Causality and responsibility judgments

Lead Research Organisation: University College London
Department Name: Experimental Psychology

Abstract

Abstracts are not currently available in GtR for all funded research. This is normally because the abstract was not required at the time of proposal submission, but may be because it included sensitive information such as personal details.

Publications

10 25 50
publication icon
Alicke MD (2015) Causal Conceptions in Social Explanation and Moral Evaluation: A Historical Tour. in Perspectives on psychological science : a journal of the Association for Psychological Science

publication icon
David Lagnado (Co-Author) (2011) Blame the skilled

publication icon
Kominsky JF (2015) Causal superseding. in Cognition

publication icon
Lagnado D (2013) Legal idioms: a framework for evidential reasoning in Argument & Computation

 
Description How do people allocate blame to different players when their football team loses, or to partners when a business venture fails, or to members of a gang for a group crime? This is a difficult problem because group members often differ in terms of their actual contributions, their intentions, knowledge and skills, and their role in the group. This project investigates how lay people attribute credit and blame in group contexts, seeking to identify common principles across various domains (e.g., law, business, team games). To explore these questions we developed a general framework for responsibility judgments based on formal work in causal modelling (Pearl, 2009).To test this framework we ran psychological studies using interactive online games and tasks to represent real world situations. Participants played or watched group games, and assigned responsibility to players for group outcomes. Key findings: (i) As predicted by our theoretical model, people were sensitive to the causal structure of a task, adjusting their attributions according to whether each agent's contribution 'made a difference' to the group outcome (Zultan, Gerstenberg & Lagnado, 2012). People assessed the extent to which an agent made a difference by comparing what actually happened to what would have happened if the agent had done something different. Moreover, in cases where the group outcome was overdetermined by several members, individual members were assigned repsonsiblity according to how close they were to making a difference. For example, when all members needed to succeed individually for the team to win, each member received high credit. In contrast, when only one member needed to succeed, members received less credit, even if they all succeeded. Their individual credit depended on how many changes were needed to turn them into a difference-maker. (ii) When assigning responsibility people cared not only about what actually happened in the game, but also about what was expected to happen before the game was played. Both prospective and retrospective factors determined people's final judgments of responsibility. We therefore revised our theoretical model, such that judged responsibility was a function both of pivotality - to what extent an agent made a difference to the outcome, and criticality - how much the agent was expected to contribute to the outcome, before the game was played (Lagnado, Gerstenberg & Zultan, 2013). This meshes with theories of legal and moral responsibility, and captures the everyday ambiguity in the term 'responsibility', which is both forward and backward looking. The new model fitted empirical results from various new studies that manipulated both pivotality and criticality. (iii) People also cared about the intentions, knowledge and skill of the group members. They attributed more blame (and punishment) to agents with bad intentions, even when this did not result in a bad outcome (Schächtele, Gerstenberg & Lagnado, 2011). They also assigned more responsibility to players who knew their actions might make a difference (Gerstenberg & Lagnado, 2012) and to players with higher skill (Gerstenberg, Ejova & Lagnado, 2011). The project also explored two related questions: (a) How do people construct their initial causal models of a situation? How do they translate messy real world contexts into causal models from which they can generate judgments of responsibility? To explore this question we focussed on the legal domain, and developed a novel theoretical account in which people use a small set of 'causal idioms' that can be combined and reused to model a large-scale problem. Ongoing empirical studies have supported this model (Lagnado et al., 2012; Fenton, Lagnado & Neil, 2013). (b) How do people communicate their causal models to other parties? To investigate this question we used a game playing domain (loosely based on football) in which participants observed two agents competing in a game, and had to describe their interactions to a third person. We found extensive use of causal and blame language, and these descriptions strongly predicted how participants assigned credit and blame to the players. In the future we plan to extend all these lines of research, with a special focus on causal and responsibility attributions in the legal domain.
Exploitation Route This project has led to five publications in international journals (including Cognition, Cognitive Science and Psychonomic Bulletin & Review) and five conference papers in the Proceedings of the Cognitive Science Society (see ROS for full details). Several more articles will be published in the next 12 months, including four more proceedings papers, two book chapters targeted at broader audiences (Oxford studies in Experimental Philosophy and Oxford Studies in Agency and Responsibility). Also in press is an article in the Annual Review of Psychology (co-authored with Steven Sloman). In addition several more articles are in preparation, including an invited chapter for the Oxford Handbook of Causal Reasoning, an invited review article for Perspectives on Psychological Science, and two empirical papers to present the final series of experiments from the project. These articles and book chapters should promote dissemination of the project findings to a wide academic audience in psychology, philosophy and cognitive science. I am also writing a more general book on causal and evidential reasoning (under contract with CUP). This is aimed at a broader audience, and will include accessible discussions of some of the project findings. The project findings have been disseminated at various international conferences, workshops and universities (see ROS for details). These include invited talks in New Orleans, Bogota, Germany, Kent, Edinburgh and Warwick. I have also organized a special symposium on causal reasoning at the International Conference of Thinking (London, 2014) and held a post-conference workshop on causality and responsibility at UCL. A particular focus over the next year will be to present the research to legal audiences. We plan to arrange seminars with the UCL Centre for Empirical Legal Studies and the UCL Judicial institute. I will also give an invited talk at an international conference on Behavioral Legal Studies at the Center for Empirical Studies of Decision Making and the Law at Hebrew University, June 9-12, 2014.

