Andy Turner's NCeSS Agenda Setting Workshop on Combining and Enhancing Data Web Page @ School of Geography, University of Leeds

Introduction
- This Web Page is for organising my information about the NCeSS Agenda Setting Workshop on Combining and Enhancing Data. Manchester, UK. 2007-01-22.
- Contents:
People
- Delegates
  - Andy Turner
  - + Others
Notes
- Key note by Ian Diamond ESRC
  - Cost of obesity in the UK ~�6Billion per year
  - "Methodology" is a word commonly used in place of "methods". I think this assists methodological research.
  - ESRC are keen to fund risky exciting research proposals.
  - UK Longitudinal Household Survey
    - Sampling and Data Collection Strategies for the Proposed UK Longitudinal Household Survey
  - ESRC committed to Virtual Safe Laboratories
    - I think I've had this idea before (perhaps at a NIEeS GridGIS meeting): Why not digitise and retain name and address information from the 2011 Census so it can be linked for epidemiology research. A scoping study and the development of a service...
  - Questions and feedback session
    - ESRC MRC DoH Welsh Office etc Public Health and obesity focus
    - Administrative data vaguaries - Nobles group at Oxford?
    - I was thinking about ESRC and ONS meta data standards adoption: Is ESRC encouraging the use of dublin core?
- How to represent uncertainty...
- Data mining
  - Correspondence - term used when variables are the same, but are recorded differently (e.g. age in CAS tables)
- There are privacy concerns with releasing 'synthetic data' containing 'synthetic individuals' "too similar" to "real individuals".
  - In MoSeS we can absolve responsibility for this by disseminating ID lists.
- Chat with Graham Squires over lunch
  - Urban Regeneration or Managing the Decline of Industrially Restructuring Cities: Low Demand Neighbourhoods and Housing Market Renewal
- Spatial Microsimulation
  - Dimitris Ballas
  - Wilson and Pownal?
  - With population data created in MoSeS we can generate multiway tables tables of census variables that are not available as CAS tables in the usual offering.
  - Should look at SimBritain and MicroMappas and collaborate with Dimitris
- Using Multiple imputation to integrate and disseminate microdata...
- Chat with Peter Halfpenny
  - I should talk to Keith Cole about work with R stats package use of Grid Enabled Census Data
  - Peter keen for MoSeS to collaborate with Dimitris
- Multi-level modelling - Getting the most from census data
  - Good stuff :)
  - Nick Best pointed out the relevance of Brown and Harvey's work
  - Simon Peters suggested a search for spatial autocorrelation aspects of covariation (e.g. at OA level). This is a good idea :)
  - This made me think that in MoSeS population initialisation we could optimise based on minimising variance and covariance...
- Bias: methods for combining data
  - NCRM link
- The General Household Survey (GHS)
- Lisa Buckner
- Jerry Reiter
- ReDRESS Learning Space Catalogue
- I'm having Deja Vu about having Deja Vu
  - I think I may have even noted that I was having Deja Vu last time, so I will search my blog and try to find out when this was... Hmmm no success...Anyway, Deja Vu is wierd.
- LINKING HOUSEHOLD SURVEY AND ADMINISTRATIVE RECORD DATA: WHAT SHOULD THE MATCHING VARIABLES BE?

Andy Turner's NCeSS Agenda Setting Workshop on Combining and Enhancing Data Web Page

Introduction

People

Notes

Validation and Metadata