Imputation and Social Indicators: The Use of Factor Analysis for Imputing Missing Data

Thursday, 14 July 2016: 09:00-10:30
Location: Hörsaal 12 (Juridicum)
RC55 Social Indicators (host committee)

Language: English and Spanish

The proposal is to use factor analysis using either quantitative and qualitative variables with the aim of proposing techniques, models or approaches for the imputation of missing data.
There is extensive literature and tradition addressing the problem of missing data where different types of imputation procedures are proposed. The procedures vary over time, from those initial decisions of a removal or replacement of records without information to sophisticated multiple imputation procedures.
The first contributions were made in 1932 by Wilks who proposed the replacement of missing data by the data available on the variable to be imputed. However, it was not until the 1970s when the systematic study and formalization start from a probabilistic point of view, mainly highlighting the work of Rubin in 1976 “Inference and Missing Data”, then ordered and systematized by Little and Rubin in their 1987 book Statistical Analysis with Missing Data.
Making new contributions from the field of multivariate techniques and systematizing the productions already made so far are a necessary task that we intend to undertake in this session.
Session Organizers:
Sandra FACHELLI, Autonomous University of Barcelona, Spain and Pedro LOPEZ-ROLDAN, Autonomous University of Barcelona, Spain
Aproximaciones a La Imputación De Ingresos Desde Los Estudios De análisis De Clase. Una Propuesta a Partir Del Uso De Técnicas De análisis Factorial.
José RODRÍGUEZ DE LA FUENTE, Instituto de Investigaciones Gino Germani - UBA, Argentina; María Clara FERNÁNDEZ MELIÁN, Instituto de Investigaciones Gino Germani - UBA, Argentina
See more of: RC55 Social Indicators
See more of: Research Committees