Transcript P4.1
Working Group on Tendency
Surveys
Overview of Programme
United Nations Statistics Division
Starting point
During the breakout session of the Scheveningen
seminar:
methodology for sentiment indicator and the need
(and options) for harmonization have been discussed
What is our focus? (avoid overlap with WG2)
The sentiment indicators are the (100%) surveybased indicators
Include both the conversion of qualitative multiple
choice questions into a single time series and the
computation of ad hoc indicators (involves choice of
the questions entering in the indicator, choice of the
methodology)
Follow-up steps taken:
What is the current situation?
What methodology exists?
(Guidelines)
What are options for harmonization?
Current situation
Worldwide, many organizations are
involved in compiling survey based
sentiment indicators
Most use their own rules and methods
In many cases produced outside the
national statistical offices.
Current situation
Result:
many practices (from the data collection to the
compilation of sentiment indicators)
many of them non-transparent and therefore
not replicable
data availability from private organizations is an
issue
some of them produce important and
recognized indicators
collaboration with private companies is essential
How do we get a better picture of
this situation?
Conduct a world-wide assessment of
country practices in compilation of
sentiment indicators
Difficulty:
Who are potential respondents?
Better picture for EU/OECD countries, but not
much known outside this group
Use NSOs as starting point
Assessment covers key question on
tendency surveys and indicator compilation
and dissemination
(results discussed separately)
Existing methodology
EC has harmonized methods for sentiment indicators
(questionnaires, coverage, timetable)
Results produced at Members States level and aggregate level on a
regular basis
still: data published at national level may differ from those published
by EC
OECD also provides guidelines
comparable to EC, but more general nature with room for country
specific requirements
While these data collection standards exist for EU/OECD countries,
existing methodology may not fit the needs of the rest of the world
Example: inclusion of the agricultural sector in the surveys
What modifications / additions to the existing guidelines are necessary
to make them more suitable for developing countries?
Question:
Harmonization
We are not looking for full harmonization at this time
Instead: sharing of best practices to extend the
existing knowledge past EU/OECD countries
consider a minimum set of international
requirements, in combination with country specific
needs
Where is harmonization necessary for global
comparison?
data collection (questions in surveys, number of
respondents, sampling methodology, classifications
used, etc.)
compilation of indicators (average of specific
questions, which kind of average, principal
components, factor analysis, etc.)
how to aggregate indicators (country weights)?
Where are we now?
World-wide assessment has been
completed
Review of existing guidelines
considering additional countries’
needs
Establishing of knowledge base