What is Robust analysis?

Does this question appeals to you?

Do you want to apply Bayesian modelling for the purpose to quantify uncertainty and support decision making? Do you feel inspired reading texts about uncertainty? Can you tell a bee from a wasp? Do you want to do a PhD in an interdisciplinary environment?

Then perhaps our new PhD project on Robust analysis is something for you. Last day to apply is October 3rd 2016.

 

Uncategorized 0

Graduate course in Risk, Uncertainty and Decision Making

Registration is now open for a PhD course in risk, uncertainty and decision making.

Register here

Last date to apply is October 2nd

Notification of acceptance will be given on September 20th

risk_measurement_400_clr_5483

We make a number of decisions every day. These decisions are based on conscious or unconscious analysis, as well as assessments of risks and uncertainties. We are often unaware of how we make these decisions, and how we compare alternatives. The decisions can be about everyday situations, like if to bring an umbrella, or about complex and urgent issues that face society, like issues connected with the environment or sustainability. The course focuses on a number of issues: How can risk be assessed? Can risk be quantified? How do risk assessments affect decision and policy making? What role does risk communication and media play? How do individuals perceive risks and risk management? How does the concept of risk influence the way we understand the world and try to control it? Risk, uncertainty, and decision analysis implies systematic efforts to understand the context and consequences of decisions.

The aim of the course is to give a deeper understanding of the theoretical perspectives and methods in risk research within different disciplines, and tools for interdisciplinary risk research. The course contains lectures and seminars. The lectures will provide rich accounts of different disciplinary perspectives.  The participants will in seminars get a chance to collaborate on questions requiring an interdisciplinary approach.

Target group: Graduate students from different disciplines

Extent: 7.5 credits

The course has two physical meetings:

  • October 13-14th in Lund (hosted by LUCRAM, Lund University), starts 9.00 the 13th, ends 15.00 the 14th
  • November 13-16th in Gothenburg (hosted by the School of Business, Economics and Law, Gothenburg University), starts 14.00 the 13th, ends 12.00 16th

Participants are expected to prepare in advance to the meetings by reading the assigned literature. The course is examined through attendance of the physical meetings and a completed paper where at least two themes of the course is discussed in relation to the participants PhD subject or field of interest.

It is possible to attend only one of the meetings and then receive a lower credit for the course (3 credits instead of 7.5).

The second meeting includes the second Nordic risk conference held in November 14-15th 2016. As course participant you have two options: (1)  To send in an abstract to the conference for an oral presentation Deadline September 13th August 29th. If accepted the participant should pay the conference fee to the conference organisers. The conference fee is 1000 SEK for PhD and master students (and 1500 SEK otherwise). (2) To participate in the conference without giving any oral presentation. If so, the course will compensate the conference fee after a completed course.

This course is funded by the research program LUCID and the research school ClimBECO and is given as part of the Lund University Centre for Risk Assessment and Managment (LUCRAM). It is open for any PhD student to attend. The number of places is limited.

Course directors: Åsa Knaggård (LU), Ullrika Sahlin (LU) och Daniel Slunge (GU).

The course will be given in English

Cost: The conference fee to attend the Nordic Chapter SRA conference

Lund – October
Thursday 13
9.30 to 12.00 Henrik Tehler (Division of Risk Management and Social Safety, LTH): Introduction to risk assessment
13.00 to 14.00 Henrik Hassel, (Division of Risk Management and Social Safety, LTH): Quantitative descriptions of risk
14.30 to 16.00 Annika Wallin (Department of Philosophy, Lund University): Individual decision making
16.00 to 17.00 Ullrika Sahlin (Centre for Environmental and Climate Research, Lund University): Probability

Friday 14
9 to 10.30 Johannes Persson, (Department of Philosophy, Lund University): Risk philosophy
11 to 12 Frederik Lundtofte (School of Economics and Management, Lund University): Risk in an economist’s perspective
13 to 15 Ullrika Sahlin and others: Mind the risk – a case-study on planning under risk and uncertainty

Göteborg – November
Sunday 13
14 to 15 Åsa Knaggård (Department of Political Science, Lund University): Policy-making under risk and uncertainty
15 to 17 Ragnar Löfstedt (King’s Centre for Risk Management, King’s College London): Risk communication in the 21st century
17.30 to 19.30 Nick Pidgeon (School of Psychology, Cardiff University): Advances in risk communication

Monday 14
9.30 to 18 Risk conference and group discussions

Tuesday 15
9 to 13 Risk conference
14.30 to 15.30 Tomas Bergström (Institute of Biomedicine, Gothenburg University): Ticks – case-study on risk perception and risk communication
16 to 17 Daniel Slunge (Centre for Environment and Sustainability, Gothenburg University): Ticks – case-study on risk perception and risk communication

Wednesday 16
9 to 10 Daniel Slunge: Ticks – case-study on risk perception and risk communication
10.30 to 12 Ullrika Sahlin: Uncertainty

Literature list [last updated 22nd September 2016]

meetings/risk research/students/teaching 2

Join the SRA-E: NORDIC meeting during the SRA conference in Bath on June 21st

The Nordic chapter of SRA Europe invites members and non-members to a chapter meeting during the SRA conference in Bath, UK.

