Previous Events

This site provides a summary of seminars held since the establishment of the WGCI.

20 November 2001: Multiple Imputation and Variance Estimation in the Environmental Sustainability Index. The seminar focused on the Environmental Sustainability Index (ESI), developed by the Yale Center of Environmental Policy and Law at Yale University, the Center for International Earth and Science Information Network (CIESIN) at Columbia University, and the World Economic Forum's Environment Task Force. The conceptual framework of the ESI consists of a 2-step procedure combining a wide set of environmentally relevant variables to key a smaller group of indicators, which are allocated to five dimensions considered to define environmental sustainability. The final index is then used to rank countries according to their progress towards environmental sustainability. The methodological questions addressed in the seminar were How to deal with missing values? using Markov Chain Monte Carlo simulations and How to derive a measure of variability for the ESI? based on the analysis of multiply imputed data.

23 January 2002: An Index assessing Progress towards the Millenium Goals based on n Estimated Distance Measure. The seminar introduced the theoretical framework for the development of a composite index based on the idea of estimating the cumulative distance from a set of targets/goals applied to the General Assembly's Millenium Declaration Goals. The strategy adopted is to use a variance-minimizing estimator of the cumulated squared distances between the current values and specified target values, where the target value corresponds to a specified percentile of the distribution of the observed current values for the target indicator.

12 February 2002: An Index assessing Progress towards the Millenium Goals based on n Estimated Distance Measure (contd.). The seminar continued the discussion on the distance based index with the practical application to the dataset of the Millenium Declaration Goals. It explored the resulting indices for a selecion of developing countries and highlighted its theoretical properties.

27 March 2002: Sequential Regression - A Method for Multiple Imputations of Missing Data. The seminar was devoted to the introduction of sequential regression multiple imputation (srmi), a technique for estimating missing values using a series of regression models to sequentially impute missing values. The reiteration of the process leads to a set of complete datasets, which are analyzed using standard statistical methods and the results are combined to final estimates of the parameters of interest. The advantages of srmi as compared to MCMC imputation are that it is less restrictive in its model assumptions and utilizes the intuitive concept of imputation through regression. The method was applied to one of the datasets used to define Least Developed Countries.

7 August 2002: Practice Session for the JSM. This seminar provided an opportunity for the members of UNSD to present their papers for the Joint Statistical Meetings of the American Statistical Association, 11-15 August 2002 in New York, to their colleagues in UNSD and DESA.

11-15 August 2002: Joint Statistical Meetings organized by the American Statistical Association, New York. The United Nations Statistics Division held a session titled Some Statistical Problems in International Statistics at the JSM 2002 in New York. Three papers were presented: Statistical Indices-Construction, Interpretation and Use in Measuring Development; A System-Approach to Environmentally-adjusted Macro-aggregates; and Deriving a Model for Missing Data in the 2002 Environmental Sustainability Index. The papers were discussed by Prof. Kjell Doksum, University of California Berkeley, and Prof. Thomas A. Louis, Johns Hopkins Bloomberg School of Public Health.

24 September 2002: Operating Characteristics of Surveys in Developing and Transition Countries. The seminar explored the role of operating characteristics such as design effects, costs, and response rates of surveys in developing countries and countries in transition. In recognition of the importance of these operating characteristics in the planning processes of surveys, the United Nations Statistics Division is currently finalizing a publication on the subject. The seminar introduced the publication, in particular chapter 12, and illustrated an example on the portability of design effects across countries using data from the Demographic and Health survey series.

31 October 2002: Measuring Governance and the Relationship with Human Development. The new Human Development Report 2002 - Deepening democracy in a fragmented world has recently been published by UNDP. The report focuses on the much-talked-about issue of good governance and analyses global trends in human development as it relates to political freedom and participation, justice, accountability, and personal security. The seminar discussed some statistical aspects arising in the attempt to measure and quantify the level of governance achieved worldwide and explain the process of indicator selection in this context. Discussants were Ms. Sakiko Fukuda-Parr, Director of the Human Development Report Office, UNDP, and Mr. Willem De Vries, Deputy Director of UNSD.

For further information please contact:
UN Statistics Division
Environment Statistics Section
DC2-1410, TWO United Nations Plaza
New York, New York 10017, USA
E-mail: contact WGCI
Tel: +1-212-963-8564