The European Network for Better Establishment Statistics is inviting academics, methodologists, experts, and users to participate in the workshop on
The Unit Problem in Business Statistics Methodology
Time Monday, November 10 2014, 08.30 – 16.30
Venue UNECE, Palais des Nations, Geneva, Switzerland
The unit problem arises when the statistical units of interest are not given and need to be constructed. The units available in administrative sources are typically created for tax or legal reasons and are not always equal to the statistical units. This is a particular concern for big and complex businesses that tend to contribute most to the statistical estimates.
The creation and maintenance of the statistical units, including profiling, is a core task often given to a separate business register department. Whatever the methods used, potential errors and effects are far from transparent and are not included in the uncertainty evaluation of the statistics produced. Misidentification of statistical units causes unit error, which will propagate to all statistics that involve the erroneously profiled or classified units.
The unit delineation and related issues require a statistical approach in order to bring them into the total error framework. This is the broad aim of this one-day workshop. Immediate goals are to form a common understanding of the unit problem, to share the related expertise across the disciplines, and to outline the important issues for development and research.
We invite survey methodologists, register specialists, academics in statistics, organisational sciences, economics and other related areas, as well as users of business statistics, to attend the workshop and take part in exchanging ideas. After each presentation, ample time for that is allocated. The workshop will close with an extensive, moderated discussion session, aiming to catalyse directions for further work.
All presentations given at the workshop have been uploaded to the website, and are linked from corresponding entries in the programme below.
A summary of the discussions held at the workshop is now available:
08.40-10.05 Session I: Wider perspectives
10.05-10.25 Coffee break
10.25-12.00 Session II: Statistical approaches I
12.00-13.00 Lunch break (participants making own arrangements)
13.00-14.40 Session III: Statistical approaches II
14.40-15.00 Coffee break [Poster: Boris Lorenc, SCB: An outline of a cognitive analysis of the profiling task]
15.00-15.25 Session IV: Longitudinal perspectives
15.30-16.30 Discussion, summary and way ahead
(moderated by the organisers)
By October 25th, 2014, by applying through the UNECE Online Meeting Registration System: https://www2.unece.org/uncdb/app/ext/meeting-registration?id=mMKQgE .
In order to maintain the workshop character of the event, the number of places is limited to 65 persons in total – we will apply the ‘first-come, first-served’ principle.
Registration will be valid after confirmation.
All participants (including speakers and organisers) need to register themselves using the UNECE Registration System (link above) and obtain their registration form. Please do not forget to bring the original of your registration form with you! Entry for delegates without a valid ID badge: Pregny Gate of the Palais de Nations only (to obtain ID badge after having obtained registeration form online).
There is no fee attached to attending the event, but ENBES is not able to give any monetary support for the participation, travel to Geneva, or stay.
The workshop is in English only.
For the practical matter (registration, travel, accomodation) please see the document prepared by the host.
Organiser and host
Organiser: ENBES, through
Host: UNECE, through