SlideShare ist ein Scribd-Unternehmen logo
1 von 35
On extrapolation Federica Russo Philosophy, Kent
Overview Extrapolation: an old problem Extrapolation in social science: the background Cook & Campbell Guala and Steel Virtues and vices of mechanism-based approaches to extrapolation How to improve the mechanism-based approach Modelling vs. using mechanisms Role of populational properties 2
The evergreen riddle of induction Goodman’s new riddle Things are GRUE  If they are green before a certain time t If they are blue and not examined before time t. All Emeralds are green … but are they also grue? Distinguishing between projectible and non-projectible properties The new riddle is a form of extrapolation  3
The riddle of extrapolation Extrapolation, or external validity, is the inference by which the results of one study, e.g. an experiment, are extended to a larger or a different population or to a different setting Distinguishing between projectible and non-projectible results   animal models and human models; experimental economics and real world economic situations; demographic models across different countries; … 4
Extrapolation in social science:Cook & Campbell Validity: the best available approximation of the truth of causal statements Types of validity Internal: confidence in causal relation within population External: confidence in generalising to other populations 5
Exporting results:an issue of external validity Generalisingto and across populations External validity refers to the approximate validity with which we can infer that the presumed causal relationship can be generalized to and across alternate measures of the cause and the effect and across different types of persons, settings, and times. (Cook and Campbell, 1979, p. 37) Generalizing to well-explicated target populations should be clearly distinguished from generalizing across populations. Each is germane to external validity: the former is crucial for ascertaining whether any research goal that specified populations have been met, and the latter is crucial for ascertaining which different populations (or subpopulations) have been affected by a treatment, i.e., for assessing how far one can generalize. (Cook and Campbell, 1979, p. 71) 6
The assessment of external validity When? How? Results of test of statistical interactions Between the selection of individuals to take part in the study and the treatment / intervention Representativeness of sample Between history (= particular conditions of the study/experiment) and treatment Possibility to replicate studies 7
Extrapolation entersthe philosophical debate An emergent awareness For once, from the right angle: Analogical reasoning and mechanisms 8
Guala: analogical reasoning External validity rests an empirical problem Solution: wisely combine field and laboratory evidence in analogical reasoning 9
The structure of such an inference can be reconstructed as follows: 1. If all directly observable features of the target and the experi- mental system are similar in structure; 2. If all the indirectly observable features have been adequately controlled in the laboratory; 3.   If there is no reason to believe that they differ in the target system; 4.   And if the outcome of the two systems at work (the data) is similar; 5. Then, the experimental and target systems are likely to be structurally similar mechanisms (or data-generating processes).  (Guala, 2005, p. 180) 10
Steel: comparative process tracing The extrapolator’s circle Challenge of successfully exporting information from model population knowing that it is limited and partial The problem of difference Challenge of providing successful methods in presence of differences between the model and the target population Solution: Comparing the mechanisms in the model and in the target at the points in which they are more likely to differ 11
Thus, efficient applications of comparative process tracing can focus on likely sources of difference in downstream stages of the mechanism. (Steel, 2008, p. 90, emphasis in the original) 12
Mechanism-based extrapolation is a step beyond Cook & Campbell tradition What’s the scope of mechanism-based approaches? Is all extrapolation practice mechanism-based? Many extrapolation practices aren’t mechanism-based, although arguably they should 13
Virtues 14
Beyond ‘nichilist’ stanceà la LaFollette & Schanks Causal Analogue Models:  can grant extrapolation Hypothetical Analogue Models: can only suggest hypotheses to test Animal models can only be HAM Extrapolation possible only if there are no differences … but that’s exactly the problem! 15
Field knowledge is important Beyond statistically-minded tradition of Cook and Campbell Stress only representativeness of samples and possibility to replicate studies 16
Against detractors of extrapolation … Even by Cook and Campbell! The priority among validity types varies with the kind of research being conducted. For persons interested in theory testing it is almost as important to show that the variables involved in the research are constructs A and B (construct validity) as it is to show that the relationship is causal and goes from one variable to another (internal validity). Few theories specify crucial target settings, population, or times to or across which generalization is desired. Consequently, external validity is of relatively little importance.[…] For investigators with theoretical interests our estimate is that the types of validity, in order of importance, are probably internal, construct, statistical conclusion, and external validity. Cook and Campbell (1979, p. 83) 17
Luckily, contrary voices exist: A primary goal in all sciences, including the social sciences, is the production of general knowledge. General knowledge is knowledge that is not confined to the particulars of time and place. Lucas (2003, p. 236) 18
Scope of mechanism-based extrapolation 19
Guala: 	Sometimes external validity takes the form of “the in vitro–in vivo problem” (biochemistry), sometimes it is called “ecological validity” (psychology), and sometimes it is called “parallelism” (economics), but the issue is always the same. (Guala, 2005, p.160) 20
Steel: The best way to introduce to topic of this book is with a few examples. Studies find that a particular substance is carcinogenic in rats. We would like to know whether it is also such in humans. A randomized controlled experiment has found that a pilot welfare-to-work program improved the economic prospects of welfare recipients. It is desired to know whether the program will be similarly effective in other locations and when implemented on a larger scale. On the basis of a controlled experiment concerning outcomes resulting from initiating anti-retroviral therapies earlier or later among HIV+ patients, a physician wishes to decide the best time to initiate this therapy for the patients she treats. [. . . ] I will use the term extrapolation to refer to inferences of this sort. (Steel, 2008, p. 3) 21
Too much in the same basket? Are external validity issues / methods the same in domains as different as economics and biology? An old issue arises again Are the natural and social sciences completely apart? Or is it a problem of having more direct and independent access to social or biological mechanisms? 22
Varieties of external validity inferences
External validities Internal/external: within/outside the sample Population: different populations of subjects Ecological: same subjects, different settings Temporal: same subjects, same settings, different times Do different types of validity require different methods of extrapolation? What population are we talking about? Cook & Campbell:  maximise representativeness of sample Mechanism-based extrapolation marks step beyond, yet, not enough emphasis on the role of socio-demo-political characteristics of populations 24
Knowledge of the mechanism Steel / Guala Knowledge of the functioning mechanism in the model and in the target Ethnographers Knowledge of mechanism in the model ‘Translate’ the mechanism in the target ,[object Object],	mechanism-based extrapolation to be developed 25
Role of the mechanism IARC: use evidence from animal models This evidence should be explicitly mechanistic Verificationist strategies during the process of research Mechanistic considerations should come at the ‘theory development stage’ through ‘modelling mechanisms’ ,[object Object],26
Direction of the inference To a larger population e.g. RCTs To the experimental setting e.g. to test theoretical explanations / general theories ,[object Object],27
How to improvethe mechanism-based approach
Modelling and using mechanisms Modelling Find out what the mechanism is: what it is made of, how it functions Infer the mechanism from observations and experiments Using For explanation Mechanisms carry explanatory power because they display how the phenomenon was brought about For external validity Mechanisms used in various ways 29
What population are we talking about? Recall the many external validities on the market: Internal/external: within/outside the sample Population: different populations of subjects Ecological: same subjects, different settings Temporal: same subjects, same settings, different times But what counts as ‘same’ or ‘different’ population? 30
The overlooked role ofpopulational properties E.g. demographic, socio-political-economic characteristics for the choice of variables or of proxies for some properties or of the statistical model, for the interpretation of results … The possibility to extrapolate strongly depends on the properties of the populations The process of extrapolation itself requires comparing the properties of the populations This has been overlooked in both the statistically-minded and the mechanism-based approach 31
Thus, efficient applications of comparative process tracing can focus on likely sources of difference in downstream stages of the mechanism. A few important qualifications about the emphasis on downstream stages should be noted. The strategy could lead to mistaken conclusions if there is a path that bypasses the downstream stage. [. . . ] Second, the mark that upstream stages leave upon the downstream stages must be distinctive in the sense that it could not have resulted from some independent causes. (Steel, 2008, p. 90) 32
Downstream … where? Appealing to the properties of the population does not exactly coincide with “the sources of difference in downstream stages of the mechanism” identified by Steel. is finding downstream differences about the populations themselves. 33
To sum up and conclude The Cook & Campbell tradition Mechanism-based extrapolation Goesbeyond Cook & Campbell ,[object Object],Shouldbe adopted in extrapolation procedures that do not involve mechanistic considerations ,[object Object],populational properties 34
Selected bibliography Cook, T. and Campbell, D. (1979). Quasi-Experimentation. Design and Analysis Issues for Field Settings. Rand MacNally, Chicago. Godfrey-Smith, P. (2003). Goodman's problem and scientic methodology. The Journal of Philosophy, 100(11):573-590. Goodman, N. (1955). Facts, Fiction, and Forecast. Cambridge University Press, Harvard. Guala, F. (2005). The methodology of experimental economics. Cambridge University Press. LaFollette, H. and Schanks, N. (1995). Two models of models in biomedical research. Philosophical Quarterly, 45:141-160. Lucas, J. W. (2003). Theory-testing, generalization, and the problem of external validity. Sociological Theory, 21(3):236-253. Steel, D. (2008). Across the boundaries. Extrapolation in biology and social science. Oxford University Press. 35

