Innovations in Small Area Estimation Methodologies
Lead Research Organisation:
University of Southampton
Department Name: Sch of Economic, Social & Political Sci
Abstract
Reliable statistics are crucial for policy relevant research. Small Area Estimation (SAE) methods generate robust reliable and consistent statistics at geographical scales for which survey data are either non-existent or too sparse to provide direct estimates of acceptable accuracy.
The last decade has seen a rapid increase in the use of SAE. Statistical agencies and Governmental organisations are actively developing their own suites of estimates. In the UK the Office for National Statistics (ONS) has responded to user demands by producing estimates of average household income for wards and using SAE to answer queries from local authorities, policy advisers and government departments. The Welsh Assembly Government (WAG) is actively seeking to develop capacity for SAE. Public Health England produces SAEs of adolescent smoking and chronic kidney disease.
Initial demands for small area statistics are now shifting to requirements for more complex statistics that extend beyond averages and proportions to encompass estimates of statistical distributions, multidimensional indicators (e.g. inequality and deprivation indicators) and methods for replacing the Census and adjusting Census results for undercount. These developing requirements pose significant methodological and applied real-world challenges. These challenges are deepened by different methodological approaches to SAE remaining largely unconnected, locked in disciplinary silos. The technical presentation of SAE also impedes more widespread uptake by social scientists and understanding by users.
The proposed programme of work aims to (a) develop novel SAE methodologies to better serve the needs of users and producers of SAE (b) bridge different methodological approaches to SAE, (c) apply SAE for answering substantive questions in the social sciences and (d) 'Mainstream' SAE within the quantitative social sciences through the creation of methodologically comprehensive and accessible resources. The project comprises three work packages of methodological innovative research designed to deepen the understanding of SAE and achieve the aforementioned aims.
The project will capitalise on a cross-disciplinary research team drawn together through an NCRM methodological network and reflecting a large part of the SAE expertise in the UK. Through long-standing collaborations with national and international agencies in the UK, Mexico and Brazil, which are placed at the centre of the project, we enjoy access to individual level secondary survey and Census data. Collaboration with key SAE users will ensure that the project remains relevant to user needs and that methodologies are used for expanding the set of small area statistics currently available. The involvement of international experts ensures the quality and relevance of the research.
Substantive outputs will include SAEs of attributes of interest to users, including income, inequality, deprivation, health, ethnicity and a realistic pseudo-Census dataset for use by other researchers. The project will advance knowledge across disciplines in the social sciences including social statistics, applied economics, human geography and sociology. It will additionally impact on the production of official and Census statistics. The project is committed to adding value to NCRM's training and capacity building activities by developing new resources.
The last decade has seen a rapid increase in the use of SAE. Statistical agencies and Governmental organisations are actively developing their own suites of estimates. In the UK the Office for National Statistics (ONS) has responded to user demands by producing estimates of average household income for wards and using SAE to answer queries from local authorities, policy advisers and government departments. The Welsh Assembly Government (WAG) is actively seeking to develop capacity for SAE. Public Health England produces SAEs of adolescent smoking and chronic kidney disease.
Initial demands for small area statistics are now shifting to requirements for more complex statistics that extend beyond averages and proportions to encompass estimates of statistical distributions, multidimensional indicators (e.g. inequality and deprivation indicators) and methods for replacing the Census and adjusting Census results for undercount. These developing requirements pose significant methodological and applied real-world challenges. These challenges are deepened by different methodological approaches to SAE remaining largely unconnected, locked in disciplinary silos. The technical presentation of SAE also impedes more widespread uptake by social scientists and understanding by users.
The proposed programme of work aims to (a) develop novel SAE methodologies to better serve the needs of users and producers of SAE (b) bridge different methodological approaches to SAE, (c) apply SAE for answering substantive questions in the social sciences and (d) 'Mainstream' SAE within the quantitative social sciences through the creation of methodologically comprehensive and accessible resources. The project comprises three work packages of methodological innovative research designed to deepen the understanding of SAE and achieve the aforementioned aims.
The project will capitalise on a cross-disciplinary research team drawn together through an NCRM methodological network and reflecting a large part of the SAE expertise in the UK. Through long-standing collaborations with national and international agencies in the UK, Mexico and Brazil, which are placed at the centre of the project, we enjoy access to individual level secondary survey and Census data. Collaboration with key SAE users will ensure that the project remains relevant to user needs and that methodologies are used for expanding the set of small area statistics currently available. The involvement of international experts ensures the quality and relevance of the research.
Substantive outputs will include SAEs of attributes of interest to users, including income, inequality, deprivation, health, ethnicity and a realistic pseudo-Census dataset for use by other researchers. The project will advance knowledge across disciplines in the social sciences including social statistics, applied economics, human geography and sociology. It will additionally impact on the production of official and Census statistics. The project is committed to adding value to NCRM's training and capacity building activities by developing new resources.
