eurostat methodological skills staff survey lesson learned final

26
Eurostat methodological skills staff survey: Lessons learned after the 1 st wave Q&A session 9 September 2015 room A2/45 11:00 https://ec.europa.eu/eusurvey/runner/MethNet2015 1

Upload: dario-buono

Post on 15-Apr-2017

41 views

Category:

Data & Analytics


1 download

TRANSCRIPT

Page 1: Eurostat methodological skills staff survey lesson learned final

Eurostat methodological skills staff survey:

Lessons learned after the 1st wave

Q&A session 9 September 2015 room A2/45 11:00https://ec.europa.eu/eusurvey/runner/MethNet2015

1

Page 2: Eurostat methodological skills staff survey lesson learned final

2

Eurostat methodological skills staff survey• Why?

To map existing methodological skills within Eurostat

• What for? Understanding better our competencies and how they are grouped

• Who? Open to all staff

• How? Linked with new flexible working conditions in Eurostat

Page 3: Eurostat methodological skills staff survey lesson learned final

3

Background

• Initiative proposed at the end of 2014 by the Eurostat 2020 discussion group B on “Permanent Methodological Development”, chaired by Director of A

• Adopted by the DM seminar in April 2015• Launched in June 2015 by Unit B1 in cooperation with

A2 under new sponsorship of Emanuele Baldacci, Director of B

• First wave during July 2015, open to all ESTAT staff• Still up and running

Page 4: Eurostat methodological skills staff survey lesson learned final

Aim of this presentation

• Snapshot after the "first wave", network visualisation tools (igraph for R)

• - Respondents• - Dimensions

• methodological areas, statistical domains, tools• training activities, active role

• - Lesson learned • - Next steps

4

Page 5: Eurostat methodological skills staff survey lesson learned final

5

When

Q&A session

2015

.7.1

2015

.7.2

2015

.7.3

2015

.7.6

2015

.7.7

2015

.7.8

2015

.7.9

2015

.7.1

0

2015

.7.1

3

2015

.7.1

4

2015

.7.1

5

2015

.7.1

6

2015

.7.1

7

2015

.7.1

9

2015

.7.2

1

2015

.7.2

3

2015

.7.2

8

2015

.7.3

1

0123456789

1010 10

1 12

1

7

3

7

1

3

1 12

1 12

1

Number of respondent per day

Launching: Cybernews email to DM & MethNet

Cybernews on Q&A

Page 6: Eurostat methodological skills staff survey lesson learned final

Dat

a A

naly

sis

Tim

e S

erie

s

Eco

nom

etric

s

Big

Dat

a A

naly

tics,

Dat

aM

inin

g an

d D

ata

Sci

ence

Sur

vey

Met

hodo

logy

Des

crip

tive

Sta

tistic

s

Dat

a P

roce

ssin

g

Dat

a V

alid

atio

n

Oth

er

Infe

rent

ial S

tatis

tics

Mod

el-b

ased

Est

imat

ion

Sta

tistic

al S

oftw

are

Sea

sona

l adj

ustm

ent

Sam

plin

g

Indi

cato

rs, c

ompo

site

and

synt

hetic

Adm

inis

trativ

e D

ata

Qua

lity

Dat

a D

isse

min

atio

n

Des

ign

and

optim

isat

ion

of s

tatis

tical

pro

cess

esIn

form

atio

n m

odel

san

d st

anda

rds

Met

adat

a m

odel

san

d st

anda

rds

Indi

ces

Dat

a In

tegr

atio

n

Line

ar A

lgeb

ra

Dat

a W

areh

ousi

ng

Ent

erpr

ise

Arc

hite

ctur

e

Sta

tistic

al c

onfid

entia

lity

Mic

ro-d

ata

acce

ss

Expertise Ranking of Methodological Areas

0

20

40

60

80

10091

60

43

30 29 27 2623 23 22 21 19 18 17 16 16

1310 10 10 10 9 7 6 5 3 1 0

Page 7: Eurostat methodological skills staff survey lesson learned final

7Agr

icul

tura

l and

Fish

erie

s S

tatis

tics

Bus

ines

s S

tatis

tics

Dem

ogra

phy

Env

ironm

enta

lS

tatis

tics

Inte

rnat

iona

l tra

de

Nat

iona

l Acc

ount

s

Pric

es

Sci

ence

and

tech

nolo

gy

Soc

ial S

tatis

tics

Tran

spor

t

Ene

rgy

Sta

tistic

Oth

er

Expertise breakdown per statistical domain

0

5

10

15

2

7

5

2

4

8

10

4

9

2

4 4

3

8

4

5

1

5 5

2

7

13

6

7

2

6

3

1

3

2 2 2

11

14

12 12

1 1

0 0 0 0

1 1

5

8

3 3

During studiesWorking in EurostatWorking elsewhereWorking on articles

Page 8: Eurostat methodological skills staff survey lesson learned final

R

SA

S

Evi

ews

Sta

ta

Mat

lab

Had

oop

Pyt

hon

SP

SS

Java

Oth

er

Expertise breakdown per statistical tool

0

5

10

15

20

10

13

8

17

10

14

12

4

8

0

4

1

8

0

5

2

7

0

3

4

1 1

3

2

1

2

4 4

16

1

10

5

6

4

5

6

7

4

7

3

During studiesWorking in EurostatWorking elsewherePersonal interest

Page 9: Eurostat methodological skills staff survey lesson learned final

9

Big

Dat

a A

naly

tics,

Dat

aM

inin

g an

d D

ata

Sci

ence

Dat

a A

naly

sis

Dat

a D

isse

min

atio

n

Dat

a In

tegr

atio

n

Dat

a P

roce

ssin

g

Dat

a V

alid

atio

n

Dat

a W

areh

ousi

ng

Des

crip

tive

Sta

tistic

sD

esig

n an

d op

timis

atio

nof

sta

tistic

al p

roce

sses

Eco

nom

etric

s

Ent

erpr

ise

Arc

hite

ctur

e

Indi

ces

Indi

cato

rs, c

ompo

site

and

synt

hetic

Infe

rent

ial S

tatis

tics

Info

rmat

ion

mod

els

and

stan

dard

sLi

near

Alg

ebra

Met

adat

a m

odel

san

d st

anda

rds

Mic

ro-d

ata

acce

ss

Mod

el-b

ased

Est

imat

ion

Qua

lity

Sea

sona

l adj

ustm

ent

Sta

tistic

al c

onfid

entia

lity

Sta

tistic

al S

oftw

are

Sur

vey

Met

hodo

logy

Sam

plin

g

Adm

inis

trativ

e D

ata

Tim

e S

erie

s

Oth

er

Competency in R and SAS by methodological areas

0

5

10

15

8

2

15

12

1

0

2

3

8

1

2

3

0

2 2

1

2

3

8

6

1 1

2 2

3

4

5

4

1

2

3

1 1

3

0 0

6

3

0

3

4 4

0 0

5

4 4

8

2

5

2

5

12

10

2

5

RSAS

Page 10: Eurostat methodological skills staff survey lesson learned final

10

Other

Energy Statistic

Transport

Social Statistics

Science and technology

Prices

National Accounts

International trade

Environmental Statistics

Demography

Business Statistics

Agricultural and Fisheries Statistics

Competency in R and SAS by statistical domains

number of people0 5 10 15

5 1

1 1

2 1

1314

3 0

5 4

7 5

1 3

3 5

3 4

5 8

3 4

RSAS

Page 11: Eurostat methodological skills staff survey lesson learned final

Other skillsDynamic Factor Analysis

Questionnaire Design

Monte Carlo, sensitivity analysis;

Fuzzy logic

Population size; analytical

hypotheses testing

Regional Statistics/

Geographic information

Land cover and

use

Classifications and metadata standards

Temporal disaggregation

Nowcasting

ArcGIS

Business cycle analysis, S-VARS, G-

VARS

Neural network

Remuneration and Pensions

Cluster analysis

Labour Market

Asylum migratio

n

Fortran, Cobol,Pascal, C

, C++,PHP, Octave

OX-metrics 11

Page 12: Eurostat methodological skills staff survey lesson learned final

12

Active Role

Informed58%

Active40%

No2%

Page 13: Eurostat methodological skills staff survey lesson learned final

Training delivering experience

No

Yes

Training delivering experience

0 5 10 15 20 25 30 35

24

31

As a trainer elsewhere

As a trainer for in-house courses

As a workshop facilitator

As a ESTP course trainer

As a ESTP course leader

At the university

Training delivering experience

0 5 10 15 20

16

15

10

12

4

20

13

Page 14: Eurostat methodological skills staff survey lesson learned final

14

Network Visualisation (Areas)

Page 15: Eurostat methodological skills staff survey lesson learned final

16

Network Visualisation (Tools)

Page 16: Eurostat methodological skills staff survey lesson learned final

17

Active role: full networkarea & tools

Page 17: Eurostat methodological skills staff survey lesson learned final

19

Classification of statistical skillsMICROECONOMETRICS MACROECONOMETRICS MULTIDIMENSIONAL

DATA ANALYSISPROCESSES CONCEPTUALISATION

Model based estimation Time series Inferential statistic Statistical confidentiality

Design and optimisation of statistical processes

Econometrics Indicators, composite and synthetic indices

Descriptive statistics Micro-data access

Metadata models and standards

Linear algebra Indices Data analysis Data validation Enterprise architecture

Sampling Seasonal adjustment Big data analytics, Data mining and data science

Data warehousing

Data integration

Survey methodology Administrative data Data processing Information models and standards

Statistical softwareQuality

Data dissemination

Page 18: Eurostat methodological skills staff survey lesson learned final

20

Network of classes

Page 19: Eurostat methodological skills staff survey lesson learned final

21

Class 1MICROECONOMETRICS

Page 20: Eurostat methodological skills staff survey lesson learned final

22

Class 2MACROECONOMETRICS

Page 21: Eurostat methodological skills staff survey lesson learned final

23

Class 3MULTIDIMENSIONAL DATA ANALYSIS

Page 22: Eurostat methodological skills staff survey lesson learned final

24

Class 4PRACTICAL PROCESSES

Page 23: Eurostat methodological skills staff survey lesson learned final

25

Class 5 CONCEPTUALISATION

Page 24: Eurostat methodological skills staff survey lesson learned final

31

Lesson learned

• About half of participants ready to take up an active role

• Strong expertise in Data analysis, Time Series and Econometrics• Most common tools SPSS, SAS and R• More than half have training delivery experience

• Identification of Centre of Competence might be based on a classification

Page 25: Eurostat methodological skills staff survey lesson learned final

32

Next steps

September 2015• Survey link will be kept open, please participate if

you have not done so yet• Eurostat Methodological Network (Yammer Group)

October 2015• New flexible way of working, network layer• DM paper with structured proposal

Page 26: Eurostat methodological skills staff survey lesson learned final

33

…to kick-off the discussion

• Any views and/or interpretation of what you have seen today?

• Do you see gaps in competencies emerging from the survey?