Weitere ähnliche Inhalte

Was ist angesagt?

weon preconference 2013 vandenbroucke counterfactual theory causality epidem...
 weon preconference 2013 vandenbroucke counterfactual theory causality epidem... weon preconference 2013 vandenbroucke counterfactual theory causality epidem...
weon preconference 2013 vandenbroucke counterfactual theory causality epidem...Bsie
 
How to read a paper
How to read a paperHow to read a paper
How to read a paperfaheta
 
Research Methodology
Research MethodologyResearch Methodology
Research MethodologyAneel Raza
 
Research design: Design of Experiment
Research design: Design of ExperimentResearch design: Design of Experiment
Research design: Design of ExperimentNirmalaPrajapati4
 
Types of Hypothesis-Advance Research Methodology
Types of Hypothesis-Advance Research MethodologyTypes of Hypothesis-Advance Research Methodology
Types of Hypothesis-Advance Research MethodologyRehan Ehsan
 
Causal design & research
Causal design & researchCausal design & research
Causal design & researchgadiabinit
 
Gemechu keneni(PhD) document
Gemechu keneni(PhD) documentGemechu keneni(PhD) document
Gemechu keneni(PhD) documentgetahun bekana
 
Weon preconference pearce variation and causation
Weon preconference pearce variation and causationWeon preconference pearce variation and causation
Weon preconference pearce variation and causationBsie
 
Planning of experiment in industrial research
Planning of experiment in industrial researchPlanning of experiment in industrial research
Planning of experiment in industrial researchpbbharate
 
Research design and experimentation
Research design and experimentationResearch design and experimentation
Research design and experimentationDr NEETHU ASOKAN
 
Projecting ‘time to event’ outcomes in technology assessment: an alternative ...
Projecting ‘time to event’ outcomes in technology assessment: an alternative ...Projecting ‘time to event’ outcomes in technology assessment: an alternative ...
Projecting ‘time to event’ outcomes in technology assessment: an alternative ...cheweb1
 
Brm (one tailed and two tailed hypothesis)
Brm (one tailed and two tailed hypothesis)Brm (one tailed and two tailed hypothesis)
Brm (one tailed and two tailed hypothesis)Upama Dwivedi
 

Was ist angesagt? (20)

weon preconference 2013 vandenbroucke counterfactual theory causality epidem...
 weon preconference 2013 vandenbroucke counterfactual theory causality epidem... weon preconference 2013 vandenbroucke counterfactual theory causality epidem...
weon preconference 2013 vandenbroucke counterfactual theory causality epidem...
 
