Session 19

Back to Schedule

Title of session: Quality standards for statistical domains

Chair: Alexandru Gherasim

Room: S3B Sukiennice

Time: 08:30 - 10:00

Date: 29 June

Session 19 - papers & presentations

Presenting AuthorAbstract
Maja Rynko
Title: <<< Quality of ICT skills indicators >>>
Basic ICT skills are a prerequisite for social inclusion in the information age. Level of population ICT skills as well as computer and Internet usage are of interest to the European Commission and the national governments. This is reflected in numerous strategic documents (e.g. Digital agenda for Europe, A new skills agenda for Europe, Long-term National Development Strategy for Poland). The EU survey on ICT usage provides a proxy on the level of ICT skills among individuals in Europe. However, it is based on individual’s self-assessment which may be prone to misjudgment (e.g. Dunning et al. 2004). Although potentially more accurate, direct assessment of skills poses a big challenge to survey design, fieldwork and budget planning. Using data from the Polish follow-up study on the Programme for the International Assessment of Adult Competencies (postPIAAC) we compare self-assessment of basic ICT skills to their direct assessments. The postPIAAC was conducted between Oct. 2014 and Feb. 2015 on over 5 thous. respondents aged 18-69. The basic questionnaire included Eurostat questions on computer and Internet usage. Respondents were also asked to conduct several tasks directly comparable to the Eurostat questions: coping files to a folder, using copy/cut/paste tools for text editing and using basic spreadsheet formulas. Among those who declared to have performed these tasks before, many did not complete the tasks in the direct assessment (13%, 8% and 65% respectively). The preliminary multivariate analysis suggests that young and higher educated people are more likely to overrate their skills in the self-assessment. The discrepancy between the actual and the declared level of ICT skills poses a question of reliability of ICT indicators based on self-assessment.
Valeriya Angelova-Tosheva
Title: <<< The method of harmonised Labour Market Areas in Europe >>>
The Labour Market Area (LMA) is a well-established and discussed concept in regional geography and statistics. The concept of LMAs has existed for almost 30 years, based on different definitions and known under various names (Labour Market Regions, Employment Zones, Commuting Zones, Travel-To-Work-Areas, Daily Urban Systems, Working Catchment Areas etc.). The need for functional geographies in the statistical practice is indisputable. Administrative boundaries often break up single LMAs. Commuting across NUTS and country boundaries can lead to significant differences between total employment (job-place-based) in a region and resident working population (domestic employment) in the same region. Indicators such as GDP per inhabitant will be affected in regions with asymmetric commuting patterns. Luxembourg, Inner London and Brussels are only a few examples of territories where employment and GDP data are distorted when presented as divided by inhabitants. For the last ten years, Eurostat has intensively worked on the EU-wide harmonisation of the LMAs concept starting with a study together with the research community to investigate the value assed, feasibility and best practices in the EU. A Task Force continued the work to make an official proposal of the Commissions position on harmonised LMAs. The approach for delineation of LMAs proposed to the Member States is a simple, transparent, reproducible, consistent, and policy independent bottom-up method that needs only commuting flows as input. As a following-up action, in the frames of a grant programme several countries tested the IT tool and proved the feasibility of implementing the harmonised method. This article aims to present and discuss the method in details, as in the future it will be the basis of creating LMAs as an operational service in a joint ESS initiative.
Ioannis Xirouchakis
Title: <<< Improving the quality of Business Statistics through Profiling >>>
Economic globalisation and the way multinational groups organise themselves have led to increasingly complex organisations and to a growing gap between their legal and economic structures. Equating the statistical unit to the legal unit has resulted to two biases: to the impression that new enterprises have been created and to double accounting for non-additive variables (such as turnover). Such biases affect the quality of economic statistics. Confronted with this common problem, several European national statistical institutes decided to go beyond the legal definition of the legal unit and to implement enterprises in the economic meaning. This distinction is commonly done by "profiling". National profiling increases the relevance and quality of business statistics. First, profiling of the most significant groups relies on establishing communication between dedicated teams within statistical institutes and group representatives. Besides, it aims at suppressing "double-counting" for non-additive variables and at delineating the economic activity of groups on the national territory. Large groups are particularly internationalized, having affiliates in different countries. Therefore, cooperation between statistical institutes is required to understand their business and structures. Since 2013 Eurostat has been supporting projects with Member States to test European collaborative profiling and to agree on a common methodology based on test results. In 2018 more than 300 multinational groups have been collaboratively profiled. The main results in terms of quality improvements include (I) consistency of the attributes of multinational groups across national borders; (II) better understanding of the groups' activities, contributing to their consistent view and treatment; (III) more accurate evaluation of their footprint on the national economies and consequently (IV) improved quality of national statistical business registers and of the EuroGroups Register containing the biggest multinational groups present in Europe. Finally, the paper investigates examples for improving data and (subsequently) statistics quality using national and European Profiling.
Thomas Balcone
Title: <<< Profiling : a new way to increase the quality of statistics on research and development >>>
Currently, statistics on Research and Development (R&D) carried out in the business sector are computed in France on the sole basis of legal units: firstly, a survey is addressed to them to collect the data and then, statistics on R&D are disseminated at legal unit level. Considering the increasing importance of the enterprise group in the French economy, it seems difficult today to go on using only the legal units to calculate business statistics. Indeed, assimilitating the legal unit to the enterprise is not relevant anymore for group's affiliates and subsidiaries. Taking into account the European definition of an enterprise will help to disseminate more consistent and relevant R&D statistics on the business sector. The French business statistic register established by the French national statistical institute (INSEE), called SIRUS, contains notably all the legal units, all the enterprises and all the links between them. The main contribution of this register is to make possible the calculation and dissemination of statistics at an other level than the legal unit one: the enterprise level. This article first describes why the data should go on being collected at legal unit level and not at enterprise one. Indeed, it seems that such a change in the data collection can be dangerous because it could result in a substantial increase of the response burden. Then, this article presents the process based on SIRUS that leads to the computation of key indicators on R&D at enterprise level. To conclude, it compares these key indicators with the ones calculated at the legal unit level to show the impact of moving to the enterprise level on French R&D statistics.
Denisa Florescu
Title: <<< European structural farm statistics — new quality rating system >>>
Eurostat, together with statistical bodies in the ESS, have adopted a quality rating system that guides how structural farm statistics derived from farm structure surveys are disseminated. The system does this by showing when the estimates are sufficiently reliable to be published, with or without warning. It is based on:
• coefficients of variations for totals and means of continuous variables;
• standard errors for proportions and counts.
The paper will also present the work carried out to harmonise methods and application of variance estimation methods within the ESS. To consistently apply the new quality rating system, Eurostat and the national statistical bodies must compute roughly the same variance estimates. Future structural farm statistics will come from the data collected from ‘Integrated Farm Statistics’, based on a modular approach. This will lead to more complex national sampling designs. Also, the paper will outline ongoing developments to integrate additional sampling design information specific to national multi-stage sampling in the estimation of variance. It will aim to accurately compute variance in Eurostat. The paper will also introduce new quality reporting based on the European Standard Quality Reporting System (ESQRS) template. This will greatly help to assess all quality dimensions, improving the quality of EU data and metadata. Farm structure surveys are the main source of information on the current state of and trends in agriculture required to monitor the common agricultural policy and other EU policies. High-quality data is essential for decision-makers.

Back to Schedule

Font Resize