Pub Date : 2023-09-14DOI: 10.23889/ijpds.v8i2.2300
Angela Sorsby
ObjectivesThis paper presents findings from administrative data analysis examining differences between ethnic groups and men and women in the number and type of requirements that make up community sentences as well as the effectiveness of different requirements in terms of successful completion of the sentence.
MethodsThe paper presents findings from analysis of the Data First probation and criminal justice linked datasets. The analysis will focus on whether:
there are differences between ethnic groups and men and women in the number and type of requirements that make up community-based orders (rehabilitation, unpaid work, curfew and accredited programmes)
some requirements are more effective in terms of successful completion of the order.
The paper presents findings from regression analysis used to examine the above relationships while controlling for other relevant variables such as age, number of previous convictions and offence type.
FindingsGraphs will be used to set out differences between broad ethnic groups and men and women in the total number of requirements which make up community-based orders and the proportion of offenders from each group which receive each of the main types of requirement namely: rehabilitation; unpaid work; curfew; and accredited programmes. Graphs will also be used to set out differences between the different order requirements in terms of successful completion. The paper will also present findings from regression analysis which will identify differences after taking account of other factors. The findings will of necessity be based on broad ethnic groups as it is unlikely that there will be sufficient numbers of people within more narrow ethnic groups to meet statistical disclosure criteria.
ConclusionThere is a lack of information on relationships between ethnicity, gender and community sentences. Better understanding of these relationships has been identified as crucial by HM Inspectorate of Probation. This paper provides more information on these relationships enabling policy decisions to be better targeted to provide equality of outcomes.
{"title":"Ethnicity, gender and community sentences","authors":"Angela Sorsby","doi":"10.23889/ijpds.v8i2.2300","DOIUrl":"https://doi.org/10.23889/ijpds.v8i2.2300","url":null,"abstract":"ObjectivesThis paper presents findings from administrative data analysis examining differences between ethnic groups and men and women in the number and type of requirements that make up community sentences as well as the effectiveness of different requirements in terms of successful completion of the sentence.
 MethodsThe paper presents findings from analysis of the Data First probation and criminal justice linked datasets. The analysis will focus on whether:
 
 there are differences between ethnic groups and men and women in the number and type of requirements that make up community-based orders (rehabilitation, unpaid work, curfew and accredited programmes)
 some requirements are more effective in terms of successful completion of the order.
 
 The paper presents findings from regression analysis used to examine the above relationships while controlling for other relevant variables such as age, number of previous convictions and offence type.
 FindingsGraphs will be used to set out differences between broad ethnic groups and men and women in the total number of requirements which make up community-based orders and the proportion of offenders from each group which receive each of the main types of requirement namely: rehabilitation; unpaid work; curfew; and accredited programmes. Graphs will also be used to set out differences between the different order requirements in terms of successful completion. The paper will also present findings from regression analysis which will identify differences after taking account of other factors. The findings will of necessity be based on broad ethnic groups as it is unlikely that there will be sufficient numbers of people within more narrow ethnic groups to meet statistical disclosure criteria.
 ConclusionThere is a lack of information on relationships between ethnicity, gender and community sentences. Better understanding of these relationships has been identified as crucial by HM Inspectorate of Probation. This paper provides more information on these relationships enabling policy decisions to be better targeted to provide equality of outcomes.","PeriodicalId":132937,"journal":{"name":"International Journal for Population Data Science","volume":"22 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-09-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134913935","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
We present the opportunities and limitations of administrative benefits data held by local authorities for data linkage projects. Whilst the richness of this data has been exploited by practitioners for administration, its potential remains little explored by researchers. We discuss data quality, sample selection and legal gateways for data sharing.
Drawing on our experience working with over 40 local authorities, we present the structure of three datasets: the Council Tax Reduction Scheme, the Single Housing Benefits Extract and the Universal Credit Data Share. We show what variables are usually included, under which legal gateways this data can be shared and how the cohorts represented within the data compare with the low-income population. We discuss how these datasets can be linked at the household level with a number of other data held by local authorities such as social rent and Council Tax arrears, Housing Benefit overpayments and Discretionary Housing Payments (DHPs).
Administrative benefits data provides a comprehensive snapshot of a household’s financial situation. Local authorities can proactively use and share this data with external data processors to fulfil their statutory duties if a legal gateway allows. By identifying households at risk of cash shortfalls before they reach a crisis point, councils can target support when administering local welfare schemes and preventing homelessness. By assessing eligibility for benefits, they can run data-driven uptake campaigns. This data captures a proportion of the population on national and local benefits within a local authority at several points in time. Attrition is of concern since households may leave datasets over time. Some will see their income rise and no longer qualify for benefits. Others will move out of the constituency.
