Small area estimation of poverty - including expenditure poverty, kilocalorie consumption, and stunting and underweight in children
The World Food Programme Kathmandu, Nepal 2005-2006 This project involved the estimation of various poverty measures in Nepal using survey and census data, to provide poverty estimates for expenditure poverty (including poverty incidence, gap and severity), kilocalorie consumption, and stunting and underweight in children. These were estimated using a combination of survey data on poverty and census data on other variables to give poverty estimates at a much finer (i.e. ilaka) level than possible from survey data alone.
Yaks near Jomsom, Nepal (2006).
Small domain estimation for Maori expenditure patterns
Statistics New Zealand Wellington, New Zealand 2005
This project was a feasibility study to assess whether Maori expenditure patters for Consumer Price Index (CPI) levels such as food, clothing, transport, etc could be estimated for different ethnic groups using survey and population census data. A copy of the report is available here.
Small area estimation of poverty - expenditure poverty including poverty incidence, gap and severity
The World Bank Manila and Palawan, Philippines 2004-2005
The project included developing statistical models for small area estimation of expenditure poverty (including poverty incidence, gap and severity), and for poverty maps on a fine scale, using Philippines survey and census data. The method uses survey-based models to make predictions using census data using models containing variables that occur in both survey and census, followed by aggregation to small area level and the mapping of results.
Market in Manila (2005).
Development of sampling methodology for water and sanitation suppliers
The World Bank Hanoi and Ho Chi Minh City, Vietnam Baku, Azerbaijan 2004-2005
The project involved developing a generic sampling methodology that could form part of a toolkit usable in countries other than those surveyed as part of the pilot project. The pilot projects were undertaken as small scale national surveys in Vietnam and Azerbaijan. The focus was methods for determining the economic sustainability of the smaller scale water and sanitation providers.
Peace billboard in Hanoi (2005).
Essentially the same methodology was used in Azerbaijan.
Free rural water bore near Ujar, Azerbaijan (2005).
Small area estimation of indicators of poverty and nutrition in Bangladesh
World Food Programme, Bangladesh Bureau of Statistics Dhaka, Bangladesh 2003-2004
The project involved combining census and sample survey data to get better estimates of poverty and nutrition at a level sufficiently fine and accurate to be used as a basis for food aid allocation. The statistical methods are applicable to a wide range of situations (for example, see also: NSF/ASA Senior Fellowship project in Washington DC). The statistical analysis was linked with the mapping expertise at WFP to produce a series of fine level maps of various poverty measures for Bangladesh.
Boats at Dhaka, Bangladesh (2004).
Measurement of the Level of Entrepreneurship in Uganda
Global Entrepreneurship Monitor (in connection with United Nations Business Council) Kampala, Uganda 2003-2004
The project involved developing a survey design ab initio, developing a fieldwork manual and materials, providing fieldwork training, and advice on analysis, for a survey design on entrepreneurship in Uganda for GEM Uganda, all done at very short notice. The project required rapid development of co-operation with the Ugandan Bureau of Statistics, who were initially not involved in the project, so that UBOS personnel were willing to provide maps of sampled areas as well as local knowledge, at minimal cost. GEM currently undertake national sample surveys in 37 countries on a regular basis.
Settlement in Kampala, near Makerere University (2003).
Small area estimation of United States employment statistics
National Science Foundation / American Statistical Association / US Bureau of Labour Statistics Washington DC, USA 2001-2004
The research focused on finding ways of estimating the number of people employed within certain industries for small subsections of states within the USA. The research has been completed and solution is at present at implementation stage. When complete, the methods will allow for better planning for present industries and also for optimal siting of new industries to ensure an adequate workforce will be available.
Washington Memorial in spring (2004).
DP-SP Project: A Consultancy on Establishment Surveys in Thailand
World Bank, National Statistics Office of Thailand Bangkok, Thailand 2002-2003
The project involved assessment and review of the business establishment sample surveys undertaken in Thailand. It also required making recommendations for improvement and further development of these surveys both in terms of technical statistical improvements to design, implementation, and analysis, and also in terms of legal context within which NSO operates.
Steve with colleagues from the National Statistics Office of Thailand at Wat Phra Kaew (2003).
The appointment involved joint research with academic staff, and teaching courses in theory of linear models for the Hilary Term at the Department of Statistics, TCD. The main focus was theoretical research into conditions for equivalent solutions to linear models under different covariance structures, residuals for such models, the use of conditional values for observations rather than their deletion and the study of the question of when these two algebraic strategies yield the same parameter solutions. These questions have application to models fitted to sample survey data, simplified methods for determining estimates in models with random components, and to small area estimation methods.
Peace billboard in Hanoi (2005).
Free rural water bore near Ujar, Azerbaijan (2005).
Settlement in Kampala, near Makerere University (2003).
Washington Memorial in spring (2004).
Steve with colleagues from the National Statistics Office of Thailand at Wat Phra Kaew (2003). 
Gateway in Dublin (1999).