Indicatori complessi

(1)

Filomena Maggino

filomena.maggino@unifi.it

Towards more participative

methods in the construction of

composite indicators

.

The case of

(2)

1.Introduction

3.Obtaining weights:

objective ad subjective approaches

2.Obtaining weights:

general principles

4.Conclusions

(3)

1.Introduction

3.Obtaining weights:

objective ad subjective approaches

2.Obtaining weights:

general principles

4.Conclusions

(4)

How to obtain this ?

By involving

individuals

in the process of

social indicators construction

GOAL

(5)

Methodology aimed

Methodology aimed at

at

at constructing

at

constructing

indicators

⇓

“

technology

technology”

technology

⇓

need

need to have specialistic

to have specialistic

to have specialistic training in

training in

training in order to

order to

apply

apply the procedure in a

the procedure in a

scientific

and

objective

(6)

Actually

the procedure,

even though scientifically

defined

defined, , , ,

is

far

_far

from being

_{from being}

objective

(7)

Indicators construction

is developed

is developed through

through

through different

different

stages

stages. . . .

⇓

Each

Each stage

stage

stage requires

requires

a

decision

decision / / / / choice

decision

choice

((((methodological

methodological or

methodological

or

or not

not

not) ) ) )

to

to be taken

be taken

be taken....

⇓

Some decisions are quite technical

others may involve value judgments

(8)

It is quite difficult to make these decisions

objective

Generally

they are taken through a process accepted and

shared by the

scientific community

But

can we share methodological decisions /choices

by

(9)

1.

1. 1. choosing

choosing

choosing analytical approach

analytical approach

2.

2. 2. choosing

choosing

choosing and

and

and obtaining weights

and

obtaining weights

3.

3. 3. choosing

choosing

and

and identifying the

identifying the

aggregating

aggregating technique

technique

4.

4. 4. choosing

choosing

choosing models

models

models and

and

and conceptual

conceptual

approaches

approaches in

in

in order to assess

order to assess

(i)

(i) robustness

robustness

((((iiiiiiii) ) ) ) discriminant capacity

discriminant capacity

In

(10)

Weights

in indicators construction aim at

assigning differential

importance

to the indicators

to be aggregated.

⇓

this process does not represent simply a

technical problem

⇓

weights judgment values

(11)

Weights

⇓

differential subjective

weights

How to obtain subjective weights

How to obtain subjective weights????

(12)

Determining

and

_and

applying differential

_{applying differential}

subjective

weights

⇓

solid

conceptual

framework

helping in clarifying how

_to

_obtain

_weights

_to

_assign

_weights

(13)

1.Introduction

3.Obtaining weights:

objective ad subjective approaches

2.Obtaining weights:

general principles

4.Conclusions

(14)

(15)

General basic conditions concerning weights

The identified weights

_are

_{non negative}

_numbers

_{add up to unity}

_are

_related

_in

_some

_way

_to

_the

corresponding

score

_{may require to be rescaled in order to have}

an

identical range

(0; 1)

(16)

Reproducing each sub-score’s contribution to CI

a criterion has to be adopted

weighting system

improvement and refinement of the adopted

model of measurement

model of measurement.

(17)

Identification of a generic weighting system needs to take into accounttake into accounttake into accounttake into account :

_{rationale and theoretical framework}

_{on which the}

measurement of the complex characteristics is founded and that will consequently regard the synthetic score

_{meaning and contribution of each sub}

_{meaning and contribution of each sub----score}

_{meaning and contribution of each sub}

_score

_{to the}

synthesis

_{quality of data and statistical adequacy of}

indicators

(18)

proportional size of weightsproportional size of weightsproportional size of weightsproportional size of weights

a.

equal

equal or

equal

differential

differential weighting

differential

aggregation technique to be adopted aggregation technique to be adopted aggregation technique to be adopted aggregation technique to be adopted

b.

compensatory

compensatory or

compensatory

non

----

compensatory

Identification of a generic weighting system needs decisions to be taken

(19)

N.B.

AAAA

whole set of weights

able to express in

a

perfect

perfect way

_perfect

way

the contribution of each indicator

General statements

does not exist

(20)

Equal vs. differential weighting

(21)

First decision

Equal Weighting

_{Different Weighting}

The choice will strongly influence

the final results

(22)

• different components different components different components different components have tohave tohave tohave to bebebebe aggregatedaggregatedaggregatedaggregated by by by by different numbers of indicators

different numbers of indicators different numbers of indicators

different numbers of indicators ( synthetic score = unbalanced structure);

• indicators exist measuring the same componentindicators exist measuring the same componentindicators exist measuring the same componentindicators exist measuring the same component (

double weighted

o

double counting

).

Doubtful

procedure mainly

when

Equal weighting

(23)

Differential weighting

•

• theoretical reflectionstheoretical reflectionstheoretical reflectionstheoretical reflectionstheoretical reflectionstheoretical reflectionstheoretical reflectionstheoretical reflections •

• methodological concernsmethodological concernsmethodological concernsmethodological concernsmethodological concernsmethodological concernsmethodological concernsmethodological concerns

Doubtful

procedure mainly

when

not supported bynot supported bynot supported bynot supported by

(24)

Compensatory and

Non

Non----compensatory

compensatory

aggregation techniques

(25)

In order to avoid incoherencies between

theoretical meaning actual application

of weights of weights

a consistent aggregating technique is needed

by considering

compensability

among

the elementary indicators

(26)

Compensatory

aggregating approach

- additive approachadditive approachadditive approachadditive approach (simple addition)

- geometrical approachgeometrical approachgeometrical approachgeometrical approach (multiplicative technique)

low values compensated by high values

synthetic score does not allow us to return to the original individual profiles

problems of interpretation

problems of interpretation 



(27)

In obtaining

ssss

ubjective

weights

the decisions / choices

do not rely on experts.

