Constructing, translating and validating a questionnaire. The example of EUROQUEST

Questionnaires are information gathering tools that are often irreplaceable in research. Multidimensional scales are best adapted to investigating diseases or disorders involving several components (such as depression or deterioration of the quality of life). These scales have been developed by psychosociologists, and are increasingly being employed on account of the fact that biomedical studies are branching out to encompass psychosocial fields.
Noting that the transfer of knowledge and experience between the psychosociological and biomedical disciplines has received little attention, both as regards the drafting and evaluation of questionnaires, INRS decided to compare them. This attempt at synthesis yielded rich results and allowed principles for drafting questionnaires to be defined along with a translation methodology that guarantees the permanence of the concepts, whatever the language. In addition, the methodology enabled both qualitative and quantitative assessment methods to be compared. Although the authors are unanimous on the necessity of measuring the reliability and the sensitivity of the validity of a questionnaire to change, there is no consensus on the concept of validity and of the statistical methods for the investigator.
Testing these methods on EUROQUEST (questionnaire on the neurotoxic symptoms of employees exposed to chemical products with effects on the nervous system) revealed the shortcomings of certain standard published methods in this respect. Calculation of the odds ratios between two items and ascending hierarchical classification, on the other hand, both turned out to be rewarding and fruitful methods.