Centro de Investigaciones Sociológicas - CIS
Verify CIS extension
Sede Electrónica
ES CA GL EU FR
The use of algorithms for automatic translation may affect the accuracy of the presented content
whastapp linkedin x email

Sample

To create a sample, you must define the reference population (which we call the universe) and make a representative selection from this population (which we call the sample).

The reference population for CIS studies is provided by the National Institute of Statistics (INE) using data from the population register, which provides information on the population by variables such as sex, municipality, nationality, and age. We then transform this data into the strata of sex, age range, and municipality size by autonomous community used in the surveys.

Sample design, that is, the selection of people for the survey, is crucial, as it allows the data obtained to be used to describe and analyze the sample population, the universe. It is important to keep in mind that a sample has a margin of error, the magnitude of which is established in the technical sheet that accompanies each study and specifies all the technical characteristics.

 

Weighing

The sample is weighted by comparing the sample composition with an appropriate delineation of the sample population. This delineation of the sample population is provided by the National Institute of Statistics (INE). Based on this, a national weighting procedure is applied, using marginal and inter-cell weights. In this way, minimum variables such as sex, age, NUTS II region (basic regions defined by the Eurostat nomenclature of territorial units for statistics), educational level (using the classification of programs, qualifications, and certifications by educational levels attained (CNED-A)), and locality size are introduced into the iterative procedure. This post-stratification weighting is also known as 'correction weighting' or 'total non-response weighting'.

The CIS dataset always provides two types of weights: a post-stratification weight at the national level (PESO) and a post-stratification weight at the autonomous community level (PESOCCAA). Depending on the sample size, some studies also offer a post-stratification weight at the provincial level (PESOPROVINCIA).

The mean values of the coefficients of the post-stratification variable at the national level (PESO) are attached at the end of the technical sheets.

For the estimation at the level of each autonomous community, the microdata file also includes the weighting for each of them (variable PESOCCAA).