User's Guide to the Youth in Focus Data Version 1.0
Contents
1. were treated as opt outs and taken off the list of people whose contact details were provided to RMR RMR received a sample that contained 12 698 focal youths and 12 276 parents for details see Tables 3a and 3b However in some cases no phone number could be found for either member of the observational unit In addition a few people 11 focal youths and 11 parents called FaCSIA to opt out after their opt out deadline and were consequently removed from the sample by RMR As a result of this cleaning process the final clean sample downloaded into CATI for interviewing included 11 757 focal youths and 11 588 parents for details see Tables 3a and 3b 20 Table 3a Wave 1 sample and mailout structure for focal youth Stratification category A B C D E F Total Mailout 1 910 1 650 490 390 398 233 4 071 Mailout 2 910 1 650 490 390 398 233 4 071 Mailout 3 910 1 649 490 390 398 232 4 069 Mailout 4 303 550 163 130 133 78 1357 Total Centrelink sample a 3 033 5 499 1 633 1 300 1 327 776 13 568 Opt out and return to sender b 259 320 96 91 61 43 870 of the Centrelink sample 8 5 58 59 7 00 46 55 64 Sample provided to RMR a b 2 774 5 179 1 537 1 209 1 266 733 12 698 No phone number late opt out c 229 390 95 90 88 49 941 of the RMR sample 8 3 7 5 6 2 74 7 0 67 74 Clean sample a b c 2545 4 789 1 442 1 119 1 178 684 11 757 Table 3b Wave 1 sample and mailout structure for parents Stratification category A B C2. participation in parent committees Remedial courses attended by focal youth Current enrolment status of focal youth e Focal youth s health Asthma ADHD Depression diagnosis disabilities hospital visits weight problems Date and reason focal youth moved out of parent s guardian s house e Type of accommodation focal youth lives in now Amount of respondent s contact with focal youth and satisfaction with focal youth s environment Quality of respondent s relationship with focal youth Financial help to focal youth Teenager difficulties Section H General questions Attitudes on unemployment benefits Importance of own and parental education ambition and job for success in life Locus of control questions Section I Health Overall rating of own health Smoking and drinking habits Asthma depression disability diagnosis other health problems Height and weight Section Y Tracking information A2 Youth questionnaire Section A Identification Post code state Date of birth Gender Section B Personal background e Aboriginal Torres Strait Islander status Country of birth COB If COB not Australia o Date of first arrival in Australia 31 O Language spoken in the family while growing up Section C Current Household e Marital de facto status gender and age of partner Living in large group share accommodation Number of family
3. Haveman Robert Gary Sandelfur Barbara Wolfe and Andrea Voyer 2001 Inequality of Family and Community Characteristics in Relation to Children s Attainments A Review of Trends in Levels and Inequality in Characteristics and Attainments and Estimated Relationships between Them University of Wisconsin unpublished working paper Israel Mark and Michael Seeborg 1998 The Impact of Youth Characteristics and Experiences on Transitions Out of Poverty Journal of Socio Economics 27 6 pp 753 76 McCoull Frances and Jocelyn Pech 2000 Trans generational income support dependence in Australia early evidence in P Saunders ed Reforming the Australian welfare state Australian Institute of Family Studies Melbourne pp 91 111 Pech Jocelyn and Frances McCoull 1998 Transgenerational poverty and income support dependence in Australia work in progress Social Security Journal 1998 2 pp 167 82 2000 Transgenerational welfare dependence myths and realities Australian 36 Social Policy 2000 1 pp 43 68 Solon Gary Mary Corcoran Roger Gordon and Deborah Laren 1991 A Longitudinal Analysis of Sibling Correlations in Economic The Journal of Human Resources 26 3 pp 509 34 Wellman Ray 2005 Transgenerational Data Set User documentation version 1 0 16 11 2005 Customer Data Services Data Management Branch FaCSIA 37
4. administrative data identified any member of the observational unit as falling into one of the following categories e Their administrative record is marked Deny Access which is a facility available to people seeking additional protection of personal information held by Centrelink for instance where there has been or is a potential fear of domestic violence e They have indicated refusal to participate in any surveys e Focal youth is deceased 3 2 TDS2 and the Australian population The numbers of focal youth present in the administrative data can be compared against other available statistics on young people born between 1 October 1987 and 31 March 1988 Table 1 shows the number of youth records in the administrative Since the responsibility of providing the updated information lies with the users of Centrelink some personal information may not be complete or regularly updated Moreover there is no requirement for individuals to update their information once they stop receiving benefits TDS User guide written by Centrelink Wellman 2005 contains more details on information available in TDS2 10 In certain circumstances young adults over 15 years of age may receive an Independent Status and be considered independent from their parents for the purposes of Centrelink payments receipt database TDS2 as well as the number of births registered during the relevant months and the Census 2006 results The 2006
5. administrative data contain information about most payments from the Australian Government to Australian families It is believed to be nearly representative of the entire Australian population with the exception of families with very high income According to FaCSIA estimates approximately 85 per cent of families with children receive Family Tax Benefit and therefore appear in the administrative data A smaller number of families receive additional support primarily due to low income Disability status could be a reason for income support receipt unrelated to low income This support may take the form of unemployment 5 Tax benefits for dependent children can take the form of a tax deduction or of a bi weekly payment direct to families The vast majority of families in Australia who are eligible for Family Tax Benefit opt for the bi weekly payment benefits payments to single parents with low income payments to couples with young children and low income and a variety of less common payments Since the administrative data include families who received substantial income support as well as families who received only the tax benefit the administrative data provide an excellent basis for studying the correlation between parents and children s receipt of income support from a combined sample of recipients and non recipients The survey is designed to collect a variety of information from a random sample of families who have appeared in th
6. members focal youth currently lives with and their age gender and relationship to focal youth Living independently or with parents guardians parental figure if lives independently date and reason for moving out of the parents house Section D Childhood living arrangements Full single parent no parent family incidence Number of houses apartments lived in Section E Family background Mother s and father s country of birth Whether lived with either or both parents at age 14 e Employment status of each parent parent substitute at focal youth s age 14 Parental education attainment school and higher certificates degrees Quality of relationship with parents Section F School School enrolment status year type of school overall number of schools attended e Plans for completing school and continuing education Assessment of own performance at school overall mathematics English e Reasons for leaving school early Certificates attained as a result of studies Section G Post School student status and highest qualification Post school studies qualification full part time status certificates Plans for completing higher degrees Section H Employment and earnings Employment status past present Number of jobs Hours and earning in all jobs main job Contract and salary arrangements Occupation and career First job history age and hours Section I J
7. number being engaged reaching an answering machine or not answered the procedure involved a callback at a different time on the next day The individual was considered uncontactable after 10 consecutive callbacks that did not result in a contact Once any attempt was made to interview an individual in an observational unit all individuals in that observational unit were followed through until either e The individual was successfully interviewed e The individual clearly refused e A sufficient number of callbacks have been made at different times and on different days to satisfy the research team that all reasonable means to contact and interview an individual have been made 6 12 Interview length and outcomes of self completion questionnaire In wave 1 all interviews were conducted over the telephone using the computer assisted telephone interviewing CATI After completing the interview over the phone the youth respondents were offered a choice of doing the self completion questionnaire either on paper to be sent by mail or on line at a secure website administered by RMR Of all youth respondents 73 4 per cent filled out self completion questionnaires Of those 48 per cent were filled out on line and 52 per cent on paper For wave 1 first interview telephone questionnaires the average interview length was 30 minutes for a focal youth and 35 minutes for a parent The time required for filling out the self completion ques