Planned Impact
The project supports impacts via tailored pathways to three key beneficiary groups: the project's major national and international partners; current and future SAE practitioners across diverse sectors; and SAE users (e.g. policy makers, media, general public).
Beneficiary Group One: Major national and international project partners
The project works collaboratively with major national and international partners - ONS (UK), Welsh Assembly Government (UK), CONEVAL (Mexico) and IBGE (Brazil) - with whom the academic team have existing track records. Partners will benefit directly via the production of new SAE estimates of co-selected outcomes of priority interest to them. They will also benefit through their project collaboration in terms of enhanced methodological understanding of alternative SAE approaches.
Impact strategy for this beneficiary group: Project partners co-designed the project foci to ensure fit to their priorities. To ensure continued impact a steering committee consisting of the project team, partner organisations and NCRM will be formed and meet twice a year. A first stakeholders' event, organised after 18 months, will offer feedback to the project team. A second stakeholders' event will be organised at the end of the project.
Beneficiary Group Two: Current and future SAE researchers (across academic, policy, commercial and third sector settings)
This project will overcome key methodological challenges and deliver a step-change in SAE methodologies. These methodological advances offer clear benefits for all current and future SAE researchers and require effective pathway strategies to ensure impact maximisation. Beyond this directly identifiable group, there are many more researchers interested in using SAE but who currently lack the technical capacity to conduct the work and still more researchers whose work or organisations could benefit from SAE but who are not aware of the technique and its potential. The project aims both to mainstream SAE awareness and upskill SAE technical capacity across these social science communities.
Impact strategy for this beneficiary group: A range of academic outputs (high-impact journal articles, national and international conferences, seminars), accessible outputs and learning resources ('how to' guides, technical manuals, new open source software packages, reports) will be produced and made available on the NCRM website as a focal point. Outputs will be disseminated at annual ESRC, NCRM events, advertised widely and regularly via email lists, the NCRM training directory, and policy contacts. Our commitment to adding value to NCRM's training and capacity building programme will result in developing - jointly with NCRM - a suite of SAE resources and training events. At the second stakeholders' event we will invite social sciences postdoctoral and PhD researchers including those from DTCs.
Beneficiary Group Three: Users of SAE estimates
The last decade has seen a rapid increase in SAE activity and a large body of users of publicly available small area estimates now exist: policy makers, media and the general public for example. These users benefit indirectly via enhanced policy and commercial decision-making enabled by spatially detailed SAE data and directly via their ability to use new information about their communities. To do so appropriately requires a degree of basic statistical literacy that is often lacking.
Impact strategy for this beneficiary group: These user groups are not the target audience for advanced training. Instead, to reach this group project estimates and outputs will come with simple non-technical SAE guides and meaning of the estimates suitable for a lay reader. Datasets produced by the project e.g. SAE of income, deprivation, health and ethnicity, will be disseminated via NCRM and the ESRC-funded Consumer Data Research Centre at Liverpool, which includes an interactive mapping portal aimed at the public and business communities.
Beneficiary Group One: Major national and international project partners
The project works collaboratively with major national and international partners - ONS (UK), Welsh Assembly Government (UK), CONEVAL (Mexico) and IBGE (Brazil) - with whom the academic team have existing track records. Partners will benefit directly via the production of new SAE estimates of co-selected outcomes of priority interest to them. They will also benefit through their project collaboration in terms of enhanced methodological understanding of alternative SAE approaches.
Impact strategy for this beneficiary group: Project partners co-designed the project foci to ensure fit to their priorities. To ensure continued impact a steering committee consisting of the project team, partner organisations and NCRM will be formed and meet twice a year. A first stakeholders' event, organised after 18 months, will offer feedback to the project team. A second stakeholders' event will be organised at the end of the project.
Beneficiary Group Two: Current and future SAE researchers (across academic, policy, commercial and third sector settings)
This project will overcome key methodological challenges and deliver a step-change in SAE methodologies. These methodological advances offer clear benefits for all current and future SAE researchers and require effective pathway strategies to ensure impact maximisation. Beyond this directly identifiable group, there are many more researchers interested in using SAE but who currently lack the technical capacity to conduct the work and still more researchers whose work or organisations could benefit from SAE but who are not aware of the technique and its potential. The project aims both to mainstream SAE awareness and upskill SAE technical capacity across these social science communities.
Impact strategy for this beneficiary group: A range of academic outputs (high-impact journal articles, national and international conferences, seminars), accessible outputs and learning resources ('how to' guides, technical manuals, new open source software packages, reports) will be produced and made available on the NCRM website as a focal point. Outputs will be disseminated at annual ESRC, NCRM events, advertised widely and regularly via email lists, the NCRM training directory, and policy contacts. Our commitment to adding value to NCRM's training and capacity building programme will result in developing - jointly with NCRM - a suite of SAE resources and training events. At the second stakeholders' event we will invite social sciences postdoctoral and PhD researchers including those from DTCs.