Use In Non-Academic Context
This project explores everyday attributions of responsibility so we expect it to be relevant and of interest to non-academic users. People are continually involved in judging responsibility, either as participators themselves, or as observers - e.g., in politics, business, law, finance, team games and social interactions. A clearer understanding of the principles that underpin these judgments, and the identification of any biases, is crucial to many areas of public life. To engage with non-academic users I plan to write an accessible popular book on causality and responsibility, largely based on the findings from this project. This is a natural follow-up to the academic book due to be finished this year. In parallel we will aim to engage the public through accessible articles in the popular press (e.g., the New Scientist, Economist, Financial Times etc) and via other forms of media (e.g., TV and Radio). An example of a recent popular article on our research findings can be seen at: http://www.npr.org/blogs/13.7/2012/11/05/164332198/are-you-responsible-for-the-outcome-of-tomorrow-s-election As well as disseminating the project findings, public engagement will also help to refine the principles and models developed in the project, paving the way for future research initiatives.
Sectors Communities and Social Services/Policy,Education,Government, Democracy and Justice

URL http://www.ucl.ac.uk/lagnado-lab/david_lagnado.html
 
Description This project focused on how people attribute causality and responsibility in group contexts, such as team games, law, politics, business and everyday social settings. It has generated novel theoretical and empirical findings that have been disseminated to academics in various fields, including psychology, law, computer science and business. To take the impact of this work beyond academic circles I have several ongoing initiatives: I am currently writing an introductory book (with Cambridge University Press) on the role of causal thinking in everyday and legal contexts. This book is aimed at a non-specialist audience, and will allow the project findings to be transmitted to the general public. As well as presenting new empirical findings, the book will also introduce readers to novel formal techniques for causal modeling and evidence evaluation. It is hoped that legal professionals, and others involved in public and private services, might benefit from this work. Once the book is published I plan to engage the public through accessible articles in the popular press (e.g., New Scientist, Economist, Financial Times etc) and via other forms of media (e.g., Radio and TV). I have numerous media contacts that will help this process. An example of a recent popular article on our research findings can be seen at: http://www.npr.org/blogs/13.7/2012/11/05/164332198/are-you-responsible-for-the-outcome-of-tomorrow-s-election Several of the themes in the project were extended to look at legal issues, and the understanding of complex legal evidence. One practical offshoot of this project has been my involvement in the re-analysis of legal cases with purported miscarriages of justice, where complex evidence has been poorly presented and evaluated. I am currently helping on several ongoing legal cases (preparing materials for court appeals), using some of the formal techniques developed in the project. A related consequence of the project is ongoing work on developing tutorials to teach statistical reasoning to legal professionals and laypeople (with a new proposal under review with ESRC). The long-term impact of this work would be a better understanding of complex probabilistic evidence by legal professionals and lay people. Following on from this, I am one of the primary organizers of a 6-month Programme on Probability and Statistics in Forensic Science at the Isaac Newton Institute for Mathematical Sciences, University of Cambridge, to take place in 2016. https://www.newton.ac.uk/event/fos I am also a key member of 'Bayes and the Law', an international consortium of statisticians, lawyers and forensic scientists aiming to improve the use and understanding of statistical evidence in court. https://sites.google.com/site/bayeslegal/home These activities are all geared towards better public and professional understanding of complex statistical and causal evidence in a legal context, and will build upon work developed in the project.
First Year Of Impact 2014
Sector Government, Democracy and Justice
Impact Types Cultural,Societal,Policy & public services

 
Description Causal cognition: Learning, reasoning and blame 
Form Of Engagement Activity A talk or presentation
Part Of Official Scheme? No
Primary Audience
Results and Impact Invited talk and seminars at Universidad Nacional de Colombia Bogotá, Colombia
Year(s) Of Engagement Activity 2014
 
Description Causal models in cognition: Learning, reasoning and blame 
Form Of Engagement Activity A talk or presentation
Part Of Official Scheme? No
Primary Audience
Results and Impact Invited talk at a conference on the cultural constitution of causal cognition
Year(s) Of Engagement Activity 2014
 
Description Causal stories and statistics 
Form Of Engagement Activity A talk or presentation
Part Of Official Scheme? No
Primary Audience
Results and Impact Invited talk at conference on evidence and causality
Year(s) Of Engagement Activity 2014
 
Description Causality and Responsibility small group workshop 
Form Of Engagement Activity Participation in an activity, workshop or similar
Part Of Official Scheme? No
Primary Audience
Results and Impact Post-conference workshop on causal responsibility with talks from various leading researchers and extended discussions
Year(s) Of Engagement Activity
 
Description Spreading the Blame: A Framework for Intuitive Judgments of Responsibility 
Form Of Engagement Activity A talk or presentation
Part Of Official Scheme? No
Primary Audience
Results and Impact New Orleans Workshop on Agency and Responsiblity
Year(s) Of Engagement Activity 2014
 
Description Spreading the blame 
Form Of Engagement Activity A talk or presentation
Part Of Official Scheme? No
Primary Audience
Results and Impact Invited talk summarising some of the key findings from the project
Year(s) Of Engagement Activity 2014
 
Description Thinking about causality 
Form Of Engagement Activity A talk or presentation
Part Of Official Scheme? No
Primary Audience
Results and Impact Invited talk for DFG Priority Program 'New Frameworks of Rationality'

Operationalizing Epistemic Concepts

Aachen, 10-13 September 2012
Year(s) Of Engagement Activity 2014