33
1 Prof. Dr. Petra Ahrweiler Innovation Lab Johannes Gutenberg University Mainz (JGU), Germany EA European Academy IMPROVING THE INNOVATION PROCESS WITH THE FIRM NEW INSIGHTS AND METHODS Masterclass

IMPROVING THE INNOVATION PROCESS WITH THE … · IMPROVING THE INNOVATION PROCESS WITH THE FIRM ... • Go to„File“ and„Models Library ... • SNA only captured the structural

  • Upload
    lamcong

  • View
    212

  • Download
    0

Embed Size (px)

Citation preview

1

Prof. Dr. PetraAhrweilerInnovationLab

JohannesGutenbergUniversityMainz(JGU),GermanyEAEuropeanAcademy

IMPROVINGTHEINNOVATIONPROCESSWITHTHEFIRM

NEWINSIGHTSANDMETHODSMasterclass

2

Contents

• Introduction

• Interactivesession 1:ScienceCafe „Utilityof new methods“• Table1:IfIhada“computerdummy”ofmyorganisation inanartificialworld:whatwouldIletitdoandtryout?

• Table2:Whatarethe“unknownunknowns”intheinnovationnetworkofmyorganisation,andwhatwouldbethebenefittogetreliableinformationaboutthem?

• Table3:Whatwouldbetheexactadded-valueformyorganisation inusingthesenewmethodologies?

• Table4:Whatcanthenewmethodologiesoffertoaddressthetop-5questionsforfuturedevelopmentofmyorganisation?

• Interimsummary and lecturer input

• Interactivesession 2:Method Lab„Hands-on“• Social NetworkAnalysis

• ABMand Netlogo

• To take home

3

Leadquestion for the ScienceCafeWhatistheutilityofthenewmethodologiesforthefuturedevelopmentofmyorganisation?

• With the new simulation methodologies counterfactual analysis ispossible

• offering abenchmark including measurable indicators for impactassessment of managerial interventions

• offering atool for „changing history“(backward counterfactualanalysis)and for „looking into the future“(forward counterfactualanalysis)

• exploring what-if questions

4

Additionaltable questions10minpertable (40min)

• Table1:IfIhada“computerdummy”ofmyorganisation inanartificialworld:whatwouldIletitdoandtryout?

• Table2:Whatarethe“unknownunknowns”intheinnovationnetworkofmyorganisation,andwhatwouldbethebenefittogetreliableinformationaboutthem?

• Table3:Whatwouldbetheexactadded-valueformyorganisationinusingthesenewmethodologies?

• Table4:Whatcanthenewmethodologiesoffertoaddressthetop-5questionsforfuturedevelopmentofmyorganisation?

5

Worldcafe methodology

• We need 4volunteers as chairs (chairs are staying attheir table updating new people towhat wassaid sofar,keeping time,and introducing their table paper)

• Please distribute equally attables (chairs write the table question onthe paper)

• Firstround:Please scribble downyour ideas how to address the table question;thenstart discussing your ideas;everybody can make notes onthe paper,after10minchange table

• Allfollowing rounds:chairs updatenewcomers about existing notes and pastdiscussions,otherwise sameas first round

• Afterlastround:chairs introduce their table notes inplenum for general discussion

• Moderatorcollectsaspectsforansweringtheleadquestion:Whatistheutilityofthenewmethodologiesforthefuturedevelopmentofmyorganisation?

6

Counterfactual analysis has ambitiousmethodological requirements

• Anadequate empirical description of the situation as it is now as abenchmark -including measurable indicators you are interested in

• while repeating the history which produced the situation if you are especiallyinterested inchanging the history to see what happens

• and with anendless-trapscenario where nothing ever happens if you areespecially interested inevaluating potentialfutures experimenting with differentinterventions

• How complex,laborious,costly,time-consuming...is this?

• CanIdoit myself to get afeeling for it?

7

How to get:anadequate empirical description ofthe situation as it is now?

• Empirical research (quantitativeand qualitative)

• Thedescription mustbe detailed enough to give you some causal ideasof what produces/changes your output indicators• Define your output indicators (e.g.revenues from new products and services,

position inthe industry etc.)

• Collect empirical data concerning the micro/macro structures and dynamics ofyour innovation network producing the relevantoutput indicators

8

Social NetworkAnalysis(SNA)

SNAisusedtoshowcentralactors(hubs)andclusters,analyseandvisualisethepositionandroleofparticularactortypes(e.g.SMEs),lookatcohesionanddensityofnetworksetc.

SNAoffers anadequate empirical descriptionof the situation as it is now as abenchmark -including measurable indicators

Here:• Nodes:funded organisations• Links/edges:„having aproject with..“

9

SNA– some examples to see what we need for it

10

Exampleforegonetworkanalysis:IndustrynetworkofUCDin2010

Sub-nodes:• UCDschools (green)• Companies(blue)

Links:• Having acollaboration with...• Having aproject with...

11

WithcompanynamesTOP10

12

Datamatrix

19

A1 A2 A3 A4 A5

A1 - 0 0 1 0

A2 0 - 1 0 0

A3 1 1 - 1 1

A4 0 1 0 - 0

A5 0 0 0 0 -

Network graph example and its matrix• Nodes:People,organisations..• Links:the activity between them

(e.g.giving advice,doing aprojectetc.)

• How does Ego‘s network look likeand how did it develop?

• What is Ego‘s position inthenetwork and how does it developover time?(centre,periphery,broker)

• Many more options:e.g.directionof graph (who asks whom for ameeting?),or other linktypes (whoplays golfwith whom?)

One matrix for one linktype...

13

Thisis scalable for a..

• small group (class inschool etc.)

• organisation (firm,university,public agency,association etc.)

• cluster (SiliconValleyetc.)

