Talk:Decision analysis and risk management

From Testiwiki
Jump to: navigation, search

Comments about DARM course (18.3.2011):

  • Are these methods used in the actual decision situation/in real world?
  • Some of issues too complicated (especially Bayes lectures);
  • Links of this course to other issues taught in environmental risk assessment;
  • More concrete examples would be useful;

Some things to take into account in planning/arranging/conducting the course

  1. Harkkatyön ohjeistukset: Category:DARM exercise. Ainakin RM harkan (osa 2) ohjeistuksessa lienee täsmentämistä. Muutenkin saa parannella niin paljon kuin osaa.
  2. Kunhan tiedostojen jakelun systeemi saadaan taas tavalla tai toisella pelaamaan, N:\YMAL\Presentations\2011\0228_DARM\materials sisältää jakeluun sopivaa tavaraa. Ainakin NRC:n kirjojen executive summaryt minulla oli alunperin tarkoituksena panna heande-sivulle (m-filesiin).
    • Understanding risk –kirjasta en saanut executive summarya pdf:na ladattua nap.edu:sta. Voisi vaikka skannata, jos ei muuten onnistu.
  3. Luento-/harkka-aikataulu on vielä aika epämääräinen ja sekava, etenkin loppupäästään (RM-osio). Jos kokonaiskuva alkaa kirkastumaan (esim. tämänpäiväisten kokemusten pohjalta), saa täsmentää kurssisivun kalenteriin vaikkapa sen kummemmin minulta kyselemättä.
  4. Sikaflunssatarinaakin voisi vähän miettiä. Esim.:
    • Taustamateriaalilinkkejä Decision analysis and risk management#General information sources on AH1N1 influenza (Swine_flu) voisi tarkistella. Kirsti on niitä keräillyt ja varmasti ovat tolkullisia, mutta olisihan se hyvä jonkun muunkin tietää mitä kaikkea sieltä löytyy. Olisiko Teemