Sin ttulo de diapositiva

1
Using a TQM approach to get high quality incoming
data in the Spanish industrial surveys Ignaci
o Arbues, Manuel González, Margarita González,
José Quesada and Pedro Revilla (previlla_at_ine.es) N
ational Statistical Institute, Spain
2

• Summary
• ? Strategies to get high quality incoming data
• Web survey opportunities
• Key points
• Joint venture
• ? Auditing by reporting enterprises

3
TQM approach to produce Spanish industrial
statistics The essential ideas The suppliers
(reporting units) and the customers (data users)
are part of the productive system The
suppliers and the customers determine the
definition of quality to be used by our
organisation The basic principles Customer
satisfaction Continuous improvement
Facts-based management People-based management
4
TQM approach the suppliers and the customers are
part of the productive system ? integrating
reporting enterprises in our productive
system How?
5
Integrating reporting enterprises in our
productive system key success factors ? Data
collection adapted to respondent conditions and
possibilities ? Improving relationship with
enterprises
6
Data collection adapted to respondent conditions
and possibilities Requested variables and
valuation rules adapted to those of the
reporting enterprises Observation units adapted
to those units for which enterprises have
available information Using different models of
questionnaires adapted to the branch and size of
the enterprises Personalised questionnaires
Transmission formats (paper, fax, electronic
questionnaires) adapted to the choice of the
enterprises
7
microdata
Reporting enterprises
Statistical agency
macrodata
Key success factor Improving the twofold
relationship between the reporting enterprises
and the statistical agency
8
Reporting enterprises position about
questionnaires so, what use is this to me?
9
Hard contradiction ? Our customers are (more or
less) satisfied ? Our suppliers are strongly
unsatisfied
solution?
Lets try to change our suppliers into customers!
10
Offering tailored data in exchange for the
questionnaires We provide the reporting
enterprises with answers to the following
questions ? What is my market share in my
business activity? ? How many enterprises
have bigger market share than mine? ? What
is the overall share of those enterprises with
bigger market share than mine?
11
(No Transcript)
12
The industrial surveys carried out by the INE
are the best example of a joint venture between
industrial enterprises and public statistical
offices. In other words, they are an excellent
illustration of the advantages that may be
achieved by means of an efficient collaboration
between public and private sector Mr. BALTASAR
ERRASTI President of the Industry Commission of
the Spanish Confederation of Enterprise
Organisations.
13
Now, sometimes, reporting enterprises say the
Statistic gives something in return!
14
A new model of relationship with
enterprises the joint venture model
15
Fellegis virtuous circle Quality
microdata Quality macrodata
Society
Statistical agency
(Enterprises)
(market shares)
16
Demings 14 Points for Management Out of the
crises 3. Cease dependence on inspection to
achieve quality. Eliminate the need for
inspection on a mass basis by building quality
into the product in the first place
17
Data editing good practice moving editing
closer to respondents Going a step further
integrating the respondents in editing
18
action to improve our relation with
enterprises ? enterprises call us when data do
not match expectations ? errors detected and
corrected questions and definitions improved
19
Feedback from enterprises ? Reporting
enterprises may have an active role on editing
and quality evaluation ? Prior knowledge about
their sectors and markets
20
Auditing by reporting enterprises ? 11000
enterprises market studies ? 70000 enterprises
sample ? ? 11000 external auditors ? only 1 in
6 have participated We have to make a lot of
efforts to inform enterprises management of our
data offer
21
EDR and Web surveys opportunities ? Changes in
technology have broadened the possibilities for
data collection ? Offering an electronic
reporting option
22
Data editing by reporting enterprises ? Paper
questionnaires ? few opportunities ? Electronic
questionnaires ? many opportunities (built-in
edits) EDR enables more editing to take place at
data collection by the reporting enterprises
23
Web advantages ? Improving accuracy ?
Improving timeliness ? Reducing cost ? Reducing
enterprise burden (?)
24
A lot of expectation about the Web surveys in the
years to come the use of Web surveys has often
been lower than expected Why the rate of using
Internet surveys is quite low, while technical
conditions are available for many of the
respondents ?
25
Advantages for reporting enterprises? ? The
tasks that take the most time are looking for the
required information and computing answers ? No
time difference keying data on a screen to fill
in a questionnaire on paper ? The reporting
enterprises benefits depend on the way metadata
support filling in questionnaires (help texts,
auto-fill rules, pre-filled data, etc.)
26
Encouraging the use of Web questionnaires by
enterprises ? Explaining respondent benefits ?
Consider Web questionnaires in a wider context of
all administrative duties and all EDR
(e-commerce, e-administration, etc.) ?
Incentives (temporary access to information, free
deliveries of tailored data, etc)
27
Offering tailored data from the Web ? When an
enterprise sends a valid form, it immediately
receives tailored data from the server ? The
Web has some advantages over paper
28
(No Transcript)
29

Turnover. Annual rate
20,0
15,0
10,0


5,0
0,0
-5,0
-10,0
-15,0
-20,0
-25,0
-30,0
Jul-05
Jan-05
Feb-05
Mar-05
Apr-06
May-06
Jun-06
Total
Sector
Your establishment
30
Editing strategy ? After Web edits no traditional
microediting is needed ? Selective editing
based on time series modelling ? all fatal
errors and the most important query errors can be
corrected before the index is disseminated for
the first time
31
Final remarks ? Still far from the joint
venture and auditing by reporting enterprises
models ? Take-up of EDR too small ? More
research needed on selective editing
Nevertheless, the combination of TQM, built-in
edits in EDR and selective editing, may be in the
near future a key success factor to get high
quality incoming data, at a lower cost and with a
lower respondent burden