Beneficiary Group Three: Users of SAE estimates
The last decade has seen a rapid increase in SAE activity and a large body of users of publicly available small area estimates now exist: policy makers, media and the general public for example. These users benefit indirectly via enhanced policy and commercial decision-making enabled by spatially detailed SAE data and directly via their ability to use new information about their communities. To do so appropriately requires a degree of basic statistical literacy that is often lacking.
Impact strategy for this beneficiary group: These user groups are not the target audience for advanced training. Instead, to reach this group project estimates and outputs will come with simple non-technical SAE guides and meaning of the estimates suitable for a lay reader. Datasets produced by the project e.g. SAE of income, deprivation, health and ethnicity, will be disseminated via NCRM and the ESRC-funded Consumer Data Research Centre at Liverpool, which includes an interactive mapping portal aimed at the public and business communities.
Organisations
- University of Southampton (Lead Research Organisation)
- Brazilian Institute of Geography and Statistics (Collaboration)
- OFFICE FOR NATIONAL STATISTICS (Collaboration)
- Oxford Policy Management (Collaboration)
- PUBLIC HEALTH ENGLAND (Collaboration)
- National Council for the Evaluation of Social Development Policy (CONEVAL) (Collaboration)
- United Nations (UN) (Collaboration)
- World Bank Group (Collaboration)
- Welsh Local Government Association (Collaboration)
Publications
Asthana S
(2016)
Equity of utilisation of cardiovascular care and mental health services in England: a cohort-based cross-sectional study using small-area estimation
in Health Services and Delivery Research
Baldermann C
(2018)
Robust Small Area Estimation under Spatial Non-stationarity
in International Statistical Review
Bianchi A
(2018)
Estimation and Testing in M-quantile Regression with Applications to Small Area Estimation
in International Statistical Review
Dutey-Magni PF
(2016)
The spatial structure of chronic morbidity: evidence from UK census returns.
in International journal of health geographics
Espuny-Pujol F
(2018)
A global optimisation approach to range-restricted survey calibration.
in Statistics and computing
Groß M
(2020)
Switching Between Different Non-Hierachical Administrative Areas via Simulated Geo-Coordinates: A Case Study for Student Residents in Berlin
in Journal of Official Statistics
Kreutzmann A
(2019)
The R Package emdi for Estimating and Mapping Regionally Disaggregated Indicators
in Journal of Statistical Software
Description | The project has already produced a number of key findings. The use of data driven transformations, careful model selection and testing can make a difference in terms of improving the precision of small area estimates; The use of robust estimation methodologies can offer protection against the misspecification of parametric model assumptions; Geographical approaches to small area estimation e.g. Spatial-microsimulation can be translated into statistical estimators for example, synthetic estimators; The latter allows us to develop uncertainty measures for the geographical approaches to small area estimation. The original objectives have been organised around three interrelated work packages. Progress has been very good and we are confident that by the end of the award we will meet all original research objectives. Below is a summary of the work carried out in each Work Package and commentary on how you have met the research objectives. For completed research, published submitted and working papers have been written. In addition, presentations (many of which invited) on the topics have been given at international conferences. WP1 devoted to the innovations in statistical methodologies for estimating distributions and complex indicators. The main outcomes from this package up to this point are as follows: ? Work on use of optimal data driven transformations for small area prediction applied to the estimation of income distributions and complex deprivation indicators has been carried out. ? Work on model diagnostics and testing has been carried out. ? A large part of the work on alternative robust methodologies has been carried out and a working paper has been written. Currently work is being carried out on the theoretical properties of some of these methods. ? Work on robust methodologies for discrete outcomes has been carried out with colleagues based in Australia (University Technology Sydney). ? Work on the estimation of uncertainty using empirical likelihood methods has been carried out. ? Methods have been applied to income data, expenditure data, health data and crime data. WP2 is devoted to geographical approaches to small area estimation. The main outcomes from this package up to this point are as follows: ? Work on benchmark and donor pool selection has been carried. ? Work on the impact weight restrictions and integer weights has been carried out. ? Work on estimating the precision of Indirect Survey Calibration (ISC) algorithms is currently being carried out by joint work between the Liverpool and Sheffield teams. ? One of the main targets of the project is to link statistical and micro-simulation methods by putting statisticians and geographers to work together. Statisticians contribute to assisting geographers with measuring the precision of ISC - type algorithms. This is further described in the work for WP3. WP3 is devoted to bridging the gap between statistical and geographical approaches. ? Considerable work has been done on joint work between statisticians and geographers on translating the indirect survey calibration algorithm into an estimator used by statisticians. Theoretical work has already taken pace and