• industry sector (see biotech example following)

• region

• country

• EU....

14

DataformatforSNA

Source: Pyka and Saviotti 2000

15

Mapping and SNA

WalterPowell,Stanford

US Biotech industry1985

16

Huge diversity of organisations in the core

Strategy in 1997: multiple cooperations

17

Networkmeasures

• densityofthenetwork• numberofcomponents• sizeofthelargestcomponent• averagedegree• averagedistance• Clustering coefficient• informationonsmall-worldproperties(shortaveragedistanceand

highdegreeoflocalclustering?)• informationonscale-freeproperties(powerlawyes/no,degree

distributionofagents?)

18

SNA– hands-on

19

We need

• Adatabase inmatrix format

• Aprogram for doing network analysis

• Please try outyourself to work with this!

20

Hands-onSNA

• Install Gephi (Windowsor MacVersion)

• DownloadGephi onto your computer (click ondmg or exedepending onWindows/Mac),draw Gephi logointoApplications,StartGephi from Applications/Programmes,Goto Extrasand change Languageto English

• Open/Re-startGephi

• Clickon„opengraph file“

• OpenGREAT_V_03.gephi

• When you see the network,you can play around and,for example:

• Have alook atthe underlying database (DataLaboratory,button atthe top)

• Changethe layout of the network (left column,lower half)

• Changethe colouring (left column,upper half)

• Calculate network measures and statistics (right column)

21

Counterfactual analysis has ambitious methodological requirements

• With the new simulation methodologies counterfactual analysis ispossible• offering abenchmark including measurable indicators for impact

assessment of managerial interventions• offering atool for „changing history“(backward counterfactual

analysis)and for „looking into the future“ (forward counterfactualanalysis)

• exploring what-if questions• How complex,laborious,costly,time-consuming...is this?• CanIdoit myself to get afeeling for it?

22

ABM– understanding the basics

2323

Agent-basedModels(ABM)

• Agent-basedModellingisacomputationalmethodologythatallowstheanalysttocreate,analyse,andexperimentwithartificialworldspopulatedbyagents(computerprograms)thatinteractinnon-trivialways.

• Agentsareunitsthathavebehaviour• Theyactwithina(simulated)environment• Agentscan

• reacttootheragents,• pursuegoals,• communicatewithotheragents,• rememberwhattheyhavedone

• Macro-levelfeaturescanemergefromtheinteractionofagents

Schellingmodelofresidentialsegregation

24

Please download Netlogo!https://ccl.northwestern.edu/netlogo/

25

ABM– hands-on

26

Hands-onABM

• Install Netlogo (Windowsor MacVersion)

• OpenNetlogo

• Goto „File“and „ModelsLibrary“

• Goto „Social Science“and „Party“

• You see the Netlogo Interface:Goto „Info“(button atthe top)and read what it is

• Goto „Code“:there you see the program (your willsee that you can even read it)

• Gobackto the Interface

• You can play around and,for example,dowhat-if counterfactual forward analysis:• Changethe number of party guests

• Changethe number of tables to standat

• Changethe tolerance level of people

• Donotforget to press„set-up“and „go“afteryou have changed things

27

Atool to change history and to look into the future exploring what-if questions

• Define your evaluativequestions (e.g.what if mycompany increases itsopeninnovationactivities:willthisimprove my innovationperformance inthe longrun?)and yourhypotheses („theory ofchange“)

27

http://cress.soc.surrey.ac.uk/skinwp/

OpenSource:CCLicence for BasicSKINmodeland allSKINderivates

Freefor any public research;For commercial use (tenders,contract research,industry projects etc.),use mustbe granted and feesapply

2828

INFSO-SKIN flow diagram

Calibrated by empirical data of 1183EU-funded research projects,3783funded organisations(universities and research institutions (RESagents),research departments of big firms(largediversified firms LDFagents)and SMEs(SMEagents),and 11244project participationsbetween 2007and 2013

29

ScreenshotInterfaceINFSO-SKIN

30

To take home

31

To take home from SNA

• Datarequirements for getting anadequate description of the situation as it is can be high...

• Sometimes SNAis good,sometimes not:it depends onyour questions and onyour outputindicators

• SNAonlycapturedthestructuralaspectsoftheinnovationlandscape.

• Actors,processesandcausesproducingthesestructuresareinvisiblebetweenfrozensnapshotsoftwonetworkstatesfollowinguponeachother.

• Managersandstrategistsareconvincedthatitwouldbeusefultoknowabouttheseproceduralaspectstofindappropriateoptionsandaccesspointsforinterventionsandchanges.

• Furthermore,SNAjustallowstoevaluatethestructuresproducedbycertainstrategies“ex post”;ex-anteevaluationisonlypossibleintheverylimitedwaysofstatisticalmodelling

• Thefutureisnotaddressedproperly

• However,youcanuseyourstatusquoanalysisasabenchmark…

32

To take home from ABM

• ABMcanshedlightintothedarknessofthefuture(copingwiththechallengesofcomplexity,understandingthedynamicsofinnovation),identifyingpotentialaccesspointsforsuccessfulintervention• Simulationresults can inform about likely future effects of managerial interventions• Some of these effects are suprising and counter-intuitive.Newknowledge is generated.Complex

contexts are made available and accessible viaexperimentation.Simulations can help and practicehow to dealwith them.

• Askingwhat-ifquestions(ex-anteevaluation)- anoptionthatisnormallynoteasilyavailableinthemanagementworld• Decisionmakerscanusescenariomodellingasaworksitefortheir‘realityconstructions’• Experiments canbeusedtogiveanindicationofthelikelyeffectofawidevarietyofmanagementmeasures• Empirical‘Un-observables’,suchasknowledgeflowsininnovationorlearningofagents,canbemeasured

33

Thank you for your participation!