Local authorities routinely process longitudinal data on households receiving means-tested benefits by administering housing benefits, council tax support, and discretionary support funds. This data provides a unique real-time insight into the socioeconomic situation of low-income households. Yet, we show that its promising potential for policy research remains largely untapped.
{"title":"What can we learn from administrative benefits data?","authors":"Juliet-Nil Uraz`, Mary-Alice Doyle, Magdalena Rossetti-Youlton","doi":"10.23889/ijpds.v8i2.2334","DOIUrl":"https://doi.org/10.23889/ijpds.v8i2.2334","url":null,"abstract":"We present the opportunities and limitations of administrative benefits data held by local authorities for data linkage projects. Whilst the richness of this data has been exploited by practitioners for administration, its potential remains little explored by researchers. We discuss data quality, sample selection and legal gateways for data sharing.
 Drawing on our experience working with over 40 local authorities, we present the structure of three datasets: the Council Tax Reduction Scheme, the Single Housing Benefits Extract and the Universal Credit Data Share. We show what variables are usually included, under which legal gateways this data can be shared and how the cohorts represented within the data compare with the low-income population. We discuss how these datasets can be linked at the household level with a number of other data held by local authorities such as social rent and Council Tax arrears, Housing Benefit overpayments and Discretionary Housing Payments (DHPs).
 Administrative benefits data provides a comprehensive snapshot of a household’s financial situation. Local authorities can proactively use and share this data with external data processors to fulfil their statutory duties if a legal gateway allows. By identifying households at risk of cash shortfalls before they reach a crisis point, councils can target support when administering local welfare schemes and preventing homelessness. By assessing eligibility for benefits, they can run data-driven uptake campaigns. This data captures a proportion of the population on national and local benefits within a local authority at several points in time. Attrition is of concern since households may leave datasets over time. Some will see their income rise and no longer qualify for benefits. Others will move out of the constituency.
 Local authorities routinely process longitudinal data on households receiving means-tested benefits by administering housing benefits, council tax support, and discretionary support funds. This data provides a unique real-time insight into the socioeconomic situation of low-income households. Yet, we show that its promising potential for policy research remains largely untapped.","PeriodicalId":132937,"journal":{"name":"International Journal for Population Data Science","volume":"29 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-09-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134913945","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-09-14DOI: 10.23889/ijpds.v8i2.2296
Louise Marryat, Petra Rauchhaus, James P Boardman, Alison McFadden, Anne Whittaker, John Frank
ObjectivesChildren of women who use substances are difficult to research at a population-level using traditional research methods due to the complexity of their lives. Resultingly, we have little robust evidence on their outcomes. This study developed an administrative data cohort of children exposed to opioids and explored health outcomes.
MethodsUsing data from birth records, antenatal records, prescription data, hospital/psychiatric hospital admissions, and drug and alcohol service data, we identified 6,408 children (born 2009-2019) in Scotland who were exposed to opioids through illicit use and/or medication assisted treatment (i.e. methadone/buprenorphine). A control group (n. 19,089) of children not exposed to opioids were matched on age of mother and Scottish Index of Multiple Deprivation. Data were described and linear and logistic regression models were used to examine the relationship between risk factors (such as drug and alcohol use in pregnancy, gestation at booking and at birth), and key early outcomes.
ResultsAlthough the majority of women had their substance use recorded in antenatal records, 28.9% did not, demonstrating the importance of using multiple administrative datasets to form the cohort. Children in the cohort were more likely to experience a range of adverse outcomes including being born early (17% born prematurely, compared with 6.5% in control group), having a below normal Apgar score (the scoring system used to assess newborns shortly after birth) (2.9% in cohort vs. 1.5% in controls), having significantly lower birthweight, length and head circumference, and more likely to be removed from their mother prior hospital discharge. Differences between the cohorts remained after controlling for other risk factors including alcohol use, and gestation.
ConclusionThis feasibility study brought together a cohort of children usually excluded from traditional forms of research. The research demonstrated early differences in outcomes between exposed children and others from similar socio-economic groups. The next stage of this research is exploring health and development outcomes in the preschool period.
{"title":"Hidden in plain sight: Using administrative data to conduct a longitudinal cohort study of children exposed to opioids in pregnancy","authors":"Louise Marryat, Petra Rauchhaus, James P Boardman, Alison McFadden, Anne Whittaker, John Frank","doi":"10.23889/ijpds.v8i2.2296","DOIUrl":"https://doi.org/10.23889/ijpds.v8i2.2296","url":null,"abstract":"ObjectivesChildren of women who use substances are difficult to research at a population-level using traditional research methods due to the complexity of their lives. Resultingly, we have little robust evidence on their outcomes. This study developed an administrative data cohort of children exposed to opioids and explored health outcomes.