has been successful. Current work on the topic focuses on developing uncertainty estimates and doing the same translation for the Iterative Proportional Fitting (IPF) approach. ? Work on the spatial variability of Census covariates is currently under way. ? Work on creating resources to be offered to the wider research audience is under way and a first output is the creation of open source software for small area estimation using the R package emdi. ? Work on full empirical comparisons is under way as part of writing up peer review publications. Since the last report there has been further significant progress on the interdisciplinary work packages (WP) 2 and 3. This involved intensive working between statisticians at the University of Southampton and Geographers at the Universities of Sheffield and Liverpool. The links between statistical and microsimulation approaches to small area estimation (SAE) have been worked out and outputs from this work in the form of invited presentations and peer review papers are currently prepared. It is expected that this work will have sustained impact as it will bring different methodological approaches to SAE closer. The impact of this work will also be sustained by developing training materials covering alternative methodological approaches that can capture a wider audience. In addition to the work in WP 2 and 3, further progress in methodological and applied work in WP1 has been achieved. This work has been delivered via invited presentations at major international conferences and meetings and peer review publications in top journals. Among others, two key examples include a paper read at the Royal Statistical Society on a general framework for producing official statistics and a paper in the Journal of Statistical Software on an R package for small area estimation. Both papers are consistent with what we promised to deliver in terms of tools that will enable wider use of SAE tools. In collaboration with another NCRM project (Borrowing Strength - a collaborative software development for Small Area Estimation) we are currently developing training materials and a workshop for disseminating the project results. According to our delivery plan, in July 2018 we organised a one-day workshop on SAE methods during which stakeholders and researchers showcased their research. The workshop took place as part of the Research Methods Festival organised by NCRM. Last but not least, this year there has been further links with related research grants funded by the EU's H2020 programme and national funding bodies. The award has led to further research currently funded by the EU's Horizon 2020 programme creating new research collaborations across the Europe and beyond. |
Exploitation Route | The main part of this work is being carried out while in close contact with organisations such as the UK Office for National Statistics, the Welsh Government, Public Health England. The Office for National Statistics participates in or advisory board meetings and in events organised as part of this project. In the impact section of this report we have outlined collaborations with such organisations that has already taken place as a result of this funding. In addition, the research has produced open source, R software that can be used by researchers, practitioners and postgraduate students for implementing and experimenting with the methods. Additional materials will be produced with the aim of creating training materials to be used in capacity building activities. |
Sectors | Education,Healthcare,Government, Democracy and Justice |
Description | The outputs and outreach associated with this award have led to a number of requests for research and KEE activity by international organisations. In late 2021 and early 2022 we have received requests for work by the United Nations and the World Bank in Colombia, Mozambique and Bangladesh. The outputs of specific interest are the R package emdi published in the Journal of Statistical Software and the paper on use of data-driven transformations published in the Journal of the Royal Statistical Society A. The latest added JRSS C article (Smith at.al, 2020) on small domain estimation for business surveys is relevant to the work of national statistical institutes. This article presents joint work with colleagues from the Central Bureau of Statistics in the Netherlands and the methodology was developed for solving methodological problems relating to the Dutch Business Surveys. The recent collaboration with OPM on estimating extreme poverty will develop further the methods initially developed as part of this award. We expect that this work will generate new impact which will be described in future submissions. Williamson, P. Currently working with the Office for National Statistics to help explore the viability of using our survey calibration approach to adjust the next national Census for under- and over-count. Moon, G. and Twigg, L. Small area estimates of adult smoking, adolescent smoking, adult obesity and adult overweight submitted to Public Health England for Local Health Website update. Report in draft awaiting further set of estimates. Tzavidis, N and Schmid, T. Work with the UK Office for National Statistics on using model-based small area methods for constructing local baskets of goods and for consumer price index methodology. Tzavidis, N, Schmid, T. and Rojas-Perilla, N. Work with the Welsh Government on using small area estimation methods for estimating discontinuities in national surveys. A report on this work is available from http://gov.wales/docs/caecd/research/2017/170628-identifying-potential-discontinuities-national-survey-summary-en.pdf Moon, G. and Twigg, L. Small area estimates of adult smoking, adolescent smoking, adult obesity and adult overweight, data sets and technical report accepted by Public Health England (PHE) for Local Health Website update, currently undergoing internal development within PHE prior to provision as open data. The award has led to further research currently funded by the EU's Horizon 2020 programme creating new links with National Statistical Offices across the Europe. |
