Anna Alexandrova

My research:

Social sciences and expertise in the age of data

What sort of knowledge can and should social sciences deliver? Is it in any way distinctive or is the category 'social science' just a label? How should our vision of social sciences change with the recent rapid proliferation of new technologies of machine learning and data analysis? I study ways in which social knowledge is produced, justified, and presented and how the ideals of what counts as good social science, reliable expertise, and responsible prediction are changing. My work with the Leverhulme Centre for Future of Intelligence since 2017 and the Centre for the Humanities and Social Change since 2018, engages the various questions in this area:

• I co-authored the report "Ethical and Societal Implications of Algorithms, Data, and Artificial Intelligence: A Roadmap for Research" for Nuffield Foundation and a paper "The Role and Limits of Principles in AI Ethics" that develops the themes of this report.
• I was part of the Limits of the Numerical project in 2015-2018 that examined the roles of quantification in social sciences and policy.
• I am a PI on the project Expertise Under Pressure, which studies ways in which experts overreach when addressing complex and value-laden phenomena.

Sciences of well-being

Is well-being a legitimate object of science and how can well-being research be practiced in a responsible way? In my writings I urge that sciences and policy should be open to multiple concepts of well-being and multiple methodologies, as well as a wider non-technocratic conception of measure validation that takes seriously the necessity for value judgments and stakeholder consultation. I argue that philosophy should develop mid-level, rather than high, theories to enable empirical work.

Recorded talks and public lectures on this topic:

Justifying Measures of Well-being: Between Psychometrics and Philosophy (Leeds, 2012)

Are social scientists experts on values? (ENPOSS keynote lecture, Krakow 2017)

Defining Mental Health (LSE Popper Seminar 2018)

When Well-being Becomes a Number (UCL STS Department Seminar, 2018)

Other writings on well-being:

"Happiness Economics as Technocracy" (with Ramandeep Singh), Behavioural Public Policy. We raise objections against one strand of happiness economics that attempts to quantify a stable effect of policies on well-being and to base public policy on this quantity.

"Mental Health Without Wellbeing" (with Sam Wren-Lewis) forthcoming in the Journal of Medicine and Philosophy. We argue that mental health is constituted by psychological preconditions of wellbeing rather than being identical to well-being or being merely absence of mental illness.

Review of "The Origins of Happiness" by Clark et al (with Ramandeep Singh), Times Literary Supplement, Nov 23rd 2018. (journal version)

"Doing Well In The Circumstance" Journal of Moral Philosophy. 2013, 10/3: 310-328. I argue that well-being claims obey contextualist semantics.

"First Person Reports and The Measurement of Happiness" Philosophical Psychology 2008. I analyze the debate between hedonists and life satisfactionists in psychology, arguing that there is no straightforward superiority of former over the latter.

"Values and the Science of Well-being: A Recipe for Mixture" in the Oxford Handbook for Philosophy of Social Science, ed. H Kincaid, New York: Oxford University Press, 2011, 625-645. I survey the meanings of well-being in social sciences raising doubts that this diversity can be handled by a single theory.

"High Fidelity Economics" (with Dan Haybron) in the Elgar Companion to Recent Economic Methodology (eds. John Davis and Wade Hands), Edward Elgar, 2011, 94-117. We argue that the advent of happiness research is changing traditional welfare economics in beneficial ways.

"Well-being as an Object of Science" Philosophy of Science 2012, 79/5:678-689. I argue that well-being sciences do not currently have a good way of securing normative adequacy of measures of well-being.

"Towards a Theory of Child Well-being" (with Ramesh Raghavan) Social Indicators Research. 2015, 121/3: 887-902. We argue that the current study of child well-being lacks an underlying theory of child well-being and make inroads towards formulating one.

"Well-being and Philosophy of Science" Philosophy Compass (journal version). 2015, 10/3:219-231. I offer a survey of the sciences of well-being from the point of view of philosophy of science and explain how philosophy of science could contribute to this field.

"Can the Science of Well-being be Objective?" British Journal of Philosophy of Science (journal version) Volume 69, Issue 2, 1 June 2018, Pages 421–445. I formulate the distinct sense in which social and medical sciences concerned with well-being are value-laden, argue that these values should not be purged and propose principles for responsible treatment of values in the science of well-being.

"Is Construct Validation Valid?" (with Dan Haybron) Philosophy of Science (journal version).83, no. 5 (December 2016): 1098-1109. We explain why social scientists believe they can measure subjective well-being in a valid way, but point out that the existing procedures for validation wrongly avoid philosophy.

"Is Well-being Measurable After All?" Public Health Ethics 2016 (journal version). This is a commentary on Dan Hausman's Valuing Health. Against Hausman's contention that well-being cannot be measured I argue that it could be, provided we adopt contextual notions of well-being and a realistic model of measurement.

Models, Explanation, and Institutions

This project stems from my 2006 doctoral dissertation. I am interested in understanding rational choice models originating in economics and now used all over social sciences. What are researchers and users warranted to conclude on the basis of these models? What is their exact role in historical explanation and policy making? What is progress in model-based social science? Is modelling as efficient of a method of inference as its high profile indicates? (On the last question I am growing increasingly skeptical). Main papers:

"Making Models Count" Philosophy of Science July 2008, 75:383-404. This is my first statement of the open-formula view of models and its advantages for understanding the role of models in mechanism design. Winner of The Philosophy of Science Association Recent PhD Essay Prize

"Progress in Economics: Lessons from the Spectrum Auctions" (with Robert Northcott) in Harold Kincaid and Don Ross (Eds.) Oxford Handbook for Philosophy of Economics, Oxford: Oxford University Press, 2009, 306-336. We give a fuller statement of the open-formula view explaining its superiority over alternative accounts. The article includes an extended account of auction design and explores its consequences for progress in economics.

"When Analytic Narratives Explain" The Journal of Philosophy of History March 2009: 3/1, 1-24. I analyze the role of rational choice models in historical explanation.

"Buyer Beware: Robustness Analyses in Economics and Biology" (with Jay Odenbaugh) in Biology and Philosophy 2011, 26:757–771, DOI 10.1007/s10539-011-9278-y. We criticize the appeal to robustness analysis of toy models arguing that it is not a method of confirmation of empirical claims.

"It's Just a Feeling: Why Economic Models Do Not Explain" (with Robert Northcott) in The Journal of Economic Methodology. (Commentary on Julian Reiss's "Explanation Paradox"). Here is the journal version. We argue that the most defensible reply to Reiss's Explanation Paradox is to reject the premise that economic models explain.

"The Prisoner's Dilemma doesn't explain much" (with Robert Northcott) in Martin Peterson (ed.), The Prisoner's Dilemma, Cambridge University Press, 2015. We use a case study of the WWI live-and-let-live system to show that Prisoner's Dilemma is not the powerful explanatory tool it is believed to be.

• I talked about this research in this discussion organised by the Cambridge Centre for Economic Pluralism.