8. original Centrelink sample minus those who chose to opt out those whose primary approach letter was returned to sender observational units that consisted of parents without youth and observational units for which no contact number could be found Another procedure specific to the wave 1 pilot was that not everybody who agreed to participate in the survey was actually interviewed The target numbers of interviews for both parents and youth around 250 were reached before the entire pilot sample was exhausted therefore in order to more accurately assess response rates RMR continued to approach individuals and record their refusal or consent to participate but did not actually interview them The Completes lines in Table B2 refer to the individuals who actually completed the interview as well as those who only agreed to be interviewed Table B2 Pilot response rates Stratification category A B C D E F Total Focal youth survey Paid respondents Clean sample size 97 97 99 102 97 101 593 Completes 39 32 36 45 45 39 236 Response rate 40 2 33 0 364 44 1 46 4 38 6 39 8 Unpaid respondents Clean sample size 97 104 104 95 94 103 597 Completes 39 32 41 34 27 24 197 Response rate 40 2 30 8 39 4 35 8 28 7 23 3 33 0 Difference in response rate with respondent payment 0 0 2 2 3 1 8 3 17 7 15 3 6 8 Overall youth response rate 40 2 31 8 37 9 40 1 37 7 30 9 36 4 Parent survey Paid respondents Clean samp
9. survey are overseas or institutionalised with no prospect of return until the end of the survey or do not speak English sufficiently well to participate b Individuals are declared uncontactable if the number listed for them is a business number if the named respondent is not at the number or if after a sufficient number of callbacks the call outcomes on all callbacks are engaged no reply answering machine fax or modem c Individuals are declared unobtainable after a sufficient number of callbacks if the call outcomes on all callbacks are carrier messages like this number is disconnected this number is uncontactable or abnormal dialing tones that are not faxes or modems 25 Table 4b Wave 1 parent outcomes Stratification category A B C D E F Total Completes 1 045 1 413 491 419 384 212 3 964 of clean strata sample 41 5 30 0 34 4 37 7 33 2 31 6 34 2 Refusals 741 1 281 428 350 376 212 3 388 of clean strata sample 29 5 27 2 30 0 31 5 32 5 31 6 29 2 Unable to participate 106 329 122 44 57 31 689 of clean strata sample 4 2 7 0 8 6 4 0 4 9 4 6 6 0 Uncontactable 366 899 221 178 211 125 2 000 of clean strata sample 14 6 19 1 15 5 16 0 18 2 18 6 17 3 Unobtainable 258 782 164 122 129 92 1 547 of clean strata sample 10 3 16 6 11 5 11 0 11 2 13 7 134 TOTAL CLEAN SAMPLE 2 516 4 704 1 426 1 113 1 157 672 11 588 Notes See Table 4a 6 19 Dat
10. 006 become available mid 2007 3 3 Currency of contact information Centrelink customers are required to provide their contact details when they first apply for payments and are required to inform Centrelink of any changes in their circumstances However since payments do not necessarily depend on the availability of current contact details payments usually being made directly into specified bank accounts customers may neglect to keep their records current Moreover individuals are not required to update personal records if they no longer receive any payment 11 See ABS 2002 and 2003 12 http www abs gov au websitedbs d3310114 nsf Home census In view of this individuals with short exposure to the income support system which took place a long time ago may be more difficult to locate than those with more recent exposure To compensate for this such individuals were over sampled see Section 4 In addition multiple approaches were employed to find missing contact details detailed in Section 6 4 Survey design 4 1 Data collection mode The primary data collection mode is computer assisted telephone interviewing CATI In addition to the questionnaire administered by telephone the focal youths are asked to complete a self completion questionnaire Youth may choose to complete this on line at a secure site or to complete a paper questionnaire which is returned by mail 4 2 Interview frequency Due to the longitudinal natur
11. Census data refer to the proportion of 263 222 18 year olds excluding overseas visitors located in Australia on the Census night of 8 August 2001 The applied proportion of 49 59 per cent is the number of births during October 1987 to March 1988 relative to the number of births which occurred during August 1987 to July 1988 as reported in ABS 2002 and 2003 Table 1 Number of young people in TDS2 and the Australian population Registered TDS2 births Census 2006 Born October December 1987 62 801 59 863 Born January March 1988 64 999 61 844 Total 127 800 121 707 130 532 Notes a The Census data show the proportion of all 18 year olds excluding overseas visitors located in Australia on the Census night of 8 August 2006 Since the Census data do not provide information on the month of birth the applied proportion of 49 59 per cent is derived based on the proportion of births in the birth cohort months October 1987 to March 1988 relative to births occurred during August 1987 to July 1988 Sources ABS 2002 and 2003 Census 2006 available from ABS website at http www abs gov au Table 1 demonstrates that the administrative dataset contains records on nearly all young adults belonging to the birth cohort Neither of the two sources used for comparison provides up to date information since they do not take into account migration and deaths We anticipate that Table 1 will be updated as the results of Australian Census 2
12. D E F Total Mailout 1 899 1 633 484 387 395 231 4 029 Mailout 2 901 1 634 488 388 397 233 4 041 Mailout 3 893 1 629 484 387 391 231 4 015 Mailout 4 300 544 160 127 132 76 1 339 Total Centrelink sample a 2 993 5 440 1 616 1 289 1 315 771 13 424 Opt out and return to sender b 310 442 127 111 90 68 1 148 of the Centrelink sample 10 4 81 7 9 86 68 88 8 6 Sample provided to RMR a b 2 683 4 998 1 489 1 178 1 225 703 12 276 No phone number late opt out c 167 294 63 65 68 31 688 of the RMR sample 6 2 59 42 55 56 44 5 6 Clean sample a b c 2516 4 704 1 426 1 113 1 157 672 11 588 Although as explained above all individuals whose primary approach letter was sent back as return to sender were removed from the sample a certain number of respondents when approached by the market research firm could not remember receiving the survey invitation and brochure and requested that these materials be re sent prior to the interview For these cases RMR compiled weekly lists that were provided to FaCSIA who then sent another copy of the primary approach letter and brochure to the participants specified current addresses Subsequently the responsibility for re sending the letters 21 and brochures was transferred to RMR In the last week of the telephone interviewing due to time constraints the respondents who requested another copy of the survey materials were offered the YIF survey website address where the bro
13. a linkage After completing the telephone survey the respondents were asked for their consent to link their survey answers to the administrative data The overwhelming majority of respondents both in the pilot and the main wave had agreed to the linking Table 5 below details percentages of respondents who consented to the data linkage Table 5 Respondents consent to data linkage per cent of total Stratification A B C D E F Total category Pilot Focal youth 92 86 92 11 97 22 95 92 100 00 96 30 95 45 Parents 92 98 100 00 94 00 90 24 94 44 96 88 94 51 Main fieldwork Focal youth 97 47 95 45 95 44 96 19 96 77 96 10 96 20 Parents 93 49 91 15 91 85 93 08 94 53 92 45 92 46 6 20 Tracking and tracing In order to minimise attrition between waves both the pilot and the main wave respondents are tracked using several step follow up procedure It is planned that 26 the follow up of respondents will be done at six monthly intervals The timetable for the pilot keep in touch strategy is as follows Wave 1 pilot March April 2006 Mailout of Thank you card July 17 2006 Mailout of Newsletter December 11 2006 Phone contact October 2007 Wave 2 pilot tentative March April 2008 The proposed timetable for the wave 1 keep in touch strategy is as follows Wave 1 fieldwork August December 2006 Mailout of Thank you card May June 2007 Phone contact September October 2007 Mailout of Newsletter March April 2008 Wave 2 field