How to read a paper
How to read a paperHow to read a paper
How to read a paper
 
Research Methodology
Research MethodologyResearch Methodology
Research Methodology
 
Experimental
ExperimentalExperimental
Experimental
 
T‑tests
T‑testsT‑tests
T‑tests
 
Research design: Design of Experiment
Research design: Design of ExperimentResearch design: Design of Experiment
Research design: Design of Experiment
 
Types of Hypothesis-Advance Research Methodology
Types of Hypothesis-Advance Research MethodologyTypes of Hypothesis-Advance Research Methodology
Types of Hypothesis-Advance Research Methodology
 
Methods for Tina
Methods for TinaMethods for Tina
Methods for Tina
 
Causal Research
Causal ResearchCausal Research
Causal Research
 
Causal design & research
Causal design & researchCausal design & research
Causal design & research
 
Gemechu keneni(PhD) document
Gemechu keneni(PhD) documentGemechu keneni(PhD) document
Gemechu keneni(PhD) document
 
4 chao chien chen
4 chao chien chen4 chao chien chen
4 chao chien chen
 
Weon preconference pearce variation and causation
Weon preconference pearce variation and causationWeon preconference pearce variation and causation
Weon preconference pearce variation and causation
 
Planning of experiment in industrial research
Planning of experiment in industrial researchPlanning of experiment in industrial research
Planning of experiment in industrial research
 
Meta analisis & criminology
Meta analisis & criminologyMeta analisis & criminology
Meta analisis & criminology
 
Research design and experimentation
Research design and experimentationResearch design and experimentation
Research design and experimentation
 
Projecting ‘time to event’ outcomes in technology assessment: an alternative ...
Projecting ‘time to event’ outcomes in technology assessment: an alternative ...Projecting ‘time to event’ outcomes in technology assessment: an alternative ...
Projecting ‘time to event’ outcomes in technology assessment: an alternative ...
 
Seminar in Meta-analysis
Seminar in Meta-analysisSeminar in Meta-analysis
Seminar in Meta-analysis
 
Experimental Research
Experimental ResearchExperimental Research
Experimental Research
 
Brm (one tailed and two tailed hypothesis)
Brm (one tailed and two tailed hypothesis)Brm (one tailed and two tailed hypothesis)
Brm (one tailed and two tailed hypothesis)
 

Andere mochten auch (10)

Extrapolación Richardson
Extrapolación RichardsonExtrapolación Richardson
Extrapolación Richardson
 
Extrapolacion
ExtrapolacionExtrapolacion
Extrapolacion
 
Métodos en prospectiva 4- extrapolación
Métodos en prospectiva 4- extrapolaciónMétodos en prospectiva 4- extrapolación
Métodos en prospectiva 4- extrapolación
 
Extrapolation
ExtrapolationExtrapolation
Extrapolation
 
Extrapolation
ExtrapolationExtrapolation
Extrapolation
 
Extrapolation
ExtrapolationExtrapolation
Extrapolation
 
Derivacion e integracion numéricas
Derivacion e integracion numéricasDerivacion e integracion numéricas
Derivacion e integracion numéricas
 
Interpolation and extrapolation
Interpolation and extrapolationInterpolation and extrapolation
Interpolation and extrapolation
 
Interpolation
InterpolationInterpolation
Interpolation
 
interpolation
interpolationinterpolation
interpolation
 

Ähnlich wie Extrapolation Kent Feb10

An Experimental Template For Case Study Research
An Experimental Template For Case Study ResearchAn Experimental Template For Case Study Research
An Experimental Template For Case Study ResearchZaara Jensen
 
The Research Process
The Research ProcessThe Research Process
The Research ProcessK. Challinor
 
13 Apr 2002 1459 AR AR158-02.tex AR158-02.SGM LaTeX2e(200105.docx
13 Apr 2002 1459 AR AR158-02.tex AR158-02.SGM LaTeX2e(200105.docx13 Apr 2002 1459 AR AR158-02.tex AR158-02.SGM LaTeX2e(200105.docx
13 Apr 2002 1459 AR AR158-02.tex AR158-02.SGM LaTeX2e(200105.docxmoggdede
 
Workshop ppt Methodology for Allied.pptx
Workshop ppt Methodology for Allied.pptxWorkshop ppt Methodology for Allied.pptx
Workshop ppt Methodology for Allied.pptxAbidAli749894
 
Malec, T. & Newman, M. (2013). Research methods Building a kn.docx
Malec, T. & Newman, M. (2013). Research methods Building a kn.docxMalec, T. & Newman, M. (2013). Research methods Building a kn.docx
Malec, T. & Newman, M. (2013). Research methods Building a kn.docxcroysierkathey
 
The Relationship Between Body Image And The Media
The Relationship Between Body Image And The MediaThe Relationship Between Body Image And The Media
The Relationship Between Body Image And The MediaJessica Myers
 
Descriptive versus Mechanistic Modeling
Descriptive versus Mechanistic ModelingDescriptive versus Mechanistic Modeling
Descriptive versus Mechanistic ModelingAshwani Dhingra
 
Statistical modeling in pharmaceutical research and development
Statistical modeling in pharmaceutical research and developmentStatistical modeling in pharmaceutical research and development
Statistical modeling in pharmaceutical research and developmentPV. Viji
 
