The quality of the contents of any survey is directly linked to the existence of validation procedures for the information initially recorded in the questionnaire. Actually, the most important part of the preparation work for the ESEE is that regarding the validation of the contents, which is made in close collaboration between the Fundación SEPI, the company which carries out the fieldwork and the companies which provide the information. This Note provides a very basic explanation of the procedure.

In general, the data recording and validation procedure is made up of four phases: the recording of the questionnaire, the validation of the contents, changing the blanks to zeros and further recordings.

The contents validations programs are a key phase in the procedure and up to five different aspects can be recognized: 1) answer, ranges, graph fulfillment and sum and order requirement controls; 2) unit controls; 3) consistency controls; 4) temporal consistency controls, and 5) financial controls. Of these, the most complex controls are those of consistency, temporal consistency and financial. They are made up by a battery with more than 100 controls regarding concrete questions of the questionnaire, which are the result of the refinement through experience of a wide range of contrasts aimed at verifying the consistency of the data. As a general rule, failure to comply with the consistency controls is not allowed. On the other hand, the temporal consistency and financial controls usually set up confidence ranges. In case of failure to comply, the data concerned are confirmed and, in case it persisted, the company which supplies the data submits the required justification. Fundación SEPI reviews and validates all these justifications.

Below are provided three straightforward examples regarding the three basic control types: of
consistency, of temporary consistency and financials.

Example 1 (Consistency control):

A16_4_1 = '1' ==> HA1_2 = '2' OR HA1_3 = '3'

Explanation: It provides non-manufacturing services, and it indicates neither to resell nor to provide services.

Example 2 (Temporary consistency control):

-25 < [(CO2007 - CO2006)/CO2006] * 100 < 25

Explanation: The cost per employee in this year (personnel costs minus indemnities paid, etc. divided by the average employment –averaged taking into account the special circumstances) presents, in comparison to the previous year, a very high variation.

Example 3 (Financial content control):

(G10_2 _2/1780) < 0,25 * G1 _9

Explanation: The personnel provided by temp agencies is, when expressed as a full-time work, very high (more than 25% of the workforce).