14. a will be linked to the administrative data and the linked data will be the basis for studying the causal mechanisms 1 5 Data management access and security Due to the sensitivity of the confidential data contained in the administrative dataset its access at the ANU is restricted to a secure data room The secure data room and procedures for using the data in it incorporate a range of measures designed to protect the data This room and the procedures are regularly audited and inspected It is anticipated that the survey data will be made accessible in the future to other interested researchers together with a summary of information from the administrative dataset The main management of the project is done through the project steering committee This committee meets several times per year and is composed of members of the 6 See Centrelink 2007 project research team and representatives from FaCSIA This committee is responsible for dissemination of the data designing conditions of use for the data and approving applications to use the data 2 Reference populations and project terminology 2 1 Birth cohort and focal youth The primary reference population for the main wave of the YIF survey is all people born between 1 October 1987 and 31 March 1988 who appear in the Centrelink administrative data at any point between 1991 and July 2006 We will refer to this group as the main birth cohort In addition there is a smaller g
15. and the ANU was signed on 5 December 2005 6 2 Pre test A small pre test was conducted in October 2005 by RMR and the project research team with several paid participants to provide preliminary feedback on all survey instruments As the main objective of the pre test was to judge participants reaction to the questions asked their answers were not recorded 6 3 Interviewer training In order to familiarise interviewers with the survey instruments and provide them with the background information on the YIF project interviewer information sessions led by the RMR project manager for YIF were run immediately prior to the pilot and the main wave of interviewing The pre pilot information session was held on 27 March 2005 and was attended on the ANU side by Drs Robert Breunig and Tue Gergens Another session was held on 15 August 2006 immediately before the main fieldwork for wave 1 and was attended by Dr Robert Breunig Professor Cobb Clark and Dr Anastasia Sartbayeva as well as the representatives of the FaCSIA Youth Bureau and the Research and Analysis Branch 6 4 Wave 1 pilot A pilot test was carried out between 27 March and 18 April 2006 The sample for the pilot consisted of observational units as described in Section 2 3 with the focal youths randomly selected from the pilot birth cohort To test the differences in response rates across stratification categories the pilot sample contained observational units from the six stra
16. aper provides important background information regarding the Youth in Focus project and the associated survey and administrative data We outline the motivation behind the project the primary research questions that the project is intended to address the funding sources and the composition of the project research team Furthermore we provide detailed information on the methodology of the project This includes the data sources the survey sampling frame the sampling methodology a brief description of the survey instruments and the proposed timeline of fieldwork We describe the pilot study that was conducted the use of respondent payments in the project the survey outcomes from the pilot and the first wave of data collection 1 Introduction 1 1 Project organisation and partners Youth in Focus YIF is a joint research project between the Australian Government Department of Families Community Services and Indigenous Affairs FaCSIA and a team of academic researchers The research team includes Professor Deborah Cobb Clark Dr Robert Breunig and Dr Tue Gergens of the Australian National University ANU Professor Jeff Borland of the University of Melbourne Professors Barbara Wolfe and Robert Haveman of the University of Wisconsin Madison and Ms Jocelyn Pech of the Australian Fair Pay Commission Secretariat previously with the Department of Family and Community Services The YIF project is supported by a five year Linkage Projec
17. ary approach letter and the first contact by RMR the mailouts were conducted approximately 2 weeks apart 14 July 28 July 11 August and 1 September 2006 the first three mailouts each containing 30 per cent of the total sample and the last one containing 10 per cent The opt out process was then conducted separately for each mailout stage Overall 13 568 young people born between 1 October 1987 and 31 March 1988 and their parent were drawn into the sample for wave 1 by Centrelink As outlined in the previous section the numbers of focal youth in each stratum of this sample were determined based on the pilot response and opt out rates The first five rows of Tables 3a and 3b contain the detailed information on the numbers of focal youth and parents in each mailout FaCSIA maintained a 1800 number throughout the duration of the opt out period so that recipients of the primary approach letters could contact FaCSIA to opt out or raise any queries about the survey From the date the first mailout sample was provided to RMR a second 1800 number was set up staffed by the market research firm to liaise with respondents As explained in Section 3 3 the accuracy of the mailing addresses of potential respondents could not be easily ensured especially for those who have last received Centrelink benefits a significant time ago Therefore the intended recipients of all primary approach letters that were returned to FaCSIA marked return to sender
18. ch may also result from a history of dependence on income support Previous studies have faced two major empirical challenges Most studies have been based upon a population of disadvantaged individuals receiving support from a particular government program These studies have not had a comparison group of program non participants to provide a compelling counterfactual to answer questions such as does dropping out of school early make youth in welfare dependent families more likely to access social assistance as adults than youths from non welfare dependent families who drop out of school early The second difficulty that previous studies have faced is the lack of complete administrative data about program participation These studies have had to rely on individuals own reports of income support receipt and have usually compared two particular points in time one during an individual s youth and one in adulthood In addition to raising measurement issues that restrict the possible observable associations and complicate interpretation 4 such studies are unable to say anything about the importance of timing or intensity of exposure to social assistance This project is designed to address and overcome the weaknesses of previous studies By using a combination of administrative data going back to 1991 and survey data gathered from both parents and children issues of timing intensity and incidence of disadvantage can be studied while contr
19. chure could be viewed and the opportunity to receive the paper copy after the interview If the respondents did not agree to this arrangement the interview was terminated and they were classified as survey refusals 6 11 Wave 1 data gathering procedure RMR made a substantial effort to contact each observational unit selected for the sample This involved e Finding a telephone number through both automated and manual white pages search using the respondent s last name e Finding a telephone number for individuals within an observational unit by asking the other observational unit member RMR in collaboration with the research team also developed protocol for declaring observational units or individuals within the observational units non contactable e If no phone number was available for either member of the observational unit and the white pages search was unsuccessful the unit was dropped from the clean sample Similarly for units consisting only of a focal youth where no phone number could be found e If only one member of the observational unit had a phone number and the automated white pages search for the other person was unsuccessful the existing number was used to get into contact with the other member of the unit e Similarly if only one member of the observational unit had a phone number and the number for the other person was found incorrect invalid then the valid number was used to get into contact with the oth
20. d by the interviewers survey participants were referred to the interviewer supervisors followed if necessary by the project manager at the market research firm or the designated FaCSIA representative 6 18 Wave 1 contact outcomes A summary of the outcomes of the wave 1 fieldwork is provided in Tables 4a and 4b Table 4a Wave 1 focal youth outcomes Stratification category A B C D E F Total Per cent of complete surveys across strata 25 2 361 129 10 3 99 57 Completes 1 027 1 472 526 420 404 231 4 079 of clean strata sample 40 4 30 7 36 5 37 5 34 2 33 8 34 7 Self completion questionnaire 840 1009 379 322 281 154 2985 online questionnaire 509 433 209 173 163 84 1571 mail questionnaire 331 576 170 149 u8 7 mm Refusals 586 1 017 323 256 278 137 2 597 of clean strata sample 230 21 2 224 229 236 20 0 221 Unable to participate 103 185 61 30 45 24 448 of clean strata sample 4 1 39 42 27 38 35 38 Uncontactable 631 1 480 378 297 339 213 3 338 of clean strata sample 24 8 30 9 26 2 26 5 28 8 31 1 28 4 Unobtainable 198 635 154 116 113 79 1 295 of clean strata sample 7 8 13 3 10 7 10 4 96 116 11 0 TOTAL CLEAN SAMPLE 2 545 4 789 1 442 1 119 1 178 684 11 757 Notes a Unable to participate category includes individuals who are out of scope and observational units where the youth is out of scope people who are unwell with no prospect of recovery until the end of the