The meeting will take place in room CB 5.6 in the Chancellors Building on Tuesday 21st June at 12.30 – 1.30pm.

It will be possible to attend the meeting via skype. Send an email to Ullrika.Sahlin [at] cec.lu.se if you want to join.

Invitation as a pdf

Info about SRA-E:NORDIC

Agenda

  1. Information about the SRA Nordic chapter

1.1. The purpose of SRA-E:NORDIC

1.2. Past activities

  1. Information of planned activities

2.1. Invitation to submit abstract and session proposals to the 2nd Nordic Chapter conference in Gothenburg Nov 2016

2.2. Planned courses in the Nordic/Baltic region (contributions are welcome)

2.3. Planned workshops in the Nordic/Baltic region (contributions are welcome)

  1. Discussions and future activities

Shared risk issues, Suggestions for policy impact, research needs etc. (contributions are welcome)


Board members:

Ullrika Sahlin, Lund University, Sweden (President for two years  – starting from June 2015)

Marja Ylönen, Technical Research Centre of Finland, Finland (President elect – to take over in June 2017 and then be president for two years)

Stian Antonsen, SINTEF, Norway

Susanne Öhman, Mid Sweden University, Sweden

Igor Kozine, Technical University of Denmark, Denmark

Aistė Balžekienė, Kaunas University of Technology, Lithuania

Student representatives:

Ingibjörg Lilja, University of Iceland, Iceland

Björn Arvidsson, Lund University, Sweden

 

Uncategorized 0

EFSAs treatment of uncertainty under trial

The way uncertainty is treated in risk assessment is under scrutiny. There is a need for this since we today face more complex problems and need to take into account possible risks for which knowledge is weak.

One example is the ongoing work in the EU on how we treat uncertainty in environmental and health risk assessments.

In March 2016 the European Food Safety Authority (EFSA) published the draft scientific Guidance document on Uncertainty in EFSA Scientific Assessment. The Guidance has been exposed to a public consultation and stakeholders provided many comments, which now have been addressed.

According to EFSA, “the Guidance is applicable to all areas of EFSA, all types of scientific assessment and all types of uncertainty affecting scientific assessment. It does not prescribe specific methods for uncertainty analysis but rather provides a harmonised and flexible framework within which different methods may be selected, according to the needs of each assessment.”

EFSA’s scientific expert panels will use the Guidance for a trial period for about two years. In 2017 the Guidance will be revised according to the problems identified. This is a great opportunity to discuss the treatment of uncertainty, not only in EFSA’s risk assessments. We wrote an oped in Sydsvenskan about this.

EFSA is the agency responsible for the risk assessments of threats posed to living organisms for the EU commission. EFSA is situated in Parma, Italy.

 

info/risk concept/risk research/Uncategorized 0

2nd Nordic chapter risk conference in Göteborg

The Nordic-Baltic chapter of Society of Risk Analysis proudly presents the call for the second Nordic chapter meeting.

The theme of the conference is “Where are we and where are we going: New insights into risk analysis in the Nordic countries”

Deadline for submission of both individual presentations and symposia is Monday 29 August.

The conference brings together individuals and organisations interested in risk assessment, risk management and risk communication in Europe. The Society of Risk Analysis Europe: Nordic Chapter promotes specifically Nordic and Baltic issues in the field of risk, and aims to facilitate debate and exchanges of information and opinion between professionals in industry, government, universities, research institutes, and consultancies. It has the ambition to convene and promote scientific and educational meetings on risk research, risk analysis and risk management in the Nordic and Baltic countries.

Themes for presentations may include:

  • risk communication and uncertainty
  • environmental governance
  • new technologies and innovation
  • risk framing
  • evidence-based policy making
  • new perspectives on risk assessment
  • risk and ethics
  • organizing in the face of risk
  • emerging risk issues in the Nordic and Baltic countries

The conference is hosted by Gothenburg Research Institute at the School of Business, Economics and Law at the University of Gothenburg.

The conference is sponsored by the Partner Program at the School of Business, Economics and Law at the University of Gothenburg.