What's the Science in Data Science? - Skipper Seabold
What's the Science in Data Science? - Skipper SeaboldWhat's the Science in Data Science? - Skipper Seabold
What's the Science in Data Science? - Skipper SeaboldPyData
 
Relevance of experimental design
 Relevance of experimental design Relevance of experimental design
Relevance of experimental designAlexander Decker
 
Relevance of experimental design
 Relevance of experimental design Relevance of experimental design
Relevance of experimental designAlexander Decker
 
Relevance of experimental design
 Relevance of experimental design Relevance of experimental design
Relevance of experimental designAlexander Decker
 
From simulated model by bio pepa to narrative language through sbml
From simulated model by bio pepa to narrative language through sbmlFrom simulated model by bio pepa to narrative language through sbml
From simulated model by bio pepa to narrative language through sbmlijctcm
 
Types of research design experiments
Types of research design   experimentsTypes of research design   experiments
Types of research design experimentsrozy_kalsi
 

Ähnlich wie Extrapolation Kent Feb10 (20)

An Experimental Template For Case Study Research
An Experimental Template For Case Study ResearchAn Experimental Template For Case Study Research
An Experimental Template For Case Study Research
 
The Research Process
The Research ProcessThe Research Process
The Research Process
 
13 Apr 2002 1459 AR AR158-02.tex AR158-02.SGM LaTeX2e(200105.docx
13 Apr 2002 1459 AR AR158-02.tex AR158-02.SGM LaTeX2e(200105.docx13 Apr 2002 1459 AR AR158-02.tex AR158-02.SGM LaTeX2e(200105.docx
13 Apr 2002 1459 AR AR158-02.tex AR158-02.SGM LaTeX2e(200105.docx
 
Causal modelling - Series of lectures on causal modelling in the social sciences
Causal modelling - Series of lectures on causal modelling in the social sciencesCausal modelling - Series of lectures on causal modelling in the social sciences
Causal modelling - Series of lectures on causal modelling in the social sciences
 
Workshop ppt Methodology for Allied.pptx
Workshop ppt Methodology for Allied.pptxWorkshop ppt Methodology for Allied.pptx
Workshop ppt Methodology for Allied.pptx
 
Malec, T. & Newman, M. (2013). Research methods Building a kn.docx
Malec, T. & Newman, M. (2013). Research methods Building a kn.docxMalec, T. & Newman, M. (2013). Research methods Building a kn.docx
Malec, T. & Newman, M. (2013). Research methods Building a kn.docx
 
02 basic researchmethod
02 basic researchmethod02 basic researchmethod
02 basic researchmethod
 
The Relationship Between Body Image And The Media
The Relationship Between Body Image And The MediaThe Relationship Between Body Image And The Media
The Relationship Between Body Image And The Media
 
Descriptive versus Mechanistic Modeling
Descriptive versus Mechanistic ModelingDescriptive versus Mechanistic Modeling
Descriptive versus Mechanistic Modeling
 
PR 2, WEEK 2.pptx
PR 2, WEEK 2.pptxPR 2, WEEK 2.pptx
PR 2, WEEK 2.pptx
 
Statistical modeling in pharmaceutical research and development
Statistical modeling in pharmaceutical research and developmentStatistical modeling in pharmaceutical research and development
Statistical modeling in pharmaceutical research and development
 
What's the Science in Data Science? - Skipper Seabold
What's the Science in Data Science? - Skipper SeaboldWhat's the Science in Data Science? - Skipper Seabold
What's the Science in Data Science? - Skipper Seabold
 
Hypothesis
HypothesisHypothesis
Hypothesis
 
Relevance of experimental design
 Relevance of experimental design Relevance of experimental design
Relevance of experimental design
 
Relevance of experimental design
 Relevance of experimental design Relevance of experimental design
Relevance of experimental design
 
Relevance of experimental design
 Relevance of experimental design Relevance of experimental design
Relevance of experimental design
 
Correlational data, causal hypotheses and validity
Correlational data, causal hypotheses and validityCorrelational data, causal hypotheses and validity
Correlational data, causal hypotheses and validity
 
From simulated model by bio pepa to narrative language through sbml
From simulated model by bio pepa to narrative language through sbmlFrom simulated model by bio pepa to narrative language through sbml
From simulated model by bio pepa to narrative language through sbml
 
Types of research design experiments
Types of research design   experimentsTypes of research design   experiments
Types of research design experiments
 
Research methodology
Research methodologyResearch methodology
Research methodology
 

Mehr von University of Amsterdam and University College London

Mehr von University of Amsterdam and University College London (20)

H-AI-BRID - Thinking and designing Human-AI systems
H-AI-BRID - Thinking and designing Human-AI systemsH-AI-BRID - Thinking and designing Human-AI systems
H-AI-BRID - Thinking and designing Human-AI systems
 
Time in QCA: a philosopher’s perspective
Time in QCA: a philosopher’s perspectiveTime in QCA: a philosopher’s perspective
Time in QCA: a philosopher’s perspective
 
Interconnected health-environmental challenges: Between the implosion of the ...
Interconnected health-environmental challenges: Between the implosion of the ...Interconnected health-environmental challenges: Between the implosion of the ...
Interconnected health-environmental challenges: Between the implosion of the ...
 
Trusting AI-generated contents: a techno-scientific approach
Trusting AI-generated contents: a techno-scientific approachTrusting AI-generated contents: a techno-scientific approach
Trusting AI-generated contents: a techno-scientific approach
 
Interconnected health-environmental challenges, Health and the Environment: c...
Interconnected health-environmental challenges, Health and the Environment: c...Interconnected health-environmental challenges, Health and the Environment: c...
Interconnected health-environmental challenges, Health and the Environment: c...
 