Sector | Government, Democracy and Justice |
Impact Types | Policy & public services |
Description | InGRID2 |
Amount | € 9,200,000 (EUR) |
Organisation | European Commission H2020 |
Sector | Public |
Country | Belgium |
Start | 02/2017 |
End | 02/2021 |
Description | MAKSWELL |
Amount | € 125,000 (EUR) |
Funding ID | 900000 |
Organisation | European Commission H2020 |
Sector | Public |
Country | Belgium |
Start | 11/2017 |
End | 06/2020 |
Description | Advising ONS |
Organisation | Office for National Statistics |
Country | United Kingdom |
Sector | Private |
PI Contribution | Tzavidis et al., Advising ONS on the implementation of the Empirical Best Predictor (EBP) and on the use of optimal transformations. Williamson, P. et al., On-going discussions with ONS about microsimulation methods. No concrete adoption/impact as of yet. |
Collaborator Contribution | This is a long term collaboration. ONS is planning to publish experimental small area estimates of poverty indicators and income distributions. Our on-going work also via this ESRC project directly feeds into the methodology used by ONS. |
Impact | This is on-going work with ONS who is also a partner organisation in the ESRC project. ONS is actively participating in project advisory meetings. It is likely that conference presentations and publications will be developed as a result of this collaboration. We will be reporting these outputs via research fish on an ongoing basis. |
Start Year | 2016 |
Description | Collaboration with National Centre for the Evaluation of Social Development Policy (CONEVAL) in Mexico |
Organisation | National Council for the Evaluation of Social Development Policy (CONEVAL) |
Country | Mexico |
Sector | Public |
PI Contribution | CONEVAL is a partner organisation in this ESRC project. CONEVAL faces methodological challenges similar to other international organisations. The research team of this project will be communicating its key findings to CONEVAL researchers. CONEVAL has been invited to participate in a project meeting and meet also researchers from other stakeholder organisations. This will be enhance the sharing of experiences in dealing with methodological problems in small area estimation. |
Collaborator Contribution | CONEVAL is represented at the steering committee of the project and offers access to survey and Census microdata necessary for applying small area methods for estimating income distributions, multidimensional deprivation, inequality and health indicators. |
Impact | Presentation at the NTTS conference organised by EUROSTAT in Brussels in March 2017. |
Start Year | 2016 |
Description | Collaboration with National Statistics Office of Brazil (IBGE) |
Organisation | Brazilian Institute of Geography and Statistics |
Country | Brazil |
Sector | Public |
PI Contribution | IBGE is a partner organisation in this ESRC project. IBGE faces methodological challenges similar to other international organisations. The methodological tools developed in this project are applicable to solving these problems. The research team of this project will be communicating its key findings to IBGE researchers. IBGE has been invited to participate in a project meeting and meet also researchers from other stakeholder organisations. This will be enhance the sharing of experiences in dealing with methodological problems in small area estimation. |
Collaborator Contribution | How IBGE will contribute has not been decided yet. One possibility is the contribution of data in a similar fashion to the contributions made by our partner in Mexico, CONEVAL. |
Impact | Not outcomes or outputs yet. |
Start Year | 2016 |
Description | Collaboration with the Welsh Government |
Organisation | Welsh Local Government Association |
Country | United Kingdom |
Sector | Public |
PI Contribution | Closer collaboration with the Welsh Government on developing methodologies for estimating survey discontinuities. A number of conference presentations and a publication are planned as a result of this work. We are also considering writing an impact case study for REF 2020 as a result of this work. |
Collaborator Contribution | Using data provided by the Welsh Government we derived estimates of survey discontinuities at National and sub-National (Local Authority and Health Board) levels for a wide range of variables. These estimates will form the basis for subsequent adjustments made to the corresponding series. |
Impact | Presentation at the NTTS conference organised by Eurostat - Brussels March 13-17 2017. Presentation at the German speaking Survey Conference in Vienna - May 2017. Presentation at the Small Area Estimation conference in Paris - July 2017. A paper from this work will be prepared and will be submitted to a peer review statistics journal. Further research work as a result of this collaboration has been identified. This will form the basis of new methodological research. We are currently investigating the possibility of writing an impact case study to be submitted in the next REF. |
Start Year | 2016 |
Description | OPM/FCDO Project on estimating Extreme Poverty |
Organisation | Oxford Policy Management |
Country | United Kingdom |
Sector | Private |
PI Contribution | Developing small area methods for estimating extreme poverty with big data sources e.g. satellite and mobile phone data. |
Collaborator Contribution | Developing methods for modelling extreme poverty. |
Impact | Inception report and prototype product - not published yet. |
