REDATAM+SP WebServer as a tool for dissemination of microdata REDATAM+SP WebServer as a tool for dissemination of microdata Alejandra Silva CELADE- Population.

REDATAM+SP WebServer as a tool REDATAM+SP WebServer as a tool for dissemination of microdatafor dissemination of microdata

Alejandra SilvaCELADE- Population Division of ECLAC, United Nations (1957 – 2007)

IPUMS- GLOBAL WorkshopLisbon, Portugal, August 2007

What is Redatam?What is Redatam?

It’s a database management tool that administrates large volumes of census microdata with hierarchical (geographical) structure down to the smallest area of the census exercise (city blocks)

Facilitates the data processing for small areas

How? by facilitating the NSO the creation of an encrypted database which could be

provided to third parties taking into account confidentiality issues

by providing the NSO’s themselves and the end users with a user friendly and fast software for processing large volumes of data, particularly for small areas

What is Redatam?What is Redatam?

It was originally (early eighties) developed at CELADE (Latin American Demographic Center/Population Division of ECLAC, United Nations) to promote access to census micro data.

The development of REDATAM was supported mainly by IDRC, CIDA, IDB and UNFPA.

REDATAM: REtrieval of census DATa for small Areas by Microcomputers

Which data can be used with Which data can be used with Redatam?Redatam?

Currently, all data sources can be imported into Redatam

- Population, agricultural, and economic censuses - Vital statistics - Household and fertility surveys - Educational and health statistics - Trade & commerce data, etc.

Characteristics of REDATAMCharacteristics of REDATAM

User friendly software

Administrates Hierarchical databases

Data stored in internal (encrypted) format (no access to individual or household records)

Highly Compressed database

Fast data processing

Multisectoral databases (combining several databases)

User can define geographical area to be processed

Hierarchical Structured DatabaseHierarchical Structured Database

HouseholdHousehold

Country

Province

Municipality

. . .

PersonPerson

DwellingDwelling

Block

Exportation of (aggregated) results to other software (Excel, ASCII,

GIS)

Thematic mapping & Graphs

External data in dbf format can be accessed

On-line Help

Dissemination of Applications on CDs (xPlan) for interactive processing

with the microdata as well as aggregated data.

On-line data processing (Redatam WebServer)

Available in English, Portuguese, French and Spanish

Characteristics of REDATAMCharacteristics of REDATAM

ProcessProcessCreateCreate

xPlanxPlanWebServerWebServer

REDATAM+SPREDATAM+SP FAMILYFAMILY

EIM & EIF EIM & EIF Demographic modulesDemographic modules

GINI moduleGINI module

Redatam+SP xPlan ApplicationsRedatam+SP xPlan Applications

• Redatam+ SP Stand Alone applications for population analysis• Redatam+SP xPlan applications for population data dissemination

Redatam Applications for Redatam Applications for Population AnalysisPopulation Analysis

The Redatam+Sp EIM and EIF Demographic Modules facilitates the indirect estimation of mortality and fertility (stand alone applications)

The Redatam+Sp GINI module designed for the estimation of inequality measures (Gini, Theil and other indicators)

Was developed to calculate child and infant mortality rates using the Brass method.

This method uses census microdata to indirectly estimate the mortality rate by utilizing survival information derived from the difference between the total number of children born alive and the number of children presently alive.

This information can be obtained from all population censuses or from specific health surveys in all the countries.

Indirect Mortality EstimationIndirect Mortality Estimation(Redatam+SP EIM)(Redatam+SP EIM)

The Brass Method to study of the fertility and mortality of a population

This method requires then the following variables: Number of women by 5 years age groups, between 15 and

49 years (total, women without declaration of children ever born, and/or children missing);

Number of children ever had, classified by the mother’s age group.

Number of children surviving, classified by the mother’s age group

Indirect Mortality EstimationIndirect Mortality Estimation(Redatam+SP EIM)(Redatam+SP EIM)

The Redatam+SP EIM module requires in the first place completing a parameters file named “EIM.inl”

Indirect Mortality EstimationIndirect Mortality Estimation(Redatam+SP EIM)(Redatam+SP EIM)

In the case of Brazil, 1991 and 2000, the mortality rates for every female cohort were estimated using the entire database and also a control variable was introduced for households to segregate according to socioeconomic quintiles.

In both cases, a weight was used because this census data correspond to a 10% sample of the population.

Once all the parameters are defined in the EIM.inl file, we are ready to run the Redatam+SP EIM module.

Indirect Mortality EstimationIndirect Mortality Estimation(Redatam+SP EIM)(Redatam+SP EIM)

Indirect Mortality EstimationIndirect Mortality Estimation(Redatam+SP EIM)(Redatam+SP EIM)

Main window with the basic parameters used to run the model

Output window with the mortality rates for each female cohort, for Brazil census1991and 2000.

The mean parity P(i), that is, the average number of children that women in each group of age ever had. A woman aged 45-49 at the end of her reproductive life has had an average of 4.22 children according to the 1991 census and 3.39 children according to the 2000 census

Infant mortality rate calculation by Quintiles Infant mortality rate calculation by Quintiles Brazil, 1991 censusBrazil, 1991 census

The Brass technique (using the Coale-Trussel variant) is utilized. Then, the q(x) values are interpolated into the West family of the model life tables of Coale & Demeny (column 9) and the corresponding mortality rates - q(1) and q(5) values are obtained.

Infant mortality Infant mortality drop in Brazil, drop in Brazil, according to according to

different sourcesdifferent sources

Brazil has reduced its infant mortality rate significantly in the last few years. Infant mortality has decreased across all socio-economic levels of the population

(quintiles), controlled by area of residence (urban versus rural) showed by the negative percentages of variation.

The magnitude of this reduction has varied in each level, reflecting social inequalities.

Infant mortality rate calculation by Infant mortality rate calculation by Quintiles Quintiles

Brazil, 1991 and 2000 censusBrazil, 1991 and 2000 census

However, it shows a large intraregional variation which, moreover, failed to decline at all in the 1990s.

Infant mortality rate calculation by Infant mortality rate calculation by Quintiles Quintiles

Brazil, 1991 and 2000 censusBrazil, 1991 and 2000 census

As a conclusion we can say that the recent drop in infant mortality in the country (as in other countries of Latin America) has not been accompanied, at least not systematically, by a decrease in the gaps between administrative units.

Redatam xPlan Applications for Redatam xPlan Applications for Microdata Dissemination – The Microdata Dissemination – The

FutureFuture Redatam Development team is focusing more and more into

the development of special applications to disseminate population data (for example on CDs or through the Web) using microdata or aggregated data.

Easy and friendly interface; without programming. Options of tabulations and processes are defined by the user Allows to process predefined indicators. Allow for graphing and

thematic mapping. Several applications have been undertaken with Census data,

vital statistics, industrial census, agricultural census, trade statistics.

CELADE, Population Division of ECLAC celebrates this years its 50th anniversary. And it will continue for many more years giving support to all countries in Population & Development issues.

Thank you!Thank you!