What steps does the CIS take to take a survey?

1) Survey design: The objectives, universe, sample and questionnaire.

2) Data collection or field work

3) Data processing

4) Results publication: How are the results of CIS surveys disseminated?


1) Survey design: The objectives, universe, sample and questionnaire.

The aims of the survey: What topics does the CIS research?

To begin to carry out a survey, it is necessary to define its aims: On which topic or topics do we want to learn the public's attitudes and opinions.

The CIS aims to gain a deeper knowledge of the attitudes and behaviour of Spanish society, and to measure its degree of stability and/or change over time. Since it was founded, it has taken nearly 2,000 surveys on widely ranging topics.

The definition of the universe (population of interest) and sample selection: Who is asked to answer the survey?

The survey universe is the population from which we want to obtain information. The population at whom CIS surveys are generally addressed holds Spanish nationality and is over 18 years of age, but sometimes they are addressed at the entire population residing in Spain (and they therefore include residents from abroad). It also carries out surveys where the universe is the young population (generally defined as that aged between 15 and 29) or some other subgroup of the population.

Given that it is impossible to interview the whole of the population, it is necessary to make a selection, take a sample of people on whom the survey will be taken. The CIS follows the necessary scientific procedures to obtain representative samples. It is crucial that these procedures be observed in order to be able to use the data obtained in the survey to describe and analyse the population from which the sample has been taken, always bearing in mind that there is always a margin of error (the size of which appears on each survey fact sheet).

Although the size and characteristics of the sample vary according to the aims of the survey, CIS surveys usually take a minimum sample of 2,500 individuals.

The writing of the questionnaire: How are the questions asked?

The questionnaires are made up of a series of questions which enable it to gather the public's opinions on different topics. In the case of the CIS, they are largely of a social and political nature. They furthermore include questions on the characteristics of the persons being interviewed (sex, age, level of qualifications, occupation), given that one of the aims is to learn whether there are differences in the opinions in each of those groups. Survey data files do not hold any information that can identify the people that have taken part in them.

CIS technical staff draw up the questionnaires based on thorough documentation which includes consultation of previous studies performed in the CIS and in other institutions, as well as those designed by international bodies devoted to carrying out surveys.

When designing the questions, special care is placed on them being understood by all the persons interviewed, irrespective of their level of qualifications or any other characteristic. Special care is also taken that they be balanced and not contain any kind of bias.

In most of the questions asked by the CIS, the answer options appear on the questionnaire and they are given a numerical code, which is what is recorded on the data file.

There are other questions on the questionnaire which are open, with no preset answer options, such as, that included in all the CIS monthly barometers, which asks the persons interviewed to express, in their own words, what they believe are Spain's three main problems. All these responses are then subsequently classified in the CIS in what is termed the “codification process”. It is thus possible to group together all those referring to “unemployment”, “housing”, “immigration”, in order to be able to count up the number of mentions each of these topics receives.


2) Data collection or field work

Data collection, also known as “field work” consists in passing the questionnaire in the people forming part of the sample.

The interviewers that take CIS surveys, go through the sections of the census included in the sample, following random routes, whereby they contact members of the public at home in order to gather their opinions on the subject of the survey.

The co-operation of these people in accepting to do the survey is fundamental for the survey results to collect opinions from all sections of the public.

The persons selected to take part in a CIS survey do so voluntarily and their collaboration is fundamental for attaining results that reflect the opinions of society as a whole.

The survey is not a test and there are no right or wrong answers: We seek their sincere opinion when they answer. At any time during the interview, the interviewee can choose not to answer a question if he/she wishes.

All the answers are anonymous, protected by the corresponding statistical secrecy and data protection laws. The answers are used as aggregates, with no individual references of any kind. No personal details or identifiers are kept on respondents and once the information they contain has been recorded, the questionnaires are destroyed.

The CIS supervises and inspects the interview process in order to assure the quality of the data collected. This is why it asks for the interviewee's telephone number, however this information is used exclusively for the CIS to be able to carry out inspections to verify that said person has, indeed, been interviewed.


3) Data processing

Once the data collection process has concluded, all the information obtained from the answers to the questionnaires is processed.

In order to learn the results and analyse public opinion on the subject of the survey, the information obtained has to be recorded. The answers the interviewees give to each question on a survey are converted into a data file or matrix structure.

In the files, each row corresponds to one questionnaire (one interviewee) and the columns contain each of the questions they have been asked. The specific answer that the interview has given to question on the questionnaire is put in the intersection of the row and column. The end file never contains information that could identify the interviewee.

Structure of a data file

Questionnaire no. Province Municipality Gender Age No. of children Appraisal of the economic situation Appraisal of state education (0-10) Appraisal of the work done by teachers Degree to which you agree with the contents of the Organic Law on Education
 0001 Murcia Ojós Male 18 0 So-so 3 So-so Totally agree
 0002 Barcelona Masnou Female 78 2 Bad 6 Bad Partially agree
 0003 Badajoz Mérida Female 45 2 Bad 7 Good Does not know
 2500 Zaragoza Calatayud Male 32 1 Good 5 Good  

The information recorded on the data files provides the survey results.


4) Results publication: How are the results of CIS surveys disseminated?

The CIS presents the results of its surveys both in documents that contain the data already processed and the matrices or files on which the data is recorded.

a) Documents that contain the data already processed (or tabulated):

  1. A “results” document, with the list of frequencies for each question, i.e., the number of times that each answer has been given and that percentage that this represents over the total number of persons who were asked said question.
  2. A “cross-reference” document for all the questions on the questionnaire with socio-demographic variables (sex, age group, habitat size, level of qualifications), political variables (ideology and voting history) and socio-economic variables (socio-economic condition and status). From these documents, for instance, it is possible to learn whether men have similar opinions to women, or whether there are differences between young and older people regarding the survey subject.

Initially, a provisional version of these documents is presented on the CIS website, in documents called “results previews”.

Subsequently, a thorough review of all the data is carried out to detect any possible error before the files becomes part of the CIS Data Bank. Once all the data has been reviewed and the verification, purging and anonymisation process has been performed, the study goes into the CIS Data Bank and the data file is then available to the public. The Law that regulates CIS operations stipulates specific deadlines for each of these steps, depending on the type of survey that has been performed.

b) Data files or matrices, showing all the answers given by interviewees, once the purging, validation and anonymisation process has been carried out

People who want to make an in-depth analysis of the results of a survey usually use the data file, as this enables them to cross any kind of information (questions), or, for example, in the case of surveys containing information on voting intention, to try to make different estimates of election results.

c) Technical documentation:

  1. A fact file with the main features of each survey: (Target population of the study, date field work was performed, size of the sample designed and the sample taken, characteristics of the sample design and, if applicable, information on the use of weighting coefficients for data processing).
  2. The data files or matrices are accompanied by the literal questionnaire used and the code book.
  3. On some occasions, when the surveys taken are particularly complex, a methodological report is provided which explains specific technical features of the survey.