 MethodsUsing data from birth records, antenatal records, prescription data, hospital/psychiatric hospital admissions, and drug and alcohol service data, we identified 6,408 children (born 2009-2019) in Scotland who were exposed to opioids through illicit use and/or medication assisted treatment (i.e. methadone/buprenorphine). A control group (n. 19,089) of children not exposed to opioids were matched on age of mother and Scottish Index of Multiple Deprivation. Data were described and linear and logistic regression models were used to examine the relationship between risk factors (such as drug and alcohol use in pregnancy, gestation at booking and at birth), and key early outcomes.
 ResultsAlthough the majority of women had their substance use recorded in antenatal records, 28.9% did not, demonstrating the importance of using multiple administrative datasets to form the cohort. Children in the cohort were more likely to experience a range of adverse outcomes including being born early (17% born prematurely, compared with 6.5% in control group), having a below normal Apgar score (the scoring system used to assess newborns shortly after birth) (2.9% in cohort vs. 1.5% in controls), having significantly lower birthweight, length and head circumference, and more likely to be removed from their mother prior hospital discharge. Differences between the cohorts remained after controlling for other risk factors including alcohol use, and gestation.
 ConclusionThis feasibility study brought together a cohort of children usually excluded from traditional forms of research. The research demonstrated early differences in outcomes between exposed children and others from similar socio-economic groups. The next stage of this research is exploring health and development outcomes in the preschool period.","PeriodicalId":132937,"journal":{"name":"International Journal for Population Data Science","volume":"18 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-09-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134913947","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-09-14DOI: 10.23889/ijpds.v8i2.2219
Pratibha Vellanki, Mary Cleaton
ObjectivesThe Integrated Data Service (IDS) is a new cross-government service that facilitates research for the public good. Key to its success are Integrated Data Assets (IDAs): de-identified, grouped datasets that are joinable on an artificial ID and themed on a given topic. The Demographic Index (DI) comprises five linked administrative datasets. We are developing a generalisable method that will link administrative and survey datasets to the DI via a customisable, reproducible pipeline, to produce IDAs.
MethodsThe method focuses on the traditional methodologies of deterministic and probabilistic data linkage and uses the Splink implementation of the Fellegi-Sunter method for probabilistic matching. The pipeline will include a tool for quality-assurance (QA) via clerical review.
We are researching a generalisable implementation of Splink, deriving the method’s control parameters using the results of the deterministic matching. Additionally, we are researching application of Locality Sensitive Hashing (LSH), a dimensionality-reduction method suggested to improve computational efficiency, for blocking. This is especially important due to the large size of the datasets involved.
ResultsWe plan to produce linked datasets with three quality levels – prioritising precision, balancing precision and recall and prioritising recall. As the datasets are always linked to the DI, the DI’s artificial ID can be used as a ‘spine’ to bring them together as assets (IDAs).
Initially, the method will be used on the 2021 England and Wales Census. Despite not including clerical matching in the method (except for quality-assurance), we anticipate a high precision and recall due to the quality of the Census and the number of linkage variables available. Thereafter, we plan for user testing with other datasets, including the Labour Market Survey.
ConclusionOur generalisable linkage pipeline for the DI will, through its IDA outputs, facilitate research for the public good. This research will directly impact government policy and responses to national health emergencies, including Covid-19, and support government priorities such as Levelling Up and the transition towards Net Zero.
{"title":"A generalisable linkage pipeline (GLADIS) to facilitate research for the public good","authors":"Pratibha Vellanki, Mary Cleaton","doi":"10.23889/ijpds.v8i2.2219","DOIUrl":"https://doi.org/10.23889/ijpds.v8i2.2219","url":null,"abstract":"ObjectivesThe Integrated Data Service (IDS) is a new cross-government service that facilitates research for the public good. Key to its success are Integrated Data Assets (IDAs): de-identified, grouped datasets that are joinable on an artificial ID and themed on a given topic. The Demographic Index (DI) comprises five linked administrative datasets. We are developing a generalisable method that will link administrative and survey datasets to the DI via a customisable, reproducible pipeline, to produce IDAs.
 MethodsThe method focuses on the traditional methodologies of deterministic and probabilistic data linkage and uses the Splink implementation of the Fellegi-Sunter method for probabilistic matching. The pipeline will include a tool for quality-assurance (QA) via clerical review.