Who Needs “Philosophy of Techno- Science”?
Who Needs “Philosophy of Techno- Science”?Who Needs “Philosophy of Techno- Science”?
Who Needs “Philosophy of Techno- Science”?
 
Philosophy of Techno-Science: Whence and Whither
Philosophy of Techno-Science: Whence and WhitherPhilosophy of Techno-Science: Whence and Whither
Philosophy of Techno-Science: Whence and Whither
 
Charting the explanatory potential of network models/network modeling in psyc...
Charting the explanatory potential of network models/network modeling in psyc...Charting the explanatory potential of network models/network modeling in psyc...
Charting the explanatory potential of network models/network modeling in psyc...
 
The implosion of medical evidence: emerging approaches for diverse practices ...
The implosion of medical evidence: emerging approaches for diverse practices ...The implosion of medical evidence: emerging approaches for diverse practices ...
The implosion of medical evidence: emerging approaches for diverse practices ...
 
On the epistemic and normative benefits of methodological pluralism
On the epistemic and normative benefits of methodological pluralismOn the epistemic and normative benefits of methodological pluralism
On the epistemic and normative benefits of methodological pluralism
 
Socio-markers and information transmission
Socio-markers and information transmissionSocio-markers and information transmission
Socio-markers and information transmission
 
Disease causation and public health interventions
Disease causation and public health interventionsDisease causation and public health interventions
Disease causation and public health interventions
 
The life-world of health and disease and the design of public health interven...
The life-world of health and disease and the design of public health interven...The life-world of health and disease and the design of public health interven...
The life-world of health and disease and the design of public health interven...
 
Towards and epistemological and ethical XAI
Towards and epistemological and ethical XAITowards and epistemological and ethical XAI
Towards and epistemological and ethical XAI
 
Value-promoting concepts in the health sciences and public health
Value-promoting concepts in the health sciences and public healthValue-promoting concepts in the health sciences and public health
Value-promoting concepts in the health sciences and public health
 
Connecting the epistemology and ethics of AI
Connecting the epistemology and ethics of AIConnecting the epistemology and ethics of AI
Connecting the epistemology and ethics of AI
 
How is Who. Empowering evidence for sustainability and public health interven...
How is Who. Empowering evidence for sustainability and public health interven...How is Who. Empowering evidence for sustainability and public health interven...
How is Who. Empowering evidence for sustainability and public health interven...
 
High technologized justice – The road map for policy & regulation. Legaltech ...
High technologized justice – The road map for policy & regulation. Legaltech ...High technologized justice – The road map for policy & regulation. Legaltech ...
High technologized justice – The road map for policy & regulation. Legaltech ...
 
Connecting the epistemology and ethics of AI
Connecting the epistemology and ethics of AIConnecting the epistemology and ethics of AI
Connecting the epistemology and ethics of AI
 
Science and values. A two-way relations
Science and values. A two-way relationsScience and values. A two-way relations
Science and values. A two-way relations
 

Kürzlich hochgeladen

Strategies for Landing an Oracle DBA Job as a Fresher
Strategies for Landing an Oracle DBA Job as a FresherStrategies for Landing an Oracle DBA Job as a Fresher
Strategies for Landing an Oracle DBA Job as a FresherRemote DBA Services
 
Apidays New York 2024 - The Good, the Bad and the Governed by David O'Neill, ...
Apidays New York 2024 - The Good, the Bad and the Governed by David O'Neill, ...Apidays New York 2024 - The Good, the Bad and the Governed by David O'Neill, ...
Apidays New York 2024 - The Good, the Bad and the Governed by David O'Neill, ...apidays
 
Exploring Multimodal Embeddings with Milvus
Exploring Multimodal Embeddings with MilvusExploring Multimodal Embeddings with Milvus
Exploring Multimodal Embeddings with MilvusZilliz
 
Apidays New York 2024 - Accelerating FinTech Innovation by Vasa Krishnan, Fin...
Apidays New York 2024 - Accelerating FinTech Innovation by Vasa Krishnan, Fin...Apidays New York 2024 - Accelerating FinTech Innovation by Vasa Krishnan, Fin...
Apidays New York 2024 - Accelerating FinTech Innovation by Vasa Krishnan, Fin...apidays
 
FWD Group - Insurer Innovation Award 2024
FWD Group - Insurer Innovation Award 2024FWD Group - Insurer Innovation Award 2024
FWD Group - Insurer Innovation Award 2024The Digital Insurer
 
ProductAnonymous-April2024-WinProductDiscovery-MelissaKlemke
ProductAnonymous-April2024-WinProductDiscovery-MelissaKlemkeProductAnonymous-April2024-WinProductDiscovery-MelissaKlemke
ProductAnonymous-April2024-WinProductDiscovery-MelissaKlemkeProduct Anonymous
 
Architecting Cloud Native Applications
Architecting Cloud Native ApplicationsArchitecting Cloud Native Applications
Architecting Cloud Native ApplicationsWSO2
 
MS Copilot expands with MS Graph connectors
MS Copilot expands with MS Graph connectorsMS Copilot expands with MS Graph connectors
MS Copilot expands with MS Graph connectorsNanddeep Nachan
 
Polkadot JAM Slides - Token2049 - By Dr. Gavin Wood
Polkadot JAM Slides - Token2049 - By Dr. Gavin WoodPolkadot JAM Slides - Token2049 - By Dr. Gavin Wood
Polkadot JAM Slides - Token2049 - By Dr. Gavin WoodJuan lago vázquez
 
Finding Java's Hidden Performance Traps @ DevoxxUK 2024
Finding Java's Hidden Performance Traps @ DevoxxUK 2024Finding Java's Hidden Performance Traps @ DevoxxUK 2024
Finding Java's Hidden Performance Traps @ DevoxxUK 2024Victor Rentea
 