Take the opportunity to arrange a session on a topic you think deserves to be in the spotlight or where you would like to engage in discussion and learn more.

A symposium is a topical session organized by two individuals who identifies the presenters for that session and their presentation titles, rather than relying on the conference organizers to assemble a topical session from individual abstracts.

A PhD-course on Risk is under development. The plan is to have it in Lund in October and in Gothenburg in association to the conference.

Sign up on the nordic chapter of sra newsletter to get information on what is going on.

^Ullrika

Uncategorized 0

It is time to reconsider how we think about risk

Risk is something that concerns us all.

Risk has become an integrated part of decision making and it is acknowledged that decisions should be informed by risks. This means that a decision maker may take other things than risk into account.

Even though society efficiently are managing risks, we are now more often facing risks that are new, tricky or complex.

New global challenges or new technology create phenomena for which we have little experience or limited understanding.

Managing the most easy and obvious problems first has left decision makers with the most tricky risks to regulate.

More inter-dependencies between different societal or natural systems is challenging for the detection and control of adverse system behavior.

The new, tricky and complex results in that the risks we face today are more uncertain than before. The classical view on risk as probability * consequences is challenged by

  1. the need to consider risks of different types and
  2. the need to face uncertainty of different levels of severity.

gbgstatuesTo combat the problem for risk assessors and decision makers of today, the Society for Risk Analysis have produced a glossary presenting a broader view on risk. The glossary make a clear distinction between risk as a concept and how we choose to measure it.

Further, the SRA glossary emphasis that the quantity and quality of available knowledge should guide how we choose to describe uncertainty. This means that uncertainty are treated even though our knowledge is too weak to support uncertainty being quantified by probabilities.

It remains to see if a broader view on risk can be of help to risk assessors and decision makers. I find that it may close the gap between risk-informed and evidence-based decision making. It is more important than ever that education and training of new assessors and decision makers get exposed to the ongoing advancement of the science of risk analysis.

So, take the opportunity to revise your view on risk. Perhaps you like it : )

/Ullrika

risk concept/risk research 0

Uncertain risks and decision making

Treating risk with high uncertainty occurs at many levels of political decision making. What do researchers and risk assessors say about how to understand and treat uncertain risks? Is there a way to take better responsibility for treating uncertainty at political decision making?

 arrows-167536_640

Within the advanced study group on calculating and communicating uncertainty supported by the Pufendorf Institute in Lund we wanted to reach out to those making decisions. This subject is huge and impossible to cover in one evening, but why not give it a try.

On April 28th we open the doors to Pufendorf for an informal discussion. The  invitation with more information is available here.

Ullrika Sahlin will start saying some words about the quantification of uncertainty. She will bring up the use of probability as a measure of uncertainty and the wonder of statistical principles at a simple level. She will give some examples of guidelines for risk assessors. After that Åsa Knaggård will talk about uncertainty and decision making from a political science point of view. We have the pleasure to be hosted by Prof Ragnar Löfstedt. He will (in swedish) talk about uncertainty and transparency. This is a very exiting topic and you will hear examples directly from the field of political decision making. We seek a simple and informal forum to discuss and create new connections.

 

Uncategorized 0

Bayesian Analysis for curious learners

I often get questions from students about opportunities to get started with Bayesian Analysis. Most courses in BA at Lund University is given on an advanced mathematical level, or as a few lectures (if all) in a statistics course.

Bayesian Analysis is useful both in research and decision making. Our students are the future scientists, consultants or policy makers for important environmental and societal challenges.

We should give them the possibility to expand their understanding of approaches to calculate and communicate uncertainty, in which Bayesian Analysis offer a coherent statistical framework.

We will invite students curious about Bayesian Analysis to a series of tutorials and discussions. Info about these tutorials will be posted here.

Monday Feb 29th. Bayesian modelling in JAGs. From 17 to 19. Held by Ullrika Sahlin

Code and examples from the sessoins are available on GitLab project

Coming up. Decision analysis using Bayesian Belief Networks. From 17 to 19. Held by Ullrika Sahlin

date-to-be decided. Bayesian modelling STAN

date-to-be decided. Bayesian hypothesis testing

 

students/teaching 0

Summary of the 1st Nordic Chapter Risk Conference


 

We did it!

Thank you all for making this a successful conference.

We cover the conference in our December newsletter from the Nordic Chapter of SRA-Europe.

Program SRANC_final

Homepage of the conference

Check out who attended the conference

It is possible to sign up for future newsletters on the nordic chapter pages on www.sraeurope.org or by clicking on this link.