 We are researching a generalisable implementation of Splink, deriving the method’s control parameters using the results of the deterministic matching. Additionally, we are researching application of Locality Sensitive Hashing (LSH), a dimensionality-reduction method suggested to improve computational efficiency, for blocking. This is especially important due to the large size of the datasets involved.
 ResultsWe plan to produce linked datasets with three quality levels – prioritising precision, balancing precision and recall and prioritising recall. As the datasets are always linked to the DI, the DI’s artificial ID can be used as a ‘spine’ to bring them together as assets (IDAs).
 Initially, the method will be used on the 2021 England and Wales Census. Despite not including clerical matching in the method (except for quality-assurance), we anticipate a high precision and recall due to the quality of the Census and the number of linkage variables available. Thereafter, we plan for user testing with other datasets, including the Labour Market Survey.
 ConclusionOur generalisable linkage pipeline for the DI will, through its IDA outputs, facilitate research for the public good. This research will directly impact government policy and responses to national health emergencies, including Covid-19, and support government priorities such as Levelling Up and the transition towards Net Zero.","PeriodicalId":132937,"journal":{"name":"International Journal for Population Data Science","volume":"85 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-09-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134913232","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
The study objectives were to (1) examine the association between children’s ethnicity and final legal orders at the end of family care proceedings (section 31 of the 1989 Children Act), and (2) test whether residential context, such as co-ethnic density and area-level deprivation, moderates this association.
Two sources of data were used for this study. The first consisted of records routinely generated by Cafcass (England) and stored in the Secure Anonymised Information Linkage (SAIL) databank, and the second was the 2021 England Census. The focus was on children whose initial care proceedings took place between 2015/2016 and 2020/2021 and concluded with a recorded final legal order outcome (N = 98,161). Three-level logistic regression models were employed to estimate the relationship between children's ethnicity and adoption, along with the potential moderating effects of co-ethnic density and area-level deprivation.
Children's ethnicity is significantly associated with placement for adoption, with white children being more likely to be subject to placement orders compared to children from all other ethnic groups combined (Asian, black, mixed or multiple, and other ethnic groups). Higher local authority co-ethnic density considerably reduces the likelihood of adoption for children of other ethnicities besides white, but not for white children. Moreover, white children living in the most deprived LSOAs are more likely to be placed for adoption than those residing in the least deprived LSOAs. However, the likelihood of placement for adoption remains consistent across all LSOA deprivation quintiles for children from ethnicities other than white. Local authority-level deprivation does not appear to moderate the relationship between children's ethnicity and adoption.
This study sheds light on the intricate relationship between ethnicity, residential context, and adoption. While previous research has indicated that white children are more likely to be adopted, the findings enhance our understanding of the underlying mechanisms influencing adoption, paving the way for a more equitable family justice system.
{"title":"The adoption paradox: Exploring the role of ethnicity, deprivation, and co-ethnic density in care proceedings in England","authors":"Bachar Alrouh, Mariam Abouelenin, Stefanie Doebler, Karen Broadhurst","doi":"10.23889/ijpds.v8i2.2252","DOIUrl":"https://doi.org/10.23889/ijpds.v8i2.2252","url":null,"abstract":"The study objectives were to (1) examine the association between children’s ethnicity and final legal orders at the end of family care proceedings (section 31 of the 1989 Children Act), and (2) test whether residential context, such as co-ethnic density and area-level deprivation, moderates this association.
 Two sources of data were used for this study. The first consisted of records routinely generated by Cafcass (England) and stored in the Secure Anonymised Information Linkage (SAIL) databank, and the second was the 2021 England Census. The focus was on children whose initial care proceedings took place between 2015/2016 and 2020/2021 and concluded with a recorded final legal order outcome (N = 98,161). Three-level logistic regression models were employed to estimate the relationship between children's ethnicity and adoption, along with the potential moderating effects of co-ethnic density and area-level deprivation.
 Children's ethnicity is significantly associated with placement for adoption, with white children being more likely to be subject to placement orders compared to children from all other ethnic groups combined (Asian, black, mixed or multiple, and other ethnic groups). Higher local authority co-ethnic density considerably reduces the likelihood of adoption for children of other ethnicities besides white, but not for white children. Moreover, white children living in the most deprived LSOAs are more likely to be placed for adoption than those residing in the least deprived LSOAs. However, the likelihood of placement for adoption remains consistent across all LSOA deprivation quintiles for children from ethnicities other than white. Local authority-level deprivation does not appear to moderate the relationship between children's ethnicity and adoption.
