Update of the underlying principles of the EurOccupations database: third reporting period: D1a1

In contrast to other survey modes, web-surveys allow for using a closed response format
for survey questions with a large number of categories, such as car types, camera types,
countries, regions, industries, and occupations. In other survey modes, these questions
require an open response format, followed by field- or office-coding. Web-surveys allow
the use of drop-downs, for example for an alphabetically sorted list of 200+ countries, or
for the use of search trees. A search tree is the most appropriate way for selfidentification
of occupation, which is a key variable in socio-economic research.
This paper summarizes the design principles underlying a Database of Occupations for
the measurement of occupations in multi-country web-surveys by means of selfidentification.
The Database was compiled as part of the FP6 funded EurOccupations project
(2006-2009, no 028987). The paper introduces the use of the occupation variable in
datasets, in particular multi-country datasets, and explores the reasons why the Database
has been designed. Using a variety of sources, the Database includes almost 1,600
occupational titles, which is judged to be sufficiently detailed for valid-self-identification.
The Database consists of a source list, a search tree and translations into the languages
of almost 50 countries. It is explained how the source list has taken into account issues
such as skill levels, corporate hierarchies, job ladders, managerial and supervisory occupations, craft versus manufacturing occupations, composite occupations and the methods used for translations. The requirements posed by web-surveys, such as readability, alphabetical sorting, coping with the category "Other" are discussed. Finally, the paper details how the Database is used in the worldwide WageIndicator web-survey. For more information see www.euroccupations.org or www.wageindicator.org.