Start Year | 2020 |
Description | Research and knowledge exchange on poverty mapping and poverty assessment in Mozambique |
Organisation | World Bank Group |
Country | United States |
Sector | Public |
PI Contribution | Research and knowledge exchange on poverty mapping and poverty assessment in Mozambique. This work is directly linked to the R package emdi published in the Journal of Statistical of Statistical Software and the paper on data-driven transformations published in the Journal of the royal Statistical Society A. |
Collaborator Contribution | Joint work with Prof. Timo Schmid (formerly at the Free University Berlin) currently at the University of Bamberg. |
Impact | Technical note on poverty mapping in Mozambique. Contributions to poverty assessment report for the country. Joint work work with Economists. |
Start Year | 2022 |
Description | Research, consultancy and capacity building for the National Statistical Office of Colombia |
Organisation | United Nations (UN) |
Country | United States |
Sector | Public |
PI Contribution | Research, consultancy and capacity building for the National Statistical Office of Colombia. This work is directly linked to the R package emdi published in the Journal of Statistical Software and the paper on the use of data driven transformations published in the Journal of the royal Statistical Society A. These outputs are directly associated with this award. |
Collaborator Contribution | The UN work is associated with members of the University of Southampton Tzavidis and Luna. |
Impact | Currently there is a new bespoke course on small area estimation. Other disciplines involved include economics and geography (GIS expertise). |
Start Year | 2021 |
Description | Research, knowledge exchange and capacity building on small area estimation in Bangladesh |
Organisation | United Nations (UN) |
Country | United States |
Sector | Public |
PI Contribution | Research and knowledge exchange and capacity building in small area estimation for the National Statistics Office of Bangladesh. |
Collaborator Contribution | No contributions by other partners. |
Impact | Developing a bespoke course and technical guidance for methodologists in Bangladesh. |
Start Year | 2021 |
Description | Small Area Estimation of Health Outcomes - Graham Moon and Liz Twigg |
Organisation | Public Health England |
Country | United Kingdom |
Sector | Public |
PI Contribution | Develop small area estimates (see above), exploring different data sources and procedures for aggregation and updating. |
Collaborator Contribution | Develop small area estimates (see above), exploring different data sources and procedures for aggregation and updating. |
Impact | Integrating national surveys to estimate small area variations in poor health and limiting long-term illness in Great Britain. G Moon, G Aitken, J Taylor, L Twigg - BMJ open, 2017, vol 7(8), pp. e016939. |
Start Year | 2016 |
Title | emdi: Estimating and Mapping Disaggregated Indicators -R package |
Description | Functions that support estimating, assessing and mapping regional disaggregated indicators. So far, estimation methods comprise the model-based approach Empirical Best Prediction (see "Small area estimation of poverty indicators" by Molina and Rao (2010)), as well as their precision estimates. The assessment of the used model is supported by a summary and diagnostic plots. For a suitable presentation of estimates, map plots can be easily created. Furthermore, results can easily be exported to excel. |
Type Of Technology | Software |
Year Produced | 2016 |
Open Source License? | Yes |
Impact | This is the first year of the project. Impact will be monitored and reported in future researchfish returns. |
URL | https://cran.r-project.org/web/packages/emdi/index.html |
Description | 'Small-area estimation of comorbidity: an indirect survey calibration approach, Euorpean Colloquium on Theoretical and Quantitative Geography, York 7-11 Sept. |
Form Of Engagement Activity | A talk or presentation |
Part Of Official Scheme? | No |
Geographic Reach | International |
Primary Audience | Professional Practitioners |
Results and Impact | Presentation on small-area estimation of comorbidity at a European Colloquium. |
Year(s) Of Engagement Activity | 2017 |
Description | Adjusting the 2021 Census for under or over count: a survey calibration problem, British Society for Population Studies Annual Conference, Liverpool, 6-8 Sept. |
Form Of Engagement Activity | A talk or presentation |
Part Of Official Scheme? | No |
Geographic Reach | National |
Primary Audience | Professional Practitioners |
Results and Impact | Presentation on Census adjustments. This is an important topic for the 2021 Census and of interest to the UK Office for National Statistics. |
Year(s) Of Engagement Activity | 2017 |
Description | Calibration in Spatial Micro-simulation: Statistical properties and potential improvements for Small Area estimation. International Conference on Trends and Perspectives in Linear Statistical Inference (LinStat). Bedlewo, Poland |
Form Of Engagement Activity | A talk or presentation |
Part Of Official Scheme? | No |
Geographic Reach | International |
Primary Audience | Professional Practitioners |
Results and Impact | Invited talk |
Year(s) Of Engagement Activity | 2018 |
URL | http://www.linstat2018.put.poznan.pl/ |
Description | Construction of Regional Consumer Price Indices using Small Area Estimation, Statistische Woche, Rostock, Germany, 2017. |
Form Of Engagement Activity | A talk or presentation |
Part Of Official Scheme? | No |
Geographic Reach | International |
Primary Audience | Professional Practitioners |