 This study sheds light on the intricate relationship between ethnicity, residential context, and adoption. While previous research has indicated that white children are more likely to be adopted, the findings enhance our understanding of the underlying mechanisms influencing adoption, paving the way for a more equitable family justice system.","PeriodicalId":132937,"journal":{"name":"International Journal for Population Data Science","volume":"30 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-09-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134913246","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-09-14DOI: 10.23889/ijpds.v8i2.2336
Yongchao Jing, Sin Yi Cheung, Lucy Griffiths, Jonathan Scourfield
ObjectivesChildren’s chances of receiving welfare interventions are found to vary by ethnicity in England, but the ethnic pattern in child welfare interventions in Wales over time is unknown. We aim to estimate the scale and trend of ethnic inequalities in intervention rates in Wales over a 12-year period, using population-based linked administrative records.
MethodsWe analyse cross-sectional administrative children’s social care data: the Children in Need (CIN) dataset from 2010 to 2016 and the Children Receiving Care and Support (CRCS) dataset from 2017 to 2021. These data are linked to both the ethnic population data to obtain the ethnic variation in child welfare intervention rates and, using the WIMD, to identify the relative deprivation level in the neighbourhoods from which children entered care. For observations in CIN/CRCS whose ethnicity are missing, we link Census 2011 to obtain the children’s ethnicity to achieve a fuller coverage. For the first time, our analysis also links children’s social care data with the religion variable in the Census data, allowing us to estimate the extent of religious inequalities in child welfare intervention.
ResultsBased on research findings in England on ethnic variation in child welfare intervention, we hypothesise higher intervention rates among Black children and a lower intervention rates among Asian children in Wales, compared to White children, controlling for deprivation status. By extension, we also hypothesise lower intervention rates among Muslim children compared to Christian children.
ConclusionsNo research to date has quantitatively documented the pattern and trend of ethnic inequalities in child welfare intervention in Wales using linked administrative data on CIN/CRCS. Our study is also the first to examine religious inequality in UK child welfare by linking social care data to the Census. Our findings will have significant implications on policy and practice in social work and particularly in children’s social services. We conclude by discussing what future research questions may emerge from the new insight we shed.
{"title":"Trends in ethnic inequality in child welfare interventions in Wales, 2010 – 2021","authors":"Yongchao Jing, Sin Yi Cheung, Lucy Griffiths, Jonathan Scourfield","doi":"10.23889/ijpds.v8i2.2336","DOIUrl":"https://doi.org/10.23889/ijpds.v8i2.2336","url":null,"abstract":"ObjectivesChildren’s chances of receiving welfare interventions are found to vary by ethnicity in England, but the ethnic pattern in child welfare interventions in Wales over time is unknown. We aim to estimate the scale and trend of ethnic inequalities in intervention rates in Wales over a 12-year period, using population-based linked administrative records.
 MethodsWe analyse cross-sectional administrative children’s social care data: the Children in Need (CIN) dataset from 2010 to 2016 and the Children Receiving Care and Support (CRCS) dataset from 2017 to 2021. These data are linked to both the ethnic population data to obtain the ethnic variation in child welfare intervention rates and, using the WIMD, to identify the relative deprivation level in the neighbourhoods from which children entered care. For observations in CIN/CRCS whose ethnicity are missing, we link Census 2011 to obtain the children’s ethnicity to achieve a fuller coverage. For the first time, our analysis also links children’s social care data with the religion variable in the Census data, allowing us to estimate the extent of religious inequalities in child welfare intervention.
 ResultsBased on research findings in England on ethnic variation in child welfare intervention, we hypothesise higher intervention rates among Black children and a lower intervention rates among Asian children in Wales, compared to White children, controlling for deprivation status. By extension, we also hypothesise lower intervention rates among Muslim children compared to Christian children.
 ConclusionsNo research to date has quantitatively documented the pattern and trend of ethnic inequalities in child welfare intervention in Wales using linked administrative data on CIN/CRCS. Our study is also the first to examine religious inequality in UK child welfare by linking social care data to the Census. Our findings will have significant implications on policy and practice in social work and particularly in children’s social services. We conclude by discussing what future research questions may emerge from the new insight we shed.","PeriodicalId":132937,"journal":{"name":"International Journal for Population Data Science","volume":"61 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-09-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134913324","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-09-14DOI: 10.23889/ijpds.v8i2.2348
Nicolas Libuy, Jorge Pacheco, Jorge Vargas
ObjectivesGovernments often struggle to accurately estimate the number of migrants using public services due to the lack of a unique national ID. We aim to study this in the context of migrant access to immunization programs in Chile and estimate vaccine coverage in school-age children.
