Purpose of the study: Identifying critical aspects in the design of the Web questionnaire related to the 2013 Italian survey on Information and Communication Technology (ICT) usage and on E-Commerce in Enterprises, and evaluating if possible difficulties in completing the questionnaire may be attributed to some structural business characteristics.
Design/methodology/approach: Detailed client-side paradata were collected during the ICT survey. On the basis of the available paradata, a set of indicators was created for each respondent company (e.g. connection time, number of days in which the connection occurs, number of visited questions). Besides, the resulting indicators were merged by enterprise with statistical information coming from the ICT survey itself and from other sources.
Findings: With reference to each navigational path, mean and percentiles of the proposed indicators indicate the most critical questions and sections of the Web questionnaire and suggest the measures to be taken to improve the accuracy of the answers. The weakness of the relationships between the proposed indicators and the structural business characteristics considered in this study (number of persons employed, turnover, etc.) is in line with the results from a previous analysis related to the Italian business Research & Development survey.
Originality/value: This can be considered as a first step towards the systematic production and use of paradata in Istat. The study has helped to define basic paradata and indicators to be produced in the framework of the new Italian Business Statistical Portal (BSP), a platform dedicated to the Web-based collection of statistical information from enterprises.
Research implications/limitations: As the BSP involves very different business surveys, standard paradata defined in that framework cannot cover all information needs and ad hoc paradata can be produced only in case of major changes in questionnaires.
Practical implications: The results from this study have been used to improve the ICT questionnaire only to some extent, as it is harmonized at the European level, and to better understand the quality of the survey data.