21. e administrative data at least once since 1991 The survey design is based around a birth cohort of youth who turned 18 just before wave 1 interviewing see Section 2 1 below For each youth we identified all individuals who received any type of payment on behalf of that youth in the history of the administrative data From these individuals who we henceforth refer to as parents we selected the person who had the longest duration of care as measured by receiving a payment on behalf of the youth or claiming the youth as a dependent over the history of the administrative data set We further implemented a set of rules designed to identify the natural mother from the administrative data Although the administrative data do not contain information on the actual family relationships the chosen method proved extremely successful in identifying the natural parents Among wave 1 respondents a natural mother was selected in 96 5 per cent and a natural parent in 98 6 per cent of cases The parents were interviewed once and at the time of writing we anticipate that the children will be interviewed approximately annually for three waves Respondents were asked to provide information on topics such as employment education physical and mental health attitudes and values family relationships and other psycho social factors the children s experiences while growing up neighbourhood and school quality With the respondents consent the survey dat
22. e of the YIF project the focal youth are expected to be interviewed at three waves conducted approximately two years apart Thus it is anticipated that the main wave youth participants will be interviewed again in September December of 2008 and 2010 and the pilot youth participants will be interviewed in March April of 2008 and 2010 4 3 Payment to respondents Based upon a test of payment to respondents in the pilot see Section 6 it was decided to pay a small monetary compensation to all participants Each participant who completed the interview received a cheque for 15 In the case of focal youth they had also to complete and return the self completion questionnaire It is important to note that the payment was not based upon responses to any particular question In fact a participant could agree to participate but subsequently refuse to answer every question and the compensation would still be paid 4 4 Sample stratification The sample of focal youths and their parents guardians was selected as a probability sample stratified on the basis of the income support history of the parent as described in Table 2 The no parental income support history category combines those observational units where there is no parent record along with those the vast majority where the parent has received only Family Tax Benefit since 1993 Each focal youth will thus belong to one of six stratification categories shown in Table 2 Table 2 Pro
23. ents and important life events The summary of topics covered by the survey instruments is provided in Appendix A The full versions of the survey questionnaires are available from the research team upon request Where possible wording from other Australian surveys was used in order to facilitate comparison of the data collected Standard Australian Bureau of Statistics questions for items such as Indigenous status were employed Questions relating to income and labour status were made similar to those in HILDA for comparability 6 Fieldwork 6 1 Contracting of market research firm A competitive tender process was undertaken to find a suitable organisation to undertake the fieldwork It was anticipated that the chosen firm would assist in finalising survey instruments designing and implementing a pre test and designing and implementing all waves of data gathering An information session was conducted at the Australian National University on 2 May 2005 for interested companies prior to the tender deadline This was attended by representatives of six different companies A phone information session was conducted for two companies which could not attend the information session Five proposals were received From these Roy Morgan Research RMR was selected as having the proposal which met the minimum requirements of the tender and which additionally offered the best value for money The contract for provision of field services between RMR
24. er member of the unit e If both members of the unit had incorrect invalid numbers the automated white pages search for both names was run at periodic intervals Some observational units provided in the sample to RMR were out of scope for other reasons An observational unit was considered to be out of scope if the focal youth was Out of scope Any individual a focal youth or a parent was considered out of scope if the individual was e Deceased e Overseas with no prospect of return before the end of the survey August 2008 e Institutionalised with no prospect of release before the end of the survey Contact details for individuals in institutions for example those confined to a correctional facility or medical treatment facility were kept and these individuals 22 were included in the sample This should enable normal interviewing to resume if and when those individuals return to living in private dwellings Similarly while sample members overseas were not contacted the contact details for individuals who moved overseas were maintained so that if they move back to Australia they can be interviewed in subsequent waves RMR in consultation with the research team developed a protocol for converting contacts with individuals to successful interviews with the entire observational unit This protocol included procedures for contacting people at different times and on different days of the week If a contact resulted in the phone
25. ers and Dr Helen Berry of the National Centre for Epidemiology and Population Health at the ANU Professor Peter Lynn of the Institute for Social and Economic Research at the University of Essex UK also provided helpful advice The survey instruments were put into operation with the help of Roy Morgan Research Pty Ltd the market research firm contracted to carry out the fieldwork This involved pre testing the questionnaires with several paid participants in October 2005 intensive practice runs of the survey instruments by the research team via a telnet access system to the CATI questionnaire and a full scale pilot of wave 1 8 Living in Australia The Household Income and Labour Dynamics in Australia Survey more information is available at http www melbourneinstitute com hilda of the survey in March April 2006 Three instruments were used in the YIF survey e Parent questionnaire which collected information on the parent s family education employment and income relationship history general values and health as well as background information about the focal youth at early stages of life e Youth questionnaire which collected information on the focal youth s household and family education employment and job search housing arrangements income and health as well as personality traits and life satisfaction e Self completion questionnaire for youth which collected information on attitudes substance abuse relationships with par
26. h CATI and self completion questionnaire responses Thus after the completion of the pilot and main fieldwork for wave 1 the ANU has four survey datafiles Table 7 outlines the existing datafiles and the numbers of variables and records available in each Note that for the focal youth who completed the telephone interview but not the self completion questionnaire the variables corresponding to the self completion questionnaire in the focal youth datafiles would contain missing data Table 7 Survey datafiles characteristics Dataset No of variables No of records Parent wave 1 pilot datafile 573 255 Focal youth wave 1 pilot datafile both telephone and self completion surveys 699 242 Parent wave 1 datafile 590 3 964 Focal youth wave 1 datafile both telephone and self completion surveys 785 4 079 29 Appendix A YI F questionnaires summary A1 Parent questionnaire Identification Date of birth gender State and postcode Relationship to focal youth Marital status Section A Household information For each person living with age gender relationship to respondent identifier for the other natural parent of the focal youth Section B Personal information Country of birth COB If COB not Australia o Date of first arrival in Australia O Language spoken in the family while growing up o English knowledge Aboriginal Torres Strait Islander status Respondent s parents occupation
27. heir own right Centrelink payments 8 The first Transgenerational Dataset TDS1 is a database that was created by FaCSIA in 1996 and informed the construction of TDS2 such as Youth Allowance or disability payments or ii at some point since 1991 they were listed as a dependent of an adult who has received a government payment Any adult who claimed payments at any time for the focal youth is also included and is referred to as a parent As mentioned earlier the relationship between parent and the focal youth is an administrative one and thus the adult may not be a natural parent In addition the dataset includes siblings identified as other children for whom the parents of the focal youths also claimed payments Again these siblings may have no blood relation to the focal youth Sibling records are linked to the parent records and through the parent to the focal youth records For all people in the administrative dataset who received payments in their own right the following information is provided e Personal details date of birth sex marital status history country of birth e Housing details home postcode history accommodation history rent type e Youth homeless or independent history e Full time student history e Family income support history e Marital status history e Individual income support history payment types and dates of receipt The entire observational unit was eliminated from the frame if the