The conference got attention in local media


Take the opportunity to go over the keynote presentations again:

Science and Proven Experience by Johannes Persson, Professor in Philosophy at Lund University, Sweden

Persson_keynote_sraNordic_Lund_2015

Risk Management in Practice: Lessons and Findings  by Åsa Boholm, Professor in social anthropology at Gothenburg University, Sweden

Boholm_keynote_sraNordic_Lund_2015

Socio-economic analysis in chemicals risk management  by Tomas Öberg, European CHemical Agency (ECHA), Finland

Oberg_keynote_sraNordic_Lund_2015

What is risk science?  by Terje Aven, Professor in Risk Analysis and Safety Management at the University of Stavanger, Norway

Aven_keynote_sraNordic_Lund_2015

How can Risk Assessment embrace the real world of Multiple Stressors, Accumulative Harm, and the Dearth of Data?  by David Gee, Visiting Fellow, Institute of Environment, Health and Societies, Brunel University, Retired Senior Adviser on Science, Policy, Emerging Issues, European Environment Agency, Denmark

Gee_keynote_sraNordic_Lund_2015

Risk risk tradeoffs: What should we do in Europe? by Ragnar Löfstedt, Professor of Risk Management and Director of King’s Centre for Risk Management, King’s College Londond, UK

Ragnar had an excellent talk without any slides.


After the conference we had a Business meeting of the SRA-E: NORDIC. We elected the board and made some decisions. Minutes are found here.

Uncategorized 0

Is Bayesian inference failing when there is a conflict between your prior and data?

Welcome to a seminar on Bayesian Inference and Prior-Data Conflict


On the 15th of December Dr. Gero Walter from Technische Universiteit Eindhoven gave a seminar about the problem of prior-data conflict, that is, when the information from data is in conflict with prior assumptions (abstract is below). His talk was followed by a hands on session where the attendees will learned how to handle and mitigate prior-data conflict in practice. The half-day was be rounded off by a complimentary lunch. 10 persons attended the seminar.

10.00 – 11.00 Talk on prior-data conflict by Dr. Gero Walter
11.15 – 12.00 Hands on exercises
12.15 – 13.00 Complimentary Lunch

The event is organised by the Advanced Study Group on Calculating and Communicating Uncertainty and takes place at the Pufendorf institute (https://goo.gl/maps/xjruS9HLzvz) at Lund University and is open to all interested and is free to attend.

Some of the exercises were made with the R statistical language. R is free and can be downloaded here: https://www.r-project.org/ Gero will do some of the exercises using his R-package luck. The package can be installed by the following lines of code

install.packages(“TeachingDemos”)
install.packages(“http://download.r-forge.r-project.org/src/contrib/luck_0.9.tar.gz”, repos = NULL, type = “source”)

Gero has prepared a document explaining Generalized Bayesian Inference with sets of conjugate priors for dealing with prior-data conflict.

The solutions to the exercises are kindly provided.

Geros talk is found at talk-lund

 

Link to the WPMSIIP workshop 2016:

The web page of the Society for Imprecise probability 

 

 

/Ullrika Sahlin on behalf of the Pufendorf Advanced Study Group on Calculating and Communicating Uncertainty 2015-2016

 

Seminar abstract

———————————

In the Bayesian approach to statistical inference, (possibly subjective) knowledge on model parameters can be expressed by so-called prior distributions. A prior distribution is updated, via Bayes’ Rule, to the so-called posterior distribution, which combines prior information and information from data into a ‘complete picture’, thus expressing our state of knowledge about model parameters after having seen the data.

A problem that then can arise is called prior-data conflict: from the viewpoint of the prior, the observations seem very surprising, i.e., the information from data is in conflict with the prior assumptions. Unfortunately, models based on conjugate priors (which allow for straight-forward calculation of the posterior), are insensitive to prior-data conflict, in the sense that the spread of the posterior distribution does not increase in case of such a conflict. The posterior then conveys a false sense of certainty, by communicating that we can quantify uncertainty on model parameters quite precisely when in fact we cannot.

It is however possible to preserve tractability and have a meaningful reaction to prior-data conflict by using sets of conjugate priors for modelling prior information. This approach, which can be seen as imprecise probability method or robust Bayesian procedure, avoids the spurious over-precision of standard Bayesian methods and allows to adequately express vague or partial prior knowledge. With the precision of prior probability statements intuitively modelled through the magnitude of the set of priors, the posterior set appropriately reflects the prior precision, the amount of data, and prior-data conflict.

 

Warning! This seminar could make you think in a different way. 2015-09-07 11.05.23

150909-ESREL_GALA_FrancescaCBruno_MS-24-300x200Ullrika and Gero together with two new made friends at the last ESREL conference in Zurich.

meetings/risk research 1