MethodsTo estimate vaccine coverage for migrant school-age children, we combined data from two databases: the Chilean National Immunization Register (which contained 77.9 million records) and the School Enrollment database (which contained around 68 million records, representing about 3.6 pupils per year). Using Splink, a Python package developed by the UK Ministry of Justice, we created a probability linkage model to link and deduplicate records of migrants who lack a unique national ID. The following linkage keys were considered in the model: first and second name, first and last name and date of birth. Linkage quality was evaluated using ‘gold standard data.
ResultsIn 2022, we find that out of 3,644,467 students enrolled in school, 140,317 of them were migrants who didn't have a Chilean national ID. Additionally, in the NIR database, 5.2 out of 77.9 million records belonged to migrants without a national ID. After removing duplicates from both databases, our linkage model determined that 52,524 of the 140,317 students without a national ID in SE were linked to NIR (37.4%). We find that excluding migrants without national IDs when estimating national vaccine coverage for school-aged children leads to an underestimation of 2%, from 86% to 88%.
ConclusionOur findings emphasize the significance of utilizing linkage techniques in order to accurately estimate access to public services for migrant populations who typically lack a national ID. By linking their records across public institutions, more reliable data can be obtained.
{"title":"Using probabilistic linkage to improve estimates of access to services among the migrant population: The case of access to immunisation programs in Chile","authors":"Nicolas Libuy, Jorge Pacheco, Jorge Vargas","doi":"10.23889/ijpds.v8i2.2348","DOIUrl":"https://doi.org/10.23889/ijpds.v8i2.2348","url":null,"abstract":"ObjectivesGovernments often struggle to accurately estimate the number of migrants using public services due to the lack of a unique national ID. We aim to study this in the context of migrant access to immunization programs in Chile and estimate vaccine coverage in school-age children.
 MethodsTo estimate vaccine coverage for migrant school-age children, we combined data from two databases: the Chilean National Immunization Register (which contained 77.9 million records) and the School Enrollment database (which contained around 68 million records, representing about 3.6 pupils per year). Using Splink, a Python package developed by the UK Ministry of Justice, we created a probability linkage model to link and deduplicate records of migrants who lack a unique national ID. The following linkage keys were considered in the model: first and second name, first and last name and date of birth. Linkage quality was evaluated using ‘gold standard data.
 ResultsIn 2022, we find that out of 3,644,467 students enrolled in school, 140,317 of them were migrants who didn't have a Chilean national ID. Additionally, in the NIR database, 5.2 out of 77.9 million records belonged to migrants without a national ID. After removing duplicates from both databases, our linkage model determined that 52,524 of the 140,317 students without a national ID in SE were linked to NIR (37.4%). We find that excluding migrants without national IDs when estimating national vaccine coverage for school-aged children leads to an underestimation of 2%, from 86% to 88%.
 ConclusionOur findings emphasize the significance of utilizing linkage techniques in order to accurately estimate access to public services for migrant populations who typically lack a national ID. By linking their records across public institutions, more reliable data can be obtained.","PeriodicalId":132937,"journal":{"name":"International Journal for Population Data Science","volume":"145 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-09-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134913330","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-09-14DOI: 10.23889/ijpds.v8i2.2356
Claire Grant, Claire Grant, Astrid Guttmann, Simone N. Vigod, Isobel Sharpe, Kinwah Fung, Hilary Brown
ObjectivesOne in 8 pregnancies are to women with disabilities. These mothers can face additional social, structural, and health-related challenges, and negative health care provider assumptions about their parenting capacity. We aimed to examine rates of newborn discharge to child protection comparing newborns of mothers with and without a disability.
MethodWe are conducting a population-based cohort study in Ontario, Canada using linked administrative health data. The cohort includes all women in Ontario with a live birth between 2003 and 2020. Diagnostic algorithms were applied to health care encounters prior to pregnancy to identify maternal disability. We will use modified Poisson regression to estimate the relative risk of discharge to child protection immediately after the birth hospital stay, comparing newborns of women with physical, sensory, developmental, and multiple disabilities to those without disabilities. Models will be adjusted for socio-demographic factors, antenatal care receipt, and maternal mental illness and substance use disorders.
ResultsThe study cohort includes of over 1.4 million newborns delivered to women with physical disabilities (n=120,014), sensory disabilities (n=39,892), developmental disabilities (n=2,182), multiple disabilities (n=8,428), and no known disability (n=1,269,633). Analyses are ongoing and results will be concluded by the conference date.
ConclusionEarly infancy is a critical period for breastfeeding and maternal-infant bonding. Findings will inform the development of tailored services and resources for supporting women with disabilities in antenatal care and after birth by identifying those most at-risk of child protection intervention, thus potentially reducing maternal-newborn separations.