Artificial Intelligence Chap.5 : Uncertainty
Artificial Intelligence Chap.5 : UncertaintyArtificial Intelligence Chap.5 : Uncertainty
Artificial Intelligence Chap.5 : UncertaintyKhushali Kathiriya
 
Cloud Frontiers: A Deep Dive into Serverless Spatial Data and FME
Cloud Frontiers:  A Deep Dive into Serverless Spatial Data and FMECloud Frontiers:  A Deep Dive into Serverless Spatial Data and FME
Cloud Frontiers: A Deep Dive into Serverless Spatial Data and FMESafe Software
 
Six Myths about Ontologies: The Basics of Formal Ontology
Six Myths about Ontologies: The Basics of Formal OntologySix Myths about Ontologies: The Basics of Formal Ontology
Six Myths about Ontologies: The Basics of Formal Ontologyjohnbeverley2021
 
Cloud Frontiers: A Deep Dive into Serverless Spatial Data and FME
Cloud Frontiers:  A Deep Dive into Serverless Spatial Data and FMECloud Frontiers:  A Deep Dive into Serverless Spatial Data and FME
Cloud Frontiers: A Deep Dive into Serverless Spatial Data and FMESafe Software
 
[BuildWithAI] Introduction to Gemini.pdf
[BuildWithAI] Introduction to Gemini.pdf[BuildWithAI] Introduction to Gemini.pdf
[BuildWithAI] Introduction to Gemini.pdfSandro Moreira
 
Platformless Horizons for Digital Adaptability
Platformless Horizons for Digital AdaptabilityPlatformless Horizons for Digital Adaptability
Platformless Horizons for Digital AdaptabilityWSO2
 
Connector Corner: Accelerate revenue generation using UiPath API-centric busi...
Connector Corner: Accelerate revenue generation using UiPath API-centric busi...Connector Corner: Accelerate revenue generation using UiPath API-centric busi...
Connector Corner: Accelerate revenue generation using UiPath API-centric busi...DianaGray10
 
Apidays New York 2024 - Passkeys: Developing APIs to enable passwordless auth...
Apidays New York 2024 - Passkeys: Developing APIs to enable passwordless auth...Apidays New York 2024 - Passkeys: Developing APIs to enable passwordless auth...
Apidays New York 2024 - Passkeys: Developing APIs to enable passwordless auth...apidays
 
Strategize a Smooth Tenant-to-tenant Migration and Copilot Takeoff
Strategize a Smooth Tenant-to-tenant Migration and Copilot TakeoffStrategize a Smooth Tenant-to-tenant Migration and Copilot Takeoff
Strategize a Smooth Tenant-to-tenant Migration and Copilot Takeoffsammart93
 
MINDCTI Revenue Release Quarter One 2024
MINDCTI Revenue Release Quarter One 2024MINDCTI Revenue Release Quarter One 2024
MINDCTI Revenue Release Quarter One 2024MIND CTI
 

Kürzlich hochgeladen (20)

Strategies for Landing an Oracle DBA Job as a Fresher
Strategies for Landing an Oracle DBA Job as a FresherStrategies for Landing an Oracle DBA Job as a Fresher
Strategies for Landing an Oracle DBA Job as a Fresher
 
Apidays New York 2024 - The Good, the Bad and the Governed by David O'Neill, ...
Apidays New York 2024 - The Good, the Bad and the Governed by David O'Neill, ...Apidays New York 2024 - The Good, the Bad and the Governed by David O'Neill, ...
Apidays New York 2024 - The Good, the Bad and the Governed by David O'Neill, ...
 
Exploring Multimodal Embeddings with Milvus
Exploring Multimodal Embeddings with MilvusExploring Multimodal Embeddings with Milvus
Exploring Multimodal Embeddings with Milvus
 
Apidays New York 2024 - Accelerating FinTech Innovation by Vasa Krishnan, Fin...
Apidays New York 2024 - Accelerating FinTech Innovation by Vasa Krishnan, Fin...Apidays New York 2024 - Accelerating FinTech Innovation by Vasa Krishnan, Fin...
Apidays New York 2024 - Accelerating FinTech Innovation by Vasa Krishnan, Fin...
 
FWD Group - Insurer Innovation Award 2024
FWD Group - Insurer Innovation Award 2024FWD Group - Insurer Innovation Award 2024
FWD Group - Insurer Innovation Award 2024
 
ProductAnonymous-April2024-WinProductDiscovery-MelissaKlemke
ProductAnonymous-April2024-WinProductDiscovery-MelissaKlemkeProductAnonymous-April2024-WinProductDiscovery-MelissaKlemke
ProductAnonymous-April2024-WinProductDiscovery-MelissaKlemke
 
Architecting Cloud Native Applications
Architecting Cloud Native ApplicationsArchitecting Cloud Native Applications
Architecting Cloud Native Applications
 
MS Copilot expands with MS Graph connectors
MS Copilot expands with MS Graph connectorsMS Copilot expands with MS Graph connectors
MS Copilot expands with MS Graph connectors
 
Polkadot JAM Slides - Token2049 - By Dr. Gavin Wood
Polkadot JAM Slides - Token2049 - By Dr. Gavin WoodPolkadot JAM Slides - Token2049 - By Dr. Gavin Wood
Polkadot JAM Slides - Token2049 - By Dr. Gavin Wood
 
Finding Java's Hidden Performance Traps @ DevoxxUK 2024
Finding Java's Hidden Performance Traps @ DevoxxUK 2024Finding Java's Hidden Performance Traps @ DevoxxUK 2024
Finding Java's Hidden Performance Traps @ DevoxxUK 2024
 
