressesoftengenerallyhavehigher attritionratesfrom longitudinalsurveys.
AccordingtotheDIACtimetableforthe survey,fivewavesofdata collectionwouldbeundertakenannually from2012to2016.Ifinterviewswereconductedannuallytherespondentswouldhavespentbetween49and
54monthsinAustralia,slightlylessthantherequired60months.Ifitisnecessarytohavedatacoveringthe fullfirst60monthsaftersettlementinAustraliainorderforthesurveytoaddressthekeypolicyquestions thereareseveraloptions.Oneoptionistoaddanadditionalwaveof data collection.Anotheroptionisto increasethe lengthoftime betweeninterviews.
Dependinguponthegeographicregionsinwhichinterviewsareconducted,there maybenochoicebutto conductinterviewsoveralonger timeperiodinorderto achievetherequiredsample size.Thereare however advantagesanddisadvantagesinhavingshorterorlongerfieldwork periodsforeachwavewhichneed consideration.Advantagesofhaving relativelyshort(2‐3month)fieldworkperiods include:
Potential efficiencies inthefieldwork thatresult from reducedcostsofmanagement;
Someadvantagesfromananalysispointofviewthatrelatetoremovingseasonalityasan issue;and
Reducingthe variabilityinlengthof timebetweenwaves for surveyrespondents. Advantagesofhavinga longerfieldwork periodinclude;
Being ableto obtain greaterlevels of community engagement; and
Potentiallyincreasing the initial responserate.
4.5SAMPLINGFRAMEAND RECRUITMENT
Thereareavarietyofways ofobtainingthesampleforalongitudinalsurveyof migrants.Byfarthebest approach isto usea pre‐existingsample framethat containsallor a high proportionof the populationof interest.Anotherpotential methodistoidentifypeoplewithinthesurveypopulationscopebyscreening households.However,because recently‐arrivedhumanitarianmigrants are onlya verysmall proportionof the population,thisapproachisnotfeasible.Othermethodssuchassnowballsamplingdonot produce representativesamplesofthepopulationofinterestandthereforecanonlybeusedtoproducefindingsfor theparticularsampleinterviewed.Theycannotbe usedtoproduceestimateswhicharerelatedtothebroader populationofinterest.
Infact,oneof themajorlimitationsofmuchoftheresearchonrefugeeshasbeenthelack ofa complete samplingframewhichcanbe usedtodraw arandom probabilitysample(Bloch,2007). Australiaisfortunatein havingwhatappearstobeagoodsamplingframeforthepurposesof thisparticularsurvey,theSettlement Database(SDB)heldbyDIAC whichcontainsarangeofinformationonmigrantstoAustraliaincludingcontact details.Thisdatabaseis generatedusinginformationfrom avarietyofsourcesincludingthe SettlementDetails form anddepartmental systems usedto process migration applications both inAustraliaandatoverseas posts. Address informationis alsocapturedwhenthe immigrantenrols for services such as Adult Migrant English Program(AMEP)classesin Australia.Addressdetailsarealsoupdatedonce amonthusingdata fromthe Department of Human Services. This provides up‐to‐date addresses based on client interactions with Centrelink and Medicare.
WhiletheSDBdoesnothavecurrentcontactinformationforallimmigrants,andthe accuracyofthecontact informationdeclines the longer the migrants havebeeninAustralia,consultations withDIAC officials suggest thatthe contactdetailsarereasonablegoodforrecentarrivals.Thisisinpartaconsequence ofthevast majority(over 90%)ofmigrantsregisteringwithMedicare,irrespectiveoftypeofvisa.WhileFamilyandSkill stream migrants generallyhavea104‐weekwaitingperiodbeforetheycanaccess Medicare,humanitarian arrivalsdonothaveanysuch waitingperiod.ThismeansthattheirMedicarecontactdetailsaremorelikelyto beup‐to‐date thanforsome other groups ofmigrants,andthusthe SDBismorelikelyto haveup‐to‐date contactdetailsforthisgroup,particularlyintheinitialperiodofsettlementinAustralia.Furtherinvestigation ofthe completenessandaccuracyof the contact informationin the SDB willneedto be undertaken duringthe developmentphaseof the survey.