{"title":"Maternal disability and newborn discharge to child protection in Ontario, Canada","authors":"Claire Grant, Claire Grant, Astrid Guttmann, Simone N. Vigod, Isobel Sharpe, Kinwah Fung, Hilary Brown","doi":"10.23889/ijpds.v8i2.2356","DOIUrl":"https://doi.org/10.23889/ijpds.v8i2.2356","url":null,"abstract":"ObjectivesOne in 8 pregnancies are to women with disabilities. These mothers can face additional social, structural, and health-related challenges, and negative health care provider assumptions about their parenting capacity. We aimed to examine rates of newborn discharge to child protection comparing newborns of mothers with and without a disability.
 MethodWe are conducting a population-based cohort study in Ontario, Canada using linked administrative health data. The cohort includes all women in Ontario with a live birth between 2003 and 2020. Diagnostic algorithms were applied to health care encounters prior to pregnancy to identify maternal disability. We will use modified Poisson regression to estimate the relative risk of discharge to child protection immediately after the birth hospital stay, comparing newborns of women with physical, sensory, developmental, and multiple disabilities to those without disabilities. Models will be adjusted for socio-demographic factors, antenatal care receipt, and maternal mental illness and substance use disorders.
 ResultsThe study cohort includes of over 1.4 million newborns delivered to women with physical disabilities (n=120,014), sensory disabilities (n=39,892), developmental disabilities (n=2,182), multiple disabilities (n=8,428), and no known disability (n=1,269,633). Analyses are ongoing and results will be concluded by the conference date.
 ConclusionEarly infancy is a critical period for breastfeeding and maternal-infant bonding. Findings will inform the development of tailored services and resources for supporting women with disabilities in antenatal care and after birth by identifying those most at-risk of child protection intervention, thus potentially reducing maternal-newborn separations.","PeriodicalId":132937,"journal":{"name":"International Journal for Population Data Science","volume":"34 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-09-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134913331","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-09-14DOI: 10.23889/ijpds.v8i2.2217
Van Phan, Felix Ritchie, Alex Bryson, John Forth, Lucy Stokes, Damian Whittard
ObjectivesGovernments acquire extensive data holdings and face increasing pressure to make these available as record-level microdata for research. However, turning data into research-ready data (RRD) is not a straightforward exercise. We demonstrate how even in simple cases researcher involvement can bring substantial rewards for effective RRD development.
MethodsThis paper reports on an ADRUK-funded project to take a dataset originally collected by the Office for National Statistics for official statistics (the UK Annual Survey of Hours and Earnings, ASHE), formally review its microanalytical characteristics, link it to Census 2011 data, and prepare a new ‘research ready dataset’ with appropriate documentation and coding. This should have been straightforward as the datasets had already been widely used as research microdata. However, the involvement of academic researchers in the production of research-ready data led to many important new insights.
ResultsThe research programme had 3 aims: testing assumptions about the data; reviewing data quality; and adding value.
Because of its sampling model, ASHE is assumed to have random non-response both longitudinally and in cross section. The research team showed that was untrue: there was higher attrition than expected, and both longitudinal and cross-sectional non-response appeared non-random..
The data quality review showed further concerns about the accuracy of some geographical indicators, and some variables of opaque provenance; in contrast, we confirmed the accuracy of administrative variables created by ONS.
As well as being important for researchers, these findings have the potential for significant effects on official statistics produced from the source data, enhancing the value of the source data.
Finally, value was added from new variables which reflected the team’s wide research interests
ConclusionOften in government the assumption is that creating RRDs is a matter of creatign files and giving access to the researchers. Insights from our work show that the deep involvement of the research community can bring rewards for both data holders and researchers. For RRDs, researcher-led construction is vital.
{"title":"Turning data into research-ready data","authors":"Van Phan, Felix Ritchie, Alex Bryson, John Forth, Lucy Stokes, Damian Whittard","doi":"10.23889/ijpds.v8i2.2217","DOIUrl":"https://doi.org/10.23889/ijpds.v8i2.2217","url":null,"abstract":"ObjectivesGovernments acquire extensive data holdings and face increasing pressure to make these available as record-level microdata for research. However, turning data into research-ready data (RRD) is not a straightforward exercise. We demonstrate how even in simple cases researcher involvement can bring substantial rewards for effective RRD development.
 MethodsThis paper reports on an ADRUK-funded project to take a dataset originally collected by the Office for National Statistics for official statistics (the UK Annual Survey of Hours and Earnings, ASHE), formally review its microanalytical characteristics, link it to Census 2011 data, and prepare a new ‘research ready dataset’ with appropriate documentation and coding. This should have been straightforward as the datasets had already been widely used as research microdata. However, the involvement of academic researchers in the production of research-ready data led to many important new insights.