Artificial Intelligence Chap.5 : Uncertainty
Artificial Intelligence Chap.5 : UncertaintyArtificial Intelligence Chap.5 : Uncertainty
Artificial Intelligence Chap.5 : Uncertainty
 
Cloud Frontiers: A Deep Dive into Serverless Spatial Data and FME
Cloud Frontiers:  A Deep Dive into Serverless Spatial Data and FMECloud Frontiers:  A Deep Dive into Serverless Spatial Data and FME
Cloud Frontiers: A Deep Dive into Serverless Spatial Data and FME
 
Six Myths about Ontologies: The Basics of Formal Ontology
Six Myths about Ontologies: The Basics of Formal OntologySix Myths about Ontologies: The Basics of Formal Ontology
Six Myths about Ontologies: The Basics of Formal Ontology
 
Cloud Frontiers: A Deep Dive into Serverless Spatial Data and FME
Cloud Frontiers:  A Deep Dive into Serverless Spatial Data and FMECloud Frontiers:  A Deep Dive into Serverless Spatial Data and FME
Cloud Frontiers: A Deep Dive into Serverless Spatial Data and FME
 
[BuildWithAI] Introduction to Gemini.pdf
[BuildWithAI] Introduction to Gemini.pdf[BuildWithAI] Introduction to Gemini.pdf
[BuildWithAI] Introduction to Gemini.pdf
 
Platformless Horizons for Digital Adaptability
Platformless Horizons for Digital AdaptabilityPlatformless Horizons for Digital Adaptability
Platformless Horizons for Digital Adaptability
 
Connector Corner: Accelerate revenue generation using UiPath API-centric busi...
Connector Corner: Accelerate revenue generation using UiPath API-centric busi...Connector Corner: Accelerate revenue generation using UiPath API-centric busi...
Connector Corner: Accelerate revenue generation using UiPath API-centric busi...
 
Apidays New York 2024 - Passkeys: Developing APIs to enable passwordless auth...
Apidays New York 2024 - Passkeys: Developing APIs to enable passwordless auth...Apidays New York 2024 - Passkeys: Developing APIs to enable passwordless auth...
Apidays New York 2024 - Passkeys: Developing APIs to enable passwordless auth...
 
Strategize a Smooth Tenant-to-tenant Migration and Copilot Takeoff
Strategize a Smooth Tenant-to-tenant Migration and Copilot TakeoffStrategize a Smooth Tenant-to-tenant Migration and Copilot Takeoff
Strategize a Smooth Tenant-to-tenant Migration and Copilot Takeoff
 
MINDCTI Revenue Release Quarter One 2024
MINDCTI Revenue Release Quarter One 2024MINDCTI Revenue Release Quarter One 2024
MINDCTI Revenue Release Quarter One 2024
 