Whiletherewill be missingor out‐of‐date address/contact informationfor somerecenthumanitarian arrivals, fromasamplerepresentativenessperspectivethekey questioniswhetherhavingout‐of‐dataaddress/contact informationis systematically relatedto the characteristics of the humanitarian arrivals. Whileit will almost certainlybe the casethatsomebiasesareintroducedintothesampleasa result ofthe missingcontact information,theextentofsuchbiasesarelikelytoberelativelysmall.Thisisanissuethatwillrequirefurther
investigation.13
Thesampleshouldbechosenrandomlyfromthein‐scopepopulationincludingallmembersofimmigrating familyunits.Whenmembersofafamilyunitapplyfor avisatogetheroneperson, themainapplicant,must meetthe visarequirements andthe otherapplicants mustbe membersofthefamilyunit. Thesurvey populationshouldincludeallmembersofimmigratingfamilyunitsandthusincludetheprimaryapplicantand
13OthermajorAustralianlongitudinalstudieshaveusedAustralianGovernmentdepartmentsadministrativedata setsasthesamplingframe andsourceofcontactinformationincludetheLongitudinalStudyofAustralian Children,whichusedMedicare dataasthesamplingframe,andJourneys Home:Alongitudinalstudyoffactors affectinghousing stability,whichused Centrelinkdataasthesamplingframe.
secondaryapplicants.This differsfromthesamplingmethodusedfor LSIA1,2and 3whichrestrictedthe samplelargelytoprimaryapplicants withlimitednumbersofsecondaryapplicants.However,the characteristics,settlementoutcomesandserviceneedsof primaryandsecondaryapplicantsmaydiffer.Given thatthe keyresearch andpolicyquestions ofthe surveyrelate toall humanitarianarrivals andnotjustprimary applicants,itisimportantthattherespondentsincludebothprimaryandsecondaryapplicants.Itshouldalso be notedthat members ofthe migratingfamilyunitmay notbe livingin the samehouseholdatthe timeofthe firstinterview.Thiscouldoccur foranumberofreasons including relationshipbreakdownor different generations of the familylivingseparately(or for a range of other reasons). It is recommendedthatan attempt be madeto includeall households resulting from themigration ofa familyunitinthesamplingframe,although this may bedifficulttoachieve inpractice.
Oftenthe first fewmonths aftersettlement inAustraliaare onesof greatchange.Itisthereforeimportantthat thesurveycollect data from peopleas soonas possibleaftertheir arrivalinAustralia.As noted above,most studieshaveconductedthefirstwaveofinterviewsaboutsixmonths afterarrivalinAustralia.Whilethe rationale for thisis notentirelyclear,itappears thataroundsix months is the periodwhichittakes many humanitarianmigranttodealwiththeinitial‘shock’ofsettlinginanewcountryandbereadytoparticipatein amajorlongitudinalsurvey. Itisthereforerecommendedthatthesamplebedrawnfromimmigrantswho arrivedinAustralia three tosix monthsbeforethe first wave of interviews.
Generally,whenadministrativedatais used as the samplingframe for asurvey,potential respondents aresent aletterbythedepartmentwhichisresponsibleforthe databasewhichadvisesthemthattheyhavebeen selected to take part in a surveyand that unless theycontact the department and advise otherwise (opt‐out) thattheirnameandcontactinformationto thefieldworkorganisationthatwillberunningthesurvey.Itis crucialthatan opt‐outratherthananopt‐inprocessbe used.Opt‐in processeswhichrequirea personto contactthedepartment and saythattheywishtoparticipateinthesurveyusuallyresultinnon‐representative samples. Ofcourse, informed consentto participate inthestudyis obtainedatthetimeof theinterview.