22nd GSS Methodology Symposium

The 22nd GSS Methodology symposium took place on Wednesday the 12th of July 2017 at the Westminster Conference Centre, 1 Victoria Street, London.

The theme for the Symposium was: Methodology: Insight; Innovation; Implication; Impact

Researchers and methodologists are driven quite naturally to look for the insight and innovation that leads to improvement in the methods that ensure that the statistics provided by the GSS year on year remains at the heart of Government policy and decision making. However, the impact of this work and its contribution to the UKSA ‘Better Statistics, Better Decisions’ Strategy for UK statistics is not always broadly recognised, simply because it does not make the press release or glossy public facing visualisation.

The aim of the 2017 GSS Methodology Symposium was to provide researchers and methodologists the opportunity, not only to share ongoing research and innovative ideas, but also to celebrate why that work is crucial and how it makes an important difference to policy and society.

As in previous years, the Symposium was extremely successful, with the informative nature of the presentations and such a wide range of topics presented and discussed throughout the day being high on the list of delegates’ favourite features.

All in all, there were five key-note presentations, two at the start of each the morning and afternoon parallel Sessions and a final talk to close the Symposium.

Morning Keynote Session:
Chair: Pete Brodie: Head of Survey Methodology & Statistical Computing, Office for National Statistics.
Professor Peter W.F. Smith: Director of ESRC ADRN, University of Southampton: ADRC-E: Making a Methodological Impact.
Abstract      Slides     Paper      
Jane Naylor: Deputy Director for Methodology, Office for National Statistics: What have the methodologists done for us?
Abstract      Notes     Slides    Blog

Afternoon Keynote session
Chair: Jason Bradbury: Programme Director for Data Collection, Office for National Statistics.
John Pullinger: National Statistician, UK Statistics Authority: Methodology – Driving Innovation in Statistics.
Abstract     Paper 
Dr James Gillan: Economic and Labour Market Statistics, Northern Ireland Statistics and Research Agency: Connect and Transform.
Abstract     Slides     

 

Closing Keynote session

Chair: Simon Compton: Head of Statistics, Competition and Markets Authority.
Sarah Henry: Director of Methods, Data, and Research, Office for National Statistics: Our data Capability – United We Stand!
Abstract     Slides     Blog

 

 

Parallel sessions
The key-note presentations were supported by eight parallel Sessions throughout the course of the day. Tom Smith from the Office for National Statistics Data Science Campus kicked off the first morning session dedicated to the work of his team. The three other morning Sessions covered new research and developments in population estimation methods, time series and longitudinal analyses, and sample design.

Following an excellent lunch, the four Sessions in the afternoon were equally diverse. Session one covered transformation and harmonisation across the GSS, including lessons for anthropology. The three Sessions to follow explored and outlined innovative work in ONS address registers, statistical linkage methodology, geography, and big data analytics.

Parallel Session 1
Data science for public good – Chair: Tom Smith, ONS Data Science Campus (DSC)
Economic statistics and data science – Louisa Nolan Abstract Slides
Developing statistics from image data – Tom Smith Abstract
Building data science capability across government – Alexis Fernquest, Gareth L. Jones Abstract Slides Paper
Parallel Session 2
Counting on developments in population estimation methods Gary Brown (ONS)
What is the population of Afghanistan? – Kim Bradford Smith, Nicola Wardrop, Francis Harper Abstract Slides Paper
Generalised Structure Preserving (GSPREE) models in Small Area Estimation: Estimating ethnic group size for local authorities – Philip Clarke, Alison Whitworth Abstract Slides
Measuring uncertainty in ONS population estimates: Tools, techniques and outcomes – Louisa Blackwell, Katy Stokes, Paulina Galezewska Abstract Slides
Parallel Session 3
Towards the future of time series and longitudinal data analyses Chair: Frances Pottier (BEIS)
Anticipating Brexit effects in time series analysis – Jennifer Davies, Duncan Elliott, Charlotte Gaughan, Atanaska Nikolova Abstract Slides Paper
The impact of moving holidays in official statistics time series – Bethan Russ, Tariq Aziz Abstract Slides Paper
Increasing frequency and improving timeliness of unemployment estimates from the UK Labour Force Survey – Duncan Elliott Abstract Slides Paper
Parallel Session 4
Sampling and sample design: Optimising the allocation Chair: Aidan Mews (MoJ)
The 2016 Annual Business Survey: A sample re-design – Megan Pope, Jonathan Digby-North Abstract Slides Paper
Which business to price? Sampling Business Price Indices – Matthew Mayhew Abstract Slides
Standard errors from audits – Sumit Rahman Abstract Slides
Parallel Session 5
Modernising methodology through transformation, collaboration, and harmonisation Chair: Gareth James (ONS)
Transformation of methodology in ONS – the why and how – Gary Brown Abstract Slides
GSS Harmonisation Programme: Past, Present and Future – Ian Sidney, Becki Aquilina Abstract Slides
Lessons from Anthropology – Tacey Laurie Abstract Slides Paper
Parallel Session 6
Linking into administrative data: Infrastructure, method, and application Chair: Sumit Rahman (BEIS)
Addressing for Census and beyond – Alistair Calder Abstract Slides
Many-to-many linkage: Finding siblings in birth registration data – Christos Chatzoglou, Charles Morris, Shelley Gammon, Julie Mills, Lynda Cooper, Theodore Manassis Abstract Slides
Household effects in Big Data research: Privacy-preserving methods in evaluating a Welsh Government fuel poverty scheme – Karen Tingay; Sian Morrison-Rees;Sarah Lowe; Charles Musselwhite Abstract Slides
Parallel Session 7
Statistical outputs: Enhancing the user interface and protecting confidentiality Chair: Nick O’ Rourke (ONSG)
The methodology of maps – Bruce Mitchell, Jeremy Brocklehurst; Anna Harris Abstract Slides
Improving workforce analysis through geographical design – Mark Baines Abstract Slides
Disclosure Control: Applying Cell-Key Perturbation to 2021 Census outputs – Keith Spicer, Stephanie Blanchard, Iain Dove Abstract Notes Slides
Parallel Session 8
Big Data analytics: Opening new avenues of innovative research and understanding Chair: Owen Abbott (ONS)
Natural Language Processing (NLP) techniques for structuring large volumes of human text data – Alessandra Sozzi, Kimberley Brett Abstract Slides
Using machine learning techniques to analyse economic statistics: a case study with HMRC Trade in Goods statistics – Andy Banks Abstract Slides
Exploring mental well-being from prisoner case-notes using text mining – Jo Lee Abstract Slides Paper

 

Sponsors & Exhibitors:
Feedback from delegates also indicated that there was a wide spread of interest in the stalls laid on by the official sponsors and exhibitors of GSSM22: The Royal Statistical Society (RSS); Southampton University (MOffStat & MDataGov), The Learning Academy (the Analytical/Data Science Branch), and the Economics Centre of Excellence (ESCoE).

A full programme of the day’s events can be found here.

And Finally…

The GSSM22 Team would like to thank everyone who helped and contributed to the organisation of GSSM22, all of those who presented, and all of those who attended. All in all, it was a very successful day. We look forward to seeing you again next year.

Please send any correspondence to methodology@ons.gov.uk