28. i nal unit iauna aaa Saas a nan a a A AE RENT 11 2A Data c lecionr uniana a E T EUN 11 3 Administrative data and sampling frame sesesesosesosososososesosesosososesosesosososososssssseses 11 3 1 The administrative database TDS2 cc ccesssssscsseceecsseceessceeeessecsaeceeeensenseeeeees 11 3 2 TDS2 and the AvistralianspopulauiOn ics tiesninariitecantnay binant aeitnnsninnd gt 12 3 3 Currency of contact information cxcciesnsssstbine i oatienastaelegsceasual acne 13 4 Survey design naiiai aea sadness sadasdiaedadvonsundad susvuabasscdiakdap a E ae 14 4 1 Data collection MOE a a N ia 14 4 2 Interview fregUency nisrani naana a a n an 14 43 Payment to respondents erences heime e a e T 14 BA Sample stratificatioti via saiisntsostiene eames E EEA EEEE AEREA 14 45 SAMple S enni eiei seater tection aE a AE A E TEES 16 4 6 Expected sample size in future waves s ssessssessesessesississesensesessesississesensesessesens 16 5 S tvey instruments issons cessive ues resorte eines Soana KANEEN Eea u KoE EEEN EEK aS 16 CCP ET e BAST A E a E TE E E E E E AE TAAA 17 6 1 Contracting of market research firm s ssssesssssereseresrsrssessststresesesrsrssestsrererenesesess 17 6 2 Pretest nnani en A Ta e a a a a a a aa aae 17 6 3 Interviewer tranne nae en ehana a Ea aE AEA 18 6 4 Wave T pilot rirorieeris aanroer iii a A OA AERA A T ERA E E EA O 18 6 5 Wave zand Waves pilotSimesreniesri ionita a aeaa E TAE EERE 18 6 6 Pilotop
29. in focus YOUTH IN FOCUS PROJECT DISCUSSION PAPER SERIES No 1 September 2007 User s Guide to the Youth in Focus Data Version 1 0 Robert Breunig Deborah Cobb Clark Tue Gorgens Anastasia Sartbayeva youthinfocus anu edu au in focus Jointly funded by the Australian Government Department of Families Community Services and Indigenous Affairs and the Australian Research Council and carried out by The Australian National University Youth in Focus Project Discussion Paper Series ISSN 1835 4025 Youth in Focus Research School of Social Sciences ANU College of Arts and Social Sciences Coombs Building 9 The Australian National University Canberra ACT 0200 Table of Contents OAA DAT A E T AE EEES EEEE EST ASEE 5 IA Bas woe LOTO Lo a PERETE E AE E TE AAE AET 6 1 1 Project organisation and partners ssssesssssseresesesrsrssesestsreresesesrsssstststerereresesrsreseses 6 K2 Project Mouvaton anaiei an a a A E AEA T I 6 1 3 Main research CCS 01 0 T PA E EE EAR aguas tuguabr sada bee iananisanisoad Meta 8 AR BEI ETON E iA E AEA E A A E ES A NA EA 8 1 5 Data management access and security s ssessssessesessesessesssessesessesissessesensesesresees 9 2 Reference populations and project terminology e seesseresrercereseeseseeresrercoresresersesesse 10 2 1 Birth cohort and LOCA ly OU Li aicis ion asiassiadeananttonasncnwiedsaion na a 10 DD AEN E E D AAA cia Beaver EEOAE A E E AE AE cite cele 10 23 Obs rvat
30. is not identifiable from the administrative data we will refer to these people as parents Due to the construction of the parent subset of the administrative data a focal youth may have several parents or guardians For the purposes of the survey only one parent was selected the criterion being the longest duration of care payment receipt for the focal youth By looking at the gender of parents and the timing of receipt from the administrative data we attempted to select the natural mother whenever 7 The two months were originally selected for use in the pilot to test for any differences in the contactability of focal youth The pilot has shown no significant difference in response rates possible As mentioned previously this methodology resulted in the natural mother being selected in 96 5 per cent and a natural parent being selected in 98 6 per cent of cases As with focal youth certain parents are considered out of scope for details see Section 6 11 2 3 Observational unit Most focal youths have a parent in the administrative dataset However in some cases there is no parent This may happen if the parent s are deceased or if the focal youth is in the administrative data as a recipient of Youth Allowance and his her parent s have never received any type of Centrelink payment In cases where both the focal youth and a parent are available in the administrative data the matched focal youth parent twosome will be refe
31. ix B2 were used to decide on wave 1 stratification as explained in the next section 6 9 Wave 1 sample On the basis of the opt out and response rates observed in the pilot test the research team determined the number of observational units to be drawn in each stratification category for the raw sample As in the pilot test the information provided for each observational unit included the names of the focal youth and the parent if there was a parent and for each of these individuals a phone number where available and address The stratum to which each observation belongs was also indicated by a stratum label A to F see Table 2 Strata information was not visible to interviewers 6 10 Wave 1 opt out process As with the pilot study FaCSIA sent a primary approach letter and information brochure to each member of the observational unit before the beginning of the wave 15 The letter and information brochure can be viewed at http youthinfocus anu edu au survey htm 1 interview period The letter contained the invitation to participate in the survey and explained how to opt out Since wave 1 and associated opt out and interviewing activities were carried out over the period of 5 months July to December 2006 it was decided that the sample for wave 1 should be drawn in four stages each stage subsequently representing a separate mailout in the survey opt out process In order to minimise the time span between the receipt of the prim
32. lative to the average for the population Another over sampled group is people who have had only a short and not recent exposure to the income support system because of the anticipated difficulty in locating them relative to those with more recent exposure see Section 3 3 As the survey is conducted over the phone it was decided not to stratify by geographic location 4 5 Sample size We selected a sample of 13 568 youths for wave 1 with the objective of having 4 000 completed interviews from youth 3 797 completed interviews from parents and 2 240 completed interviews from observational units 4 6 Expected sample size in future waves We expect attrition rates of 15 per cent between waves 1 and 2 and 10 per cent between waves 2 and 3 for youth We thus expect to have 3 460 completed interviews from youth in wave 2 and 3 115 completed interviews in wave 3 5 Survey instruments Development of the survey instruments for wave 1 was an iterative process involving discussion between the project research team FaCSIA staff the TDS2 development team external technical advisors and other interested agencies The teams working on the questionnaires drew on existing international and domestic surveys as well as expert advice from a range of individuals including Ms Nicole Watson and Professor Mark Wooden at the Melbourne Institute of Applied Economic and Social Research who have responsibilities for HILDA as well as Professor Bryan Rodg
33. le size 93 91 103 97 99 98 581 Completes 36 31 38 40 45 34 224 Response rate 38 7 34 1 36 9 41 2 45 5 34 7 38 6 Unpaid respondents Clean sample size 89 100 102 100 96 102 589 Completes 34 21 40 26 31 38 190 Response rate 38 2 21 0 39 2 26 0 32 3 37 3 32 3 Difference in response rate with respondent payment 0 5 13 1 2 3 15 2 13 2 2 6 6 3 Overall parent response rate 38 5 27 2 38 0 33 5 39 0 36 0 35 4 35 Bibliography An Chong Bum Donna Ginther Robert Haveman and Barbara Wolfe 1996 The Window Problem in Studies of Children s Attainments A Methodological Exploration Journal of the American Statistical Association 91 pp 970 82 Australian Bureau of Statistics ABS 2002 Births Australia 2002 http www abs gov au AUSSTATS abs nsf DetailsPage 3301 02002 OpenDoc ument 2003 Births Australia 2003 http www abs gov au AUSSTATS abs nsf DetailsPage 3301 02003 OpenDoc ument Centrelink 2007 A guide to Australian Government Payments http www centrelink gov au internet internet nsf filestores co029_0702 file co 029_0702en pdf Gottschalk Peter 1996 Is the Correlation in Welfare Participation Across Generations Spurious Journal of Public Economics 63 pp 1 25 Haveman Robert and Barbara Wolfe 1995 The Determinants of Children s Attainments A Review of Methods and Findings Journal of Economic Literature 33 4 December pp 1829 78