Results and Impact | Presentation on the possibility of using small area methods for constructing consumer price indices. This work is of importance to the UK Office for National Statistics and to other National Statistical Institutes in Europe and beyond. The topic will be further look as part of a new EU-H2020 grant (MAKSWELL). Please see links with other projects reported in Researchfish. |
Year(s) Of Engagement Activity | 2017 |
Description | Departures from normality: The performance of the EBP under different types of transformations. Small Area Estimation Conference, Maastricht, August 17-19, 2016. |
Form Of Engagement Activity | A talk or presentation |
Part Of Official Scheme? | No |
Geographic Reach | International |
Primary Audience | Professional Practitioners |
Results and Impact | Communicate research work on a key aim of this research grant (Work Package 1) i.e. the use of optimal transformations in prediction models. This is a topic of great relevance for practitioners at National Statistics Institutes working in small area estimation. For example, the Office for National Statistics (ONS) in the UK is using transformations for producing experimental estimates of small area statistics. Our work directly informs the work by colleagues at ONS. |
Year(s) Of Engagement Activity | 2016 |
Description | Domain Estimation of Survey Discontinuities, ESRA , Lisbon, Portugal, 2017. |
Form Of Engagement Activity | A talk or presentation |
Part Of Official Scheme? | No |
Geographic Reach | International |
Primary Audience | Professional Practitioners |
Results and Impact | Invited presentation on the estimation of survey discontinuities using small area methods. The presentation was given at the conference of the European Survey Research Association. This is joint work with the Welsh Government and is motivated by a methodological problem of interest to the Welsh Government. The presentation is an example of joint work between academic researchers and professional practitioners that relates to the production of survey statistics in Wales. |
Year(s) Of Engagement Activity | 2017 |
Description | Domain Estimation of Survey Discontinuities. 49th Scientific Meeting of the Italian Statistical Society, Palermo, Italy. |
Form Of Engagement Activity | A talk or presentation |
Part Of Official Scheme? | No |
Geographic Reach | International |
Primary Audience | Professional Practitioners |
Results and Impact | Invited talk |
Year(s) Of Engagement Activity | 2018 |
Description | Domain Estimation of Survey Discontinuities. Small Area Estimation Conference, Shanghai, China. |
Form Of Engagement Activity | A talk or presentation |
Part Of Official Scheme? | No |
Geographic Reach | International |
Primary Audience | Professional Practitioners |
Results and Impact | Invited talk |
Year(s) Of Engagement Activity | 2018 |
URL | http://www.icsa.org/icsa/events/sae2018-small-area-estimation-and-other-topics-current-interest-surv... |
Description | Estimation of Quantiles based on Fay-Herriot Models, Small Area Estimation Conference, Paris, France, 2017. |
Form Of Engagement Activity | A talk or presentation |
Part Of Official Scheme? | No |
Geographic Reach | International |
Primary Audience | Professional Practitioners |
Results and Impact | Presentation at the SAE conference in Paris presenting research outputs from the ESRC/NCRM research grant on small area estimation. |
Year(s) Of Engagement Activity | 2017 |
Description | Exploring the Robustness of Log-Gamma vs. Normal for Random Effect Distributions: The Case of Count Data. International Conference on Robust Statistics. University of Wollongong, Wollongong, Australia, July 3-7, 2017 |
Form Of Engagement Activity | A talk or presentation |
Part Of Official Scheme? | No |
Geographic Reach | International |
Primary Audience | Professional Practitioners |
Results and Impact | Talk on robust small area methods for discrete outcomes. This is a presentation on one of the outputs of the ESRC/NCRM grant on SAE. |
Year(s) Of Engagement Activity | 2017 |
Description | From start to finish: A framework for the production of small area official statistics, Santiago de Chile, Chile |
Form Of Engagement Activity | A talk or presentation |
Part Of Official Scheme? | No |
Geographic Reach | International |
Primary Audience | Professional Practitioners |
Results and Impact | Invited presentation |
Year(s) Of Engagement Activity | 2018 |
URL | https://www.cepal.org/sites/default/files/courses/files/marco-estandarizado-para-produccion-estadist... |
Description | Invited talk at the UN World Data Forum. |
Form Of Engagement Activity | A talk or presentation |
Part Of Official Scheme? | No |
Geographic Reach | International |
Primary Audience | Professional Practitioners |
Results and Impact | Invited talk at the UN World Data Forum. |
Year(s) Of Engagement Activity | 2021 |
Description | Measurement of Wealth when the Sample Size is Small, Eurosystem Household Finance and Consumption Survey - Workshop, Frankfurt am Main, Germany, 2017. |
Form Of Engagement Activity | Participation in an activity, workshop or similar |
Part Of Official Scheme? | No |
Geographic Reach | International |
Primary Audience | Professional Practitioners |
Results and Impact | Users' workshop on measuring wealth using small area estimation methods |
Year(s) Of Engagement Activity | 2017 |
Description | Multilevel Modelling of Counts with Gamma-Poisson Model, International Biometric Conference, Victoria, Canada (10-15 July 2016). |
Form Of Engagement Activity | A talk or presentation |
Part Of Official Scheme? | No |
Geographic Reach | International |
Primary Audience | Professional Practitioners |
Results and Impact | Communicate novel methodological research on models for discrete data. |
Year(s) Of Engagement Activity | 2016 |
Description | Multilevel modelling of survey data under two-stage design: inference for regression parameter and small domain means by using an empirical likelihood approach", NTTS2017 conference, Eurostat, Brussels |