Extrapolation Kent Feb10

  • 1. On extrapolation Federica Russo Philosophy, Kent
  • 2. Overview Extrapolation: an old problem Extrapolation in social science: the background Cook & Campbell Guala and Steel Virtues and vices of mechanism-based approaches to extrapolation How to improve the mechanism-based approach Modelling vs. using mechanisms Role of populational properties 2
  • 3. The evergreen riddle of induction Goodman’s new riddle Things are GRUE If they are green before a certain time t If they are blue and not examined before time t. All Emeralds are green … but are they also grue? Distinguishing between projectible and non-projectible properties The new riddle is a form of extrapolation 3
  • 4. The riddle of extrapolation Extrapolation, or external validity, is the inference by which the results of one study, e.g. an experiment, are extended to a larger or a different population or to a different setting Distinguishing between projectible and non-projectible results animal models and human models; experimental economics and real world economic situations; demographic models across different countries; … 4
  • 5. Extrapolation in social science:Cook & Campbell Validity: the best available approximation of the truth of causal statements Types of validity Internal: confidence in causal relation within population External: confidence in generalising to other populations 5
  • 6. Exporting results:an issue of external validity Generalisingto and across populations External validity refers to the approximate validity with which we can infer that the presumed causal relationship can be generalized to and across alternate measures of the cause and the effect and across different types of persons, settings, and times. (Cook and Campbell, 1979, p. 37) Generalizing to well-explicated target populations should be clearly distinguished from generalizing across populations. Each is germane to external validity: the former is crucial for ascertaining whether any research goal that specified populations have been met, and the latter is crucial for ascertaining which different populations (or subpopulations) have been affected by a treatment, i.e., for assessing how far one can generalize. (Cook and Campbell, 1979, p. 71) 6
  • 7. The assessment of external validity When? How? Results of test of statistical interactions Between the selection of individuals to take part in the study and the treatment / intervention Representativeness of sample Between history (= particular conditions of the study/experiment) and treatment Possibility to replicate studies 7
  • 8. Extrapolation entersthe philosophical debate An emergent awareness For once, from the right angle: Analogical reasoning and mechanisms 8
  • 9. Guala: analogical reasoning External validity rests an empirical problem Solution: wisely combine field and laboratory evidence in analogical reasoning 9
  • 10. The structure of such an inference can be reconstructed as follows: 1. If all directly observable features of the target and the experi- mental system are similar in structure; 2. If all the indirectly observable features have been adequately controlled in the laboratory; 3. If there is no reason to believe that they differ in the target system; 4. And if the outcome of the two systems at work (the data) is similar; 5. Then, the experimental and target systems are likely to be structurally similar mechanisms (or data-generating processes). (Guala, 2005, p. 180) 10
  • 11. Steel: comparative process tracing The extrapolator’s circle Challenge of successfully exporting information from model population knowing that it is limited and partial The problem of difference Challenge of providing successful methods in presence of differences between the model and the target population Solution: Comparing the mechanisms in the model and in the target at the points in which they are more likely to differ 11
  • 12. Thus, efficient applications of comparative process tracing can focus on likely sources of difference in downstream stages of the mechanism. (Steel, 2008, p. 90, emphasis in the original) 12
  • 13. Mechanism-based extrapolation is a step beyond Cook & Campbell tradition What’s the scope of mechanism-based approaches? Is all extrapolation practice mechanism-based? Many extrapolation practices aren’t mechanism-based, although arguably they should 13
  • 15. Beyond ‘nichilist’ stanceà la LaFollette & Schanks Causal Analogue Models: can grant extrapolation Hypothetical Analogue Models: can only suggest hypotheses to test Animal models can only be HAM Extrapolation possible only if there are no differences … but that’s exactly the problem! 15
  • 16. Field knowledge is important Beyond statistically-minded tradition of Cook and Campbell Stress only representativeness of samples and possibility to replicate studies 16
  • 17. Against detractors of extrapolation … Even by Cook and Campbell! The priority among validity types varies with the kind of research being conducted. For persons interested in theory testing it is almost as important to show that the variables involved in the research are constructs A and B (construct validity) as it is to show that the relationship is causal and goes from one variable to another (internal validity). Few theories specify crucial target settings, population, or times to or across which generalization is desired. Consequently, external validity is of relatively little importance.[…] For investigators with theoretical interests our estimate is that the types of validity, in order of importance, are probably internal, construct, statistical conclusion, and external validity. Cook and Campbell (1979, p. 83) 17
  • 18. Luckily, contrary voices exist: A primary goal in all sciences, including the social sciences, is the production of general knowledge. General knowledge is knowledge that is not confined to the particulars of time and place. Lucas (2003, p. 236) 18
  • 19. Scope of mechanism-based extrapolation 19
  • 20. Guala: Sometimes external validity takes the form of “the in vitro–in vivo problem” (biochemistry), sometimes it is called “ecological validity” (psychology), and sometimes it is called “parallelism” (economics), but the issue is always the same. (Guala, 2005, p.160) 20
  • 21. Steel: The best way to introduce to topic of this book is with a few examples. Studies find that a particular substance is carcinogenic in rats. We would like to know whether it is also such in humans. A randomized controlled experiment has found that a pilot welfare-to-work program improved the economic prospects of welfare recipients. It is desired to know whether the program will be similarly effective in other locations and when implemented on a larger scale. On the basis of a controlled experiment concerning outcomes resulting from initiating anti-retroviral therapies earlier or later among HIV+ patients, a physician wishes to decide the best time to initiate this therapy for the patients she treats. [. . . ] I will use the term extrapolation to refer to inferences of this sort. (Steel, 2008, p. 3) 21
  • 22. Too much in the same basket? Are external validity issues / methods the same in domains as different as economics and biology? An old issue arises again Are the natural and social sciences completely apart? Or is it a problem of having more direct and independent access to social or biological mechanisms? 22
  • 23. Varieties of external validity inferences
  • 24. External validities Internal/external: within/outside the sample Population: different populations of subjects Ecological: same subjects, different settings Temporal: same subjects, same settings, different times Do different types of validity require different methods of extrapolation? What population are we talking about? Cook & Campbell: maximise representativeness of sample Mechanism-based extrapolation marks step beyond, yet, not enough emphasis on the role of socio-demo-political characteristics of populations 24
  • 25.
  • 26.
  • 27.
  • 28. How to improvethe mechanism-based approach
  • 29. Modelling and using mechanisms Modelling Find out what the mechanism is: what it is made of, how it functions Infer the mechanism from observations and experiments Using For explanation Mechanisms carry explanatory power because they display how the phenomenon was brought about For external validity Mechanisms used in various ways 29
  • 30. What population are we talking about? Recall the many external validities on the market: Internal/external: within/outside the sample Population: different populations of subjects Ecological: same subjects, different settings Temporal: same subjects, same settings, different times But what counts as ‘same’ or ‘different’ population? 30
  • 31. The overlooked role ofpopulational properties E.g. demographic, socio-political-economic characteristics for the choice of variables or of proxies for some properties or of the statistical model, for the interpretation of results … The possibility to extrapolate strongly depends on the properties of the populations The process of extrapolation itself requires comparing the properties of the populations This has been overlooked in both the statistically-minded and the mechanism-based approach 31
  • 32. Thus, efficient applications of comparative process tracing can focus on likely sources of difference in downstream stages of the mechanism. A few important qualifications about the emphasis on downstream stages should be noted. The strategy could lead to mistaken conclusions if there is a path that bypasses the downstream stage. [. . . ] Second, the mark that upstream stages leave upon the downstream stages must be distinctive in the sense that it could not have resulted from some independent causes. (Steel, 2008, p. 90) 32
  • 33. Downstream … where? Appealing to the properties of the population does not exactly coincide with “the sources of difference in downstream stages of the mechanism” identified by Steel. is finding downstream differences about the populations themselves. 33
  • 34.
  • 35. Selected bibliography Cook, T. and Campbell, D. (1979). Quasi-Experimentation. Design and Analysis Issues for Field Settings. Rand MacNally, Chicago. Godfrey-Smith, P. (2003). Goodman's problem and scientic methodology. The Journal of Philosophy, 100(11):573-590. Goodman, N. (1955). Facts, Fiction, and Forecast. Cambridge University Press, Harvard. Guala, F. (2005). The methodology of experimental economics. Cambridge University Press. LaFollette, H. and Schanks, N. (1995). Two models of models in biomedical research. Philosophical Quarterly, 45:141-160. Lucas, J. W. (2003). Theory-testing, generalization, and the problem of external validity. Sociological Theory, 21(3):236-253. Steel, D. (2008). Across the boundaries. Extrapolation in biology and social science. Oxford University Press. 35