34. n category A to F identifier Pbmonth Parent survey only Parent month of birth Pbyear Parent survey only Parent year of birth Home_state Home state of respondent Home_post Postcode of respondent Ccount Call count total number of calls made to the respondent s telephone number Avint Time of the telephone interview Totaltime Total call time time of all telephone calls made to the respondent scrla youth screx parent Asks whether the respondent recalls receiving the survey invitation and information brochure scr1 scr2 Asks whether the respondent is willing to conduct the telephone interview and or arranges suitable appointment The rest of the variables in the survey datasets are named according to the following naming scheme wave questionnaire identifier section identifier question number Wave questionnaire identifier is represented by a single letter a through z and corresponds to a parent telephone questionnaire wave 1 pilot b youth telephone questionnaire wave 1 pilot c youth self completion questionnaire wave 1 pilot d parent telephone questionnaire wave 1 e youth telephone questionnaire wave 1 f youth self completion questionnaire wave 1 The wave questionnaire identifier is followed by a section identifier which is a single letter a through z The question number follows as it appears in questionnaire i e 24a If the question is repea
35. nds Quality of relationship with each parent substitute Education values Own friends and parents attitudes to education Life events and childhood Possible traumatic events that happened to focal youth Teenager problems and contact with authorities e Relationship with parents during childhood Overall rating of childhood 33 Appendix B Additional information on wave 1 pilot B1 I tems tested in pilot Survey respondent reaction to the primary approach letter and information leaflet For all of the survey instruments Average interview length of the survey instruments Routing procedures in the survey instruments so that all survey respondents were asked the questions appropriate to them The response categories for scalar questions so that survey respondents were distributed across categories Identify questions which generated high item non response rates Identify questions posing any other specific problem for survey respondents Identify categorical questions which generated a large number of other responses Whether or not useful information was being obtained from open ended questions Any other survey instrument problems or issues Procedures for obtaining survey respondent consent to linking Centrelink administrative records with the survey data of the individual Procedures for contacting and interviewing individuals in observational units Best methods to contact individuals Propo
36. ob search and main activity Job search methods and length Main activity if not in school employed or looking for work Section J Family formation e Date of start of current and first marriage de facto relationship Number of marriages de facto relationships Partner s education and employment status Partner s hours and earning in all jobs main job Children date of birth living arrangements for the child money paid received for child support other than government 32 Section K Housing and income Housing arrangements renting own Rent mortgage payment amounts Combined own and partner s income for 2005 2006 from wages salary own business workers compensation investment etc Financial help from parents Section L Health Overall health and whether it provides any limitations to work e Effect of emotional problems on daily activities Emotional health in the past 4 weeks Presence of Asthma ADHD Depression diagnosis Smoking yes no e Height and weight Section M Attitudes Attitudes on unemployment benefits Importance of own and parental education ambition and job for success in life Section N Tracking questions and SCQ A3 Youth self completion questionnaire Lifestyle and health Leisure and recreation activities Locus of control questions Physical activity Smoking and drinking habits Use of marijuana Family and frie
37. olling for a range of background and demographic factors The nature of the Australian payments system managed by Centrelink provides a population sampling frame of young Australians who grew up in a range of family circumstances In particular the same administrative database used to manage income support payments is also used to manage childcare subsidies which are not means tested and tax rebates for dependent children which are means tested and not paid to the top 15 per cent approximately of the income distribution Thus this data source from which the survey sample is drawn provides consistent administrative data not only for disadvantaged families but also for a large comparison group of middle and upper middle income families 3 See Haveman et al 2001 4 See An et al 1996 1 3 Main research questions The data generated by the project will be the foundation for research on a wide variety of issues including youth health education and employment as well as the influence of intergenerational factors on young people s socio economic outcomes The overarching goal of the YIF project is to understand the consequences of growing up in disadvantage on economic social and demographic outcomes in early adulthood To this end the research project has two main objectives The first objective is to describe the overall correlation between parents and young adult children s outcomes The second objective is to investigate
38. portion of people across income support stratification categories Snai Proportion Expected Achieved Achieved a Stratification category code in admin proportion proportion proportion r T d dataset in sample of youth of parents obs units No parental income support history A 40 9 25 0 25 2 26 4 29 0 Heavy exposure to income support programs more than six total years on Income support 7 0 D o 1 S o 3 0 han si ly i pport B 27 5 34 9 36 1 35 7 32 3 First exposure to income support system after 1998 C 8 5 10 7 12 9 12 4 12 4 First exposure to income support system between 1994 and 1998 and less than three total years on income support D 8 5 10 8 10 3 10 6 11 2 First exposure prior to 1994 and less than six total years on income support E 9 5 12 1 9 9 9 7 9 6 First exposure to income support system between 1994 and 1998 and more than three but less than six total years on income support F 5 1 6 5 5 7 5 4 5 6 Notes a In this table the term income support refers to any kind of Centrelink benefits the parents may have received excluding family payments Carer Allowance and maternity payments Table 2 shows that individuals belonging to the strata B to F have been over sampled in comparison to the proportions observed in the administrative data Individuals with heavy exposure to the income support system were over sampled since they have been found in other surveys to have lower response rates re
39. roup of young people born in July 1987 and January 1988 only which is used in the pilot stages of the survey We will refer to these young adults as the pilot birth cohort For both cohorts the selected young person in each family will be referred to as the focal youth There are two ways that a youth can appear in the administrative data One is if the youth has received a payment such as Youth Allowance directly from the Australian government in his her own name The other is that the youth has been listed as the dependent child of another individual for the purpose of obtaining eligibility and amount of payment such as Family Tax Benefit or Parenting Payment For the survey purposes certain focal youths are considered out of scope for details see Section 6 11 2 2 Parents The secondary reference population includes all adults who at any point since 1991 have received government payments for the focal youth or who had the focal youth registered as their dependent for payment purposes The vast majority of cases will be where the adult received Family Tax Benefit previously Family Allowance which is the annual tax benefit to help families with the costs of raising children The adult in question may be a birth parent an adopted parent a foster parent an older sibling a relative or a non relative who cared or is caring for the focal youth Although the family relationship between the focal youth and these adults
40. rred to as the observational unit In cases where there is no parent record the observational unit is the focal youth Each observational unit therefore has either one or two members depending upon whether there is a record for a parent Note that the observational unit is not necessarily a complete family nor a household The proportion of one person focal youth only observational units in the administrative data is 1 7 per cent with the two person focal youth and parent observational units constituting the overwhelming majority at 98 3 per cent 2 4 Data collection unit The data collection unit is the individual Different instruments are used for data collection from parents and youths as detailed below 3 Administrative data and sampling frame 3 1 The administrative database TDS2 The frame for this study is an administrative database referred to as TDS2 built from Centrelink payment records Centrelink represents a one stop shopfront for receipt of almost all government benefits including parenting and family payments unemployment benefits old age and disability pensions and provides an excellent sampling frame because most transfer payments are administered through this one agency at the national level All focal youth who appeared in the Centrelink payment records at any time since 1991 are included in this database A focal youth may appear in the administrative data for two reasons i they have claimed in t