Form Of Engagement Activity | A talk or presentation |
Part Of Official Scheme? | No |
Geographic Reach | International |
Primary Audience | Professional Practitioners |
Results and Impact | Presentation linked to work stand in WP1 and in particular to empirical likelihood methods applied to domain estimation. Work stand lead is Dr Yves Berger. |
Year(s) Of Engagement Activity | 2017 |
Description | Multinomial mixed model via extended "poisson trick", Australian Statistical Conference, Canberra, Australia (5-9 December 2016). |
Form Of Engagement Activity | A talk or presentation |
Part Of Official Scheme? | No |
Geographic Reach | National |
Primary Audience | Professional Practitioners |
Results and Impact | Communicate novel methodological work on random effects models for discrete outcomes. |
Year(s) Of Engagement Activity | 2016 |
Description | NCRM course in applied quantile-M-quantile and expectile regression (University of Southampton) |
Form Of Engagement Activity | Participation in an activity, workshop or similar |
Part Of Official Scheme? | No |
Geographic Reach | National |
Primary Audience | Professional Practitioners |
Results and Impact | 2 day training workshop on applied quantile , M-qantile and expectable regression and applications in small area estimation. Southampton, February 9 and 10, 2017. |
Year(s) Of Engagement Activity | 2017 |
URL | https://www.ncrm.ac.uk/training/show.php?article=6569 |
Description | On the links between spatial micro-simulation and statistical small area estimation methods". Presented at the 8th ESRC Research Methods Festival, Bath, UK, 3-5 July 2018. |
Form Of Engagement Activity | A talk or presentation |
Part Of Official Scheme? | No |
Geographic Reach | International |
Primary Audience | Postgraduate students |
Results and Impact | Invited session |
Year(s) Of Engagement Activity | 2018 |
URL | https://www.ncrm.ac.uk/RMF2018/home.php |
Description | Simulating regional and sub-regional health data to examine the role of population health on regional economic performance, Regional Studies Association Annual International (British and Irish Section) Annual Conference, Harrogate, 22-24 August. |
Form Of Engagement Activity | A talk or presentation |
Part Of Official Scheme? | No |
Geographic Reach | National |
Primary Audience | Professional Practitioners |
Results and Impact | Talk on simulating regional and sub-regional health data. |
Year(s) Of Engagement Activity | 2017 |
Description | Small Area Estimation in R with Application to Mexican Income Data, NTTS 2017, Brussels, Belgium. |
Form Of Engagement Activity | A talk or presentation |
Part Of Official Scheme? | No |
Geographic Reach | International |
Primary Audience | Professional Practitioners |
Results and Impact | Presentation at the conference of the Statistical Office of the European Commission (EUROSTAT). |
Year(s) Of Engagement Activity | 2017 |
Description | Small Area Estimation in R with application to Mexican income data, Survey Conference, Vienna, Austria, 2017. |
Form Of Engagement Activity | A talk or presentation |
Part Of Official Scheme? | No |
Geographic Reach | International |
Primary Audience | Professional Practitioners |
Results and Impact | Presentation to an academic and practitioner audience on a new software for implementing small area estimation with open source software. |
Year(s) Of Engagement Activity | 2017 |
Description | Spatial variability of census variables, Regional Studies Association Annual International (British and Irish Section) Annual Conference, Harrogate, 22-24 August. |
Form Of Engagement Activity | A talk or presentation |
Part Of Official Scheme? | No |
Geographic Reach | National |
Primary Audience | Professional Practitioners |
Results and Impact | Presentation on spatial variability of census variables. This is an important topic for Census research and of interest to the UK Office for National Statistics. |
Year(s) Of Engagement Activity | 2017 |
Description | The utility of geodemographic indicators in small area estimates of limiting long-term illness, presented at International Symposium in Medical Geography, Angers, France, July 2017 (also presented to NCRM project mid term conference) |
Form Of Engagement Activity | A talk or presentation |
Part Of Official Scheme? | No |
Geographic Reach | International |
Primary Audience | Professional Practitioners |
Results and Impact | Talk on the utility of geodemographic indicators in small area estimates of limiting long-term illness. This is output related to the application of small area methods with health data as outlined in the ESRC/NCRM aims and objectives. |
Year(s) Of Engagement Activity | 2017 |
Description | Understanding the geography of social structure, International Population Geography conference, Seattle, 28-30 June. |
Form Of Engagement Activity | A talk or presentation |
Part Of Official Scheme? | No |
Geographic Reach | International |
Primary Audience | Professional Practitioners |
Results and Impact | Presentation on understanding the geography of social structure given at the international Population Geography conference. |
Year(s) Of Engagement Activity | 2017 |
Description | emdi: An R package for estimating and mapping regional disaggregated indicators, Poster session, Small Area Estimation Conference, Maastricht, August 17-19, 2016. |
Form Of Engagement Activity | Participation in an activity, workshop or similar |
Part Of Official Scheme? | No |
Geographic Reach | International |
Primary Audience | Professional Practitioners |
Results and Impact | Presentation on the newly developed open source small area software emdi. The software has been recently released and we will be monitoring its use via downloads and the requests we receive. |
Year(s) Of Engagement Activity | 2016 |