ESSNet on Statistical Disclosure Control

Task 1. Dissemination of CENEX results


1.a Extending the handbook

The CENEX project produced a first version of an SDC handbook. It was foreseen that the final handbook would not be achieved within the one year project. The aim of this task will be to develop the handbook further to achieve a more final version. The work for this task will be divided into two sub-tasks.

1.a.1.: Consultation of NSIs.

1.a.2.: Improve the handbook with more examples and recommendations

1.a.1. Consultation of NSIs.
One partner (NL) will coordinate this consultation. We foresee that we will ask several NSIs for their general opinion of the handbook and in addition we will ask a selected group of NSIs (partners in the ESSNet) to study the handbook in more depth and provide us with their findings. We will devote the first project meeting in May/June 2008 to discuss the results of this consultation. This work and the meeting will lead us to an (internal) report with recommendations for improvements. Based on these conclusions we will then improve the handbook.
In year 2 of the project we will again ask the testers their opinions of the improved handbook. After that test-round we will have input for completing the last version of the handbook in this project.
Partners: (NL, SE, PT, SI, BG, HU, RO, LU, NO).
The Czech office showed serious interest in this project and will contribute, but will not become a partner.
The Lux. NSI (STATEC) can only contribute to this topic in year 2.
Deliverables: Test reports by the testing countries after 4 months in year 1 and also after 4 months in year 2
1.a.2. Enhance the handbook: Examples and recommendations
In the field of Statistical Disclosure Control it is not always easy to give straightforward advice to non-expert users. A good SDC-plan for a specify survey or publication always requires more expert knowledge. Therefore it is not always easy to give a simple recipe without knowing the specific local situation. E.g. when protecting a survey data set external factors like whether a register is available should be taken into account. Simple rules can only be given for certain situations. Nevertheless we will do our best to include recommendations where appropriate and more examples and case studies to make the handbook more relevant.
The reports produced in subtask 1.a.1 will guide us to improve and extend the handbook.
The aim at the end of the ESSnet project will be that a more stable version of the handbook will be ready. We are reasonably confident that a publisher can be found to print an ESSNet SDC handbook. Some funds to facilitate this process will be allocated in the budget.
Partners: (NL, UK, IT, DE, SW, URV)
Deliverables: An intermediate version of the handbook after year 1 and a final ESSNet version of the handbook after year 2.

1.b. Follow up of the CENEX inventory

1.b.1. Addressing the needs of less experienced countries

This task will involve a detailed study of the inventory results in order to draw preliminary conclusions. In addition a few countries identified as in need of support will be contacted and the results of those consultations will inform for other parts of this project. Invitations to the courses, visits and/or changes to the software could be a result of the actions here.
Partners: (NL)
Deliverable: A report of these activities will be produced in year 1

1.b.2. Tailored courses and follow up visits.

A continuous effort of training in the SDC field for NSI staff is essential. The basis of a good standard of SDC-work across the NSIs is a good understanding of the fundamental issues and a good knowledge of the tools available. It is too risky to give these tools to non-experienced users. We already have a tradition of SDC-courses, a successful training event was held in Budapest as part of the CENEX project. However bringing together 25 participants from almost an equal number of member states leads to very isolated people with some SDC-knowledge. We will try to improve on this approach by organising courses for a selected group of countries. In the ESSNet we will offer a course/workshop for 4 or 5 countries. Each country will be allowed to send 4 or 5 participants. Part of these training sessions will be a course along the lines of the CENEX course, but also we will allocate time to a practical training session when the participants will use their own datasets and have the opportunity to use the software on their own data.
The general setup of these training sessions will be a two-day session on micro data followed by a two day session on tabular data. The larger part of the second days will be devoted to practical work on the participantís datasets and tables. By extending the course to 4 days and training 4/5 participants per country we expect to achieve a greater impact. We plan one training event per year during this ESSNet.

Follow-up visits might look promising, but it is a great burden on the shoulders of the small European group of SDC-specialists. If needed such a visit could be made, but alternatively it could be more effective the other way round, with the countries visiting the ESSNet members. Alternatively a help-desk might solve many problems.
Partners: NL, IT, DE
Deliverables: SDC-courses one in each year. If needed SDC-trainees from countries will be received in our offices and two visits (one each year) by Istat will be made to promote the Italian risk approach.