41. rtion of invalid phone numbers The result of all attempts to contact sample individuals and to track down individuals for whom phone number information was incorrect Percentage of the provided sample that appeared to be uncontactable The number of focal youth parent interviews which were conducted on one and two different phone numbers The number of observational units where complete responses were obtained from both individuals The number of observational units where it was not possible to interview both individuals in the observational unit the number of individuals who could not be interviewed and the reason why they could not be interviewed Compare the impact on response rates of offering a payment for participation The pilot study was also used to develop and test fieldwork manuals for the use of supervisors and interviewers Data coding and provision Check that coding protocols do not produce any out of range codes Test procedure and format for supplying data to the University 34 B2 Pilot response rates Table B2 below provides summary information on the strata specific response rates for focal youth and parents interviewed during the wave 1 pilot The table also outlines differences in response rates between the groups that were offered monetary compensation for participating in the survey and the groups that were not The response rates are calculated as a proportion of the clean sample The clean sample is the
42. s and family holidays at respondent s age 14 Section D Education and employment Highest year of school completed and the type of school Highest educational degree attained e Employment status and number of jobs Hours and earning in all jobs main job Contract and salary arrangements Occupation Job search and main activity if not employed Section E Income from other sources Combined own and partner s income for 2005 2006 from wages salary own business workers compensation interest or dividends rental property etc Section F Children and relationships If respondent is a natural parent of focal youth Total number of children e Number of marriages de facto relationships For each marriage relationship o Date of start date when started living together o Reason and date of ending o Whether spouse partner is the other natural parent For the other natural parent O Country of birth o Education and employment o Amount of contact with respondent regarding the focal youth 30 If respondent is not a natural parent of focal youth the following is asked about each natural parent Relationship to respondent Country of birth Education and employment Date when stopped living with focal youth Section G Questions about Focal Youth s education and youth Focal youth s general performance in school and participation in extra curricular activities Respondents
43. shed that individuals who grow up in families that are dependent upon income support are themselves much more likely to be dependent upon income support as adults 2 At the same time the majority of children who grow up in income support dependent families are not reliant upon income support IARC Linkage Project LP0347164 entitled The Intergenerational Transmission of Dependence on Income Support Patterns Causation and Implications for Australian Social Policy Research 2 See for example Pech and McCoull 1998 and McCoull and Pech 2000 for Australian evidence and Solon et al 1991 Gottschalk 1996 Haveman and Wolfe 1995 Israel and Seeborg 1998 and Haveman et al 2001 for international evidence as adults An important objective of the project is identifying the factors which contribute to or prevent the need for social assistance amongst young adults Although the association between growing up in an income support dependent family and reliance upon social assistance as an adult are well established there has been only limited research on assessing which factors underlie this relationship A lack of educational qualifications early childbearing poor health and disability all contribute to reducing a young person s labour market opportunities and increasing the chances of needing social assistance Determining the causal effect of these risk factors is difficult since while all can cause dependence on income support ea
44. ssiasitaxceanuissiohumionnieaaneniaiiamtnuieats 29 Appendix A YIF questionnaires SUMmAary sssesesesssosesosososesosesosesosesosesosesosososososssssseses 30 AT PAPC IE CUCSE OME GILG ngri srn e a Aa aus E aE 30 A2 Yo th questionnaite nse iier ra aka N a A a A A aa eee 31 A3 Youth self completion questionnaire sic csscseastseterteensdeciastosoroatyientnantonusgereuccaspanetes 33 Appendix B Additional information on wave 1 pilot sssessssosososososososososososssssssseseses 34 BT Items test edin piloteseniaceisiniennn e a aud A 34 B2 Pilot response fates muyunan nab quae E a E Raat 35 BUDE O 2a hy spessccsits dancsvshdscusacisesusavetinctsdecwsusuasasouaseSueansnsssseuadehenocsuseasacanessmsusansadeudalgesevenvepaeens 36 List of tables Table 1 Number of young people in TDS2 and the Australian population 13 Table 2 Proportion of people across income support stratification categories 15 Table 3a Wave 1 sample and mailout structure for focal youth 21 Table 3b Wave 1 sample and mailout structure for parents 21 Table 4a Wave l focal youth Outcomes 6 och Me ies r all ao Sd 25 Table 4b Wave I parentoute mes eco ee we esl ae el tine eee A ted sok 26 Table 5 Respondents consent to data linkage per cent of total 26 Table 6 Survey dataset introductory variable description eee eee 28 Table 7 Survey datafiles characteristics a a o ooon aaa 29 Table b2s VHOETESpOnse Takes oye woes on De ce ah ae le Sea Deas 35 Overview This p
45. t grant from the Australian Research Council ARC In addition to its initial commitment of resources for the project FaCSIA has provided support for background research relevant to the project Centrelink has given in kind support by providing staff time and expertise for work related to building the administrative data set The ANU is the lead institution and has primary responsibility for coordinating the work of the research team FaCSIA and Roy Morgan Research Pty Ltd RMR the market research firm conducting the survey and for reporting to the ARC 1 2 Project motivation The overarching goal of the Youth in Focus project is to understand the ways in which economic and social disadvantage might be transferred from one generation to the next The project uses an innovative combination of survey and administrative data to explore some of the consequences for young Australians of growing up in disadvantaged families Our focus is on outcomes in the early adult years when young people are moving into higher education entering the labour market starting families and generally establishing themselves as independent adults Developing a fuller understanding of reasons why disadvantaged youth succeed or fail to succeed is a necessary first step in formulating sensible policies targeted towards breaking any cycle of dependence and promoting the social and economic independence of Australian youth In particular it is well establi
46. t out piocesSisssasire iiiaae oaan a ATAS aE EA R 18 6 7 Privacy and confidentiality issues e e ssesssesssseressseereresesrsrsreseststereresesrsreseststerenes 19 6 8 VWV ave pilot OULCONICS is onemaneunennneia eoa i a E AN 19 6 9 Wave T sample eirin e PEE AT EEEE E RE 19 6 10 Wav GO POUL processi eisai a e a AE EERE 19 6 11 Wave 1 data gathering procedure esssssessssersssstereresesrsrsseseststeresesesrssssestsrerenes 22 6 12 Interview length and outcomes of self completion questionnaire 0 23 SE Br hog V2 bl 1 ATA EAEE ENEA EE S EE a EOE 24 6 14 Paym nt to respondents ic cccsscarneinti neria aa a aE 24 6 15 Foreign language intervie WS os baceaseasccesn sti occ sancstilopenea cabdencovaaescatavmrcareaubetienad ai 24 6 16 Refusal conversion and follow up i swicniesieasicsinccyees deans ni ses ase tnaerenimne anaes 24 6 17 Interviewer monitoring and complaint handling protocol eee 25 6 18 Wave 1 Contact outcomes cccccceccccccccccccccccccececeseececseseseeseesesesesesseeseecesesesesesesesesesens 25 6 19 atta MIA CS rs cassis cssactanadvaal oea e E EE E A 26 6 20 TCI BN and trading ecm aruana aiian a a a a e pea a aa 26 6 2L WAVES QAI Eo EEEE EEEE ERE E A E E A 27 VEDA A E ELE Ee a E 27 7 1 Variable naming conventions siicsnicaisace paatismatatennsmgseanentans neniaeasane in eeaatetigtenen 27 7 2 Labeling serana ean r dy nd EE gu aaa nade ENE sa ias 29 Tavourvey datafiles characteristics ic