 ResultsThe research programme had 3 aims: testing assumptions about the data; reviewing data quality; and adding value.
 Because of its sampling model, ASHE is assumed to have random non-response both longitudinally and in cross section. The research team showed that was untrue: there was higher attrition than expected, and both longitudinal and cross-sectional non-response appeared non-random..
 The data quality review showed further concerns about the accuracy of some geographical indicators, and some variables of opaque provenance; in contrast, we confirmed the accuracy of administrative variables created by ONS.
 As well as being important for researchers, these findings have the potential for significant effects on official statistics produced from the source data, enhancing the value of the source data.
 Finally, value was added from new variables which reflected the team’s wide research interests
 ConclusionOften in government the assumption is that creating RRDs is a matter of creatign files and giving access to the researchers. Insights from our work show that the deep involvement of the research community can bring rewards for both data holders and researchers. For RRDs, researcher-led construction is vital.","PeriodicalId":132937,"journal":{"name":"International Journal for Population Data Science","volume":"2013 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-09-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134913336","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-09-14DOI: 10.23889/ijpds.v8i2.2259
Louise Marryat, Camila Biazus-Dalcin, Hazel Booth, Sarah Gray, Joan Love, Andrea Mohan, Sreekanth Thekkumkara
ObjectivesAdministrative data research requires trust that data will be used sensitively and wisely. People who use drugs are frequently stigmatised, and trust may be a particular barrier. This project aims to understand the perceptions of people who use drugs around the use of their administrative health data for research purposes.
MethodsThis project will work with Restoration Fife, a third-sector organisation based in Fife, Scotland, that supports people who use drugs. We are conducting focus groups exploring how administrative health data are used in research from the perspectives of people who use drugs, including discussion around different types/sources of data. Data will be analysed using the Framework approach. We will also work with an artist and members of Restoration Fife, to co-produce a short, animated film about how administrative data are used in research around drug use, in order to educate the wider population about how their data are used.
Results*This presentation will discuss findings from the focus groups on the perceptions of usage of administrative data for different types of research. It will also discuss the use of administrative data in the context of findings from previous studies involving general populations and populations with other vulnerabilities, such as care-experienced populations and people with mental health difficulties. We will also provide a viewing of the film within this paper session.
*This project is funded by Research Data Scotland and runs from April to September 2023: all results will therefore be available by the time of the ADR conference in November 2023.
ConclusionsEvidence suggests low levels of public awareness of how and why data are used. We know little about perceptions of people who use drugs, for whom trust of services may be a particular issue. This study uses innovative methods to provide a platform for voices rarely heard in this context.
{"title":"‘My Data’ project: Understanding the perceptions of administrative data usage for research of people who use drugs","authors":"Louise Marryat, Camila Biazus-Dalcin, Hazel Booth, Sarah Gray, Joan Love, Andrea Mohan, Sreekanth Thekkumkara","doi":"10.23889/ijpds.v8i2.2259","DOIUrl":"https://doi.org/10.23889/ijpds.v8i2.2259","url":null,"abstract":"ObjectivesAdministrative data research requires trust that data will be used sensitively and wisely. People who use drugs are frequently stigmatised, and trust may be a particular barrier. This project aims to understand the perceptions of people who use drugs around the use of their administrative health data for research purposes.
 MethodsThis project will work with Restoration Fife, a third-sector organisation based in Fife, Scotland, that supports people who use drugs. We are conducting focus groups exploring how administrative health data are used in research from the perspectives of people who use drugs, including discussion around different types/sources of data. Data will be analysed using the Framework approach. We will also work with an artist and members of Restoration Fife, to co-produce a short, animated film about how administrative data are used in research around drug use, in order to educate the wider population about how their data are used.
 Results*This presentation will discuss findings from the focus groups on the perceptions of usage of administrative data for different types of research. It will also discuss the use of administrative data in the context of findings from previous studies involving general populations and populations with other vulnerabilities, such as care-experienced populations and people with mental health difficulties. We will also provide a viewing of the film within this paper session.
 *This project is funded by Research Data Scotland and runs from April to September 2023: all results will therefore be available by the time of the ADR conference in November 2023.
 ConclusionsEvidence suggests low levels of public awareness of how and why data are used. We know little about perceptions of people who use drugs, for whom trust of services may be a particular issue. This study uses innovative methods to provide a platform for voices rarely heard in this context.","PeriodicalId":132937,"journal":{"name":"International Journal for Population Data Science","volume":"145 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-09-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134913338","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：