Statements in obtaining

subjective weights

(28)

_{theoretical issue}_{theoretical issue}_{theoretical issue}_{theoretical issue}

“importance” is a distinct construct?

_{psychometric properties of importance ratings}_{psychometric properties of importance ratings}_{psychometric properties of importance ratings}_{psychometric properties of importance ratings}

internal consistency and test-retest reliability Identification of a subjective weighting system

needs to take into accounttake into accounttake into accounttake into account :

Statements in obtaining

subjective weights

(29)

the criterion the level

the techniques the approach

Identification of a subjective weighting system: a modelmodelmodelmodel should be chosen by considering

Statements in obtaining

subjective weights

(30)

1.Introduction

3.Obtaining weights:

objective ad subjective approaches

2.Obtaining weights:

general principles

4.Conclusions

(31)

Approaches for obtaining weights

can produce:

a.

statistical statistical statistical statistical methodsmethodsmethodsmethods

b.

multimultimultimulti----attribute attribute attribute approachesattribute approachesapproachesapproaches

c.

scaling approachesscaling approachesscaling approachesscaling approaches

Approaches for obtaining

weights

““““objectiveobjectiveobjectiveobjective weightsweightsweights””””weights

⇒

(32)

a.

a. Statistical methods

Statistical methods

Approaches for obtaining

weights

(33)

(34)

b.

b. Multi

Multi

Multi----attributes approaches

Multi

attributes approaches

Approaches for obtaining

weights

(35)

Multi

Multi----Attribute Models

Attribute Models

allow

subjective importance weights to be identified at

subjective level

through an indirect approach

by

Multi-attributes

approaches

• managing a certain number of combined comparisons

• applying methods aimed at making decision among different available alternatives

(36)

1. 1. 1.

1. MultiMultiMulti----Attribute Decision MakingMulti Attribute Decision MakingAttribute Decision MakingAttribute Decision Making:

_{Analytic Hierarchy Process}_{Analytic Hierarchy Process}_{Analytic Hierarchy Process}_{Analytic Hierarchy Process (AHP) (pairwise}

comparison)

Among these models we can distinguish:

2. 2. 2.

2. MultiMultiMulti----Attribute Compositional ModelMulti Attribute Compositional ModelAttribute Compositional ModelAttribute Compositional Modelssss:

_{Conjoint Analysis}_{Conjoint Analysis (CA)}_{Conjoint Analysis}_{Conjoint Analysis}

Multi-attributes

approaches

(37)

c.

c. Scaling approaches

Scaling approaches

Approaches for obtaining

weights

(38)

DimensionalityDimensionalityDimensionalityDimensionality Nature of dataNature of dataNature of dataNature of data

Scaling techniqueScaling techniqueScaling techniqueScaling technique

Criterion for testing the modelCriterion for testing the modelCriterion for testing the modelCriterion for testing the model Standard of measurementStandard of measurementStandard of measurementStandard of measurement

Contribution to the measurement of each multiple Contribution to the measurement of each multiple Contribution to the measurement of each multiple Contribution to the measurement of each multiple measures measures measures measures

They

They can

can

can be

be

be classifi

classifi

classified through

ed through

ed through their features

ed through

their features

their features::::

(39)

(40)

wh

whoooose nature is comparative or preferential

se nature is comparative or preferential

((((in in in in yellow in the previous table)yellow in the previous table)yellow in the previous table)yellow in the previous table)

produced by a comparative scaling technique

(in pink in the previous table) (in pink in the previous table) (in pink in the previous table) (in pink in the previous table)

Among all these approaches we

need

to

select those approaches that utilize data:

(41)

-- group weightinggroup weightinggroup weightinggroup weighting: group weightinggroup weightinggroup weightinggroup weighting

-

Thurstone model

(

differential scale

) -

Unfolding model

(

perceptual mapping

)

-- individual weightingindividual weightingindividual weightingindividual weighting: individual weightingindividual weightingindividual weightingindividual weighting

-

Conjoint model

Scaling models allowing subjective

weights to be obtained

weights to be obtained are

are

(42)

1.Introduction

3.Obtaining weights:

objective ad subjective approaches

2.Obtaining weights:

general principles

4.Conclusions

(43)

This work aims at

systematically framing the issue and

showing the possible approaches in order to

obtaining weights in a subjective perspective

⇓

many technical issues should be clarified.

(44)

Difficulties

Obtaining subjective weights requires and relies

on the accomplishment of large survey projects

aimed at collecting “importance” data.

- Time

- Resources

- Sampling

- Field work

- …

Conclusions

(45)

«Constructing composite indicators should take

into account the agreement among citizens

concerning the importance to be assigned to each

indicator»

(Hagerty and Land, 2007)

Seen in this perspective, this topic can be placed

in the ambit of an improvement of democratic

participation to decisions (“res publica”).

(46)

Presentation designer: trapani.marco@gmail.com

Thank you for your attention