47. ta in approximately equal proportions In addition to investigate the possibility of introducing respondent payments to encourage participation each stratum was randomly divided into two halves one of which was offered a 15 compensation payment and the other was not see also Section 6 14 6 5 Wave 2 and wave 3 pilots Given the longitudinal nature of this project the focal youth respondents of the wave 1 pilot will be approached again for the pilots in waves 2 and 3 6 6 Pilot opt out process After drawing the initial raw sample of about 1 400 observational units Centrelink provided FaCSIA with their contact information which included name of focal youth name of parent and for each of these individuals a phone number where available and address Before passing the sample to RMR FaCSIA sent a primary approach letter and information brochure to each member of the observational unit These materials contained sufficient information to enable people selected in the sample to reach an informed decision on whether or not they wished to participate in the survey Those who did not wish to participate were requested to call a 1800 number managed by FaCSIA to opt out of the survey Survey respondents could also use this number for any queries about the survey After the end of the opt out period the 1800 number was transferred to RMR who then took the responsibility for 14 The letter and information brochure can be viewed at http yo
48. ted several times e g for different members of household a _ is added to indicate which time it appears 28 So for example the question J22 Do you only pay only receive OR both pay AND receive any financial support for this child excluding any money from the government in the youth telephone questionnaire for wave 1 would have a variable name ej22_1 when asked for the first child ej22_2 when asked for the second child etc The data file is rectangular In other words the maximum number of variables are created for each respondent and are coded missing for anyone of whom the question was not asked For example the maximum number of children the focal youth may have was set at 8 thus ej22_1 to ej22_8 exist for each observation even though the data would be largely missing for individuals with few or no children 7 2 Labeling The variables are labelled in the following way Question text These labels have a limit of 80 characters including the question number colon and spaces If the question is more than 80 characters it is abbreviated or shortened in a way that keeps the question identifiable So for the above example the label is ej22_1 do you pay receive or both any financial support for this child 7 3 Survey datafiles characteristics For each stage of the survey RMR provided to the ANU two computer files the first of which contains parent CATI survey responses and the second focal yout
49. the causal mechanisms behind these relationships paying particular attention to the intergenerational correlation in income support receipt The literature points to several mechanisms through which disadvantage might be passed from one generation to the next however little is known about their relative importance Identifying which mechanisms are most important in accounting for the intergenerational correlation in disadvantage is a necessary first step in formulating evidence based policies targeted towards breaking any cycle of welfare dependence Data collection is designed to focus on four possible transmission mechanisms low educational attainment early fertility poor health and or disability and attitudinal factors All of these may depend upon the socio economic status and income support history of a young person s family and may affect a youth s future receipt of income support Information gathered from matched pairs of youths and parents lengthy administrative data and a medium length panel survey provide the mechanisms by which these research questions may be addressed 1 4 Data sources There are two data sources for the project The first is administrative data collected by the Australian Government Centrelink from all recipients of and applicants for any type of government payment The second data source is a survey of a sample of individuals selected from this administrative data specifically for the YIF project The
50. tionnaire whether on paper or on line was estimated at 10 minutes 16 Among potential respondents who were declared uncontactable with call outcomes recorded as engaged no reply answering machine or fax 11 81 per cent of focal youth and 8 61 per cent of parents received less than 9 callbacks and 4 21 per cent of youth and 5 45 per cent of parents received less than 5 callbacks 23 6 13 Timeline The fieldwork for wave 1 primarily occurred between 15 August 2006 and 16 December 2006 Self completion questionnaires were accepted by RMR until 15 January 2007 6 14 Payment to respondents In order to test the effect of payment to respondents on response rates and characteristics of individual survey participants the pilot for wave 1 was designed so that a randomly chosen half of respondents in each stratum would be offered a payment of 15 by the market research firm and the other half would not The response rates were uniformly higher in the paid groups especially among respondents who were heavily exposed to the income support programs and therefore the decision was made to offer the 15 payment to all participants of the main wave The payments were mailed to respondents by RMR in the form of a cheque upon the completion of the telephone interview parents or upon the receipt of the self completion questionnaire by the market research firm focal youth 6 15 Foreign language interviews Due to the complex nature of the s
51. urvey questions it was decided that the interviews were to be conducted in the English language only For the respondents who did not know English sufficiently well to participate a separate code was created to record as a reason for the interview termination Among parents 410 interviews 3 54 per cent of the clean sample were terminated for this reason and among focal youth the respective number of terminated interviews was 51 0 43 per cent of the clean sample 6 16 Refusal conversion and follow up The research team government organisations involved and the market research firm engaged to undertake the survey ensured that the participants were given the fullest information about the project in order to convert as many refusals as possible Particular care was taken to emphasise the academic research nature of the survey and the government support of the YIF project to encourage participation Due to the longitudinal nature of the survey the focal youths are contacted several times between the waves in order to minimise attrition The contacts are evenly spaced between waves and include a thank you card newsletter with a summary of the survey results and a phone call closer to the next wave 24 6 17 Interviewer monitoring and complaint handling protocol The complaint handling protocol for the YIF survey was approved by the ANU Human Research Ethics Committee In the case of complaints or queries not satisfactorily addresse
52. uthinfocus anu edu au survey htm respondent liaison 6 7 Privacy and confidentiality issues The procedures for handling confidential data were set out in the YIF Project Ethics Protocol which was approved by the ANU Human Research Ethics Committee on 11 February 2005 Due to the sensitive nature of the administrative data all project personnel including all interviewers and supervisors at RMR have had to sign Deeds of Confidentiality relating to the treatment of personal contact and other information 6 8 Wave 1 pilot outcomes The information gathered in the pilot was used to judge the respondents reaction to the questions the effect of payment to respondents on response rates and the accuracy of the CATI programming particularly as it related to questionnaire routing procedures The full list of items tested in the pilot is available in Appendix B1 After the interviews for the pilot were completed RMR submitted a technical report on the outcome of the pilot study to the ANU The report critically considered strategies and methodologies tested in the pilot study and made recommendations regarding a few proposed changes Analysis of the pilot data and the technical report resulted in the shortening of the parent questionnaire correction of several glitches in the CATI programming and a positive decision to use monetary compensation to improve the response rates In addition the response rates observed in the pilot see Append
53. work September December 2008 6 21 Waves 2 and 3 Wave 2 and 3 are currently planned to be conducted approximately two years apart Pre tests and pilots will precede the main fieldwork similarly to wave 1 For all focal youths who were interviewed in wave 1 interviews at waves 2 and 3 will again be conducted by telephone Those individuals who agree to do the survey will also be provided with the self completion questionnaire to fill out either on line or on paper Individuals focal youth and parents who are not interviewed at wave 1 will be interviewed at the first wave in which they enter the sample using the wave 1 questionnaire 7 Survey dataset 7 1 Variable naming conventions The opening variables in each dataset provide summary information on the respondent and the interview Table 6 provides a list of these variables 27 Table 6 Survey dataset introductory variable description Variable name Explanation CentrelinkID Random identification number assigned by Centrelink postfixed by 0 for parent and 1 for focal youth iiin Permutation of the CentrelinkID an integer variable that is the same for parent and focal youth belonging to the same observational unit Surveyid String version of the iiin variable iisn Survey number identifies surveys of parents and focal youth main vs pilot stages of each survey wave also identifies payment and non payment respondent groups for wave 1 pilot Economic Stratificatio
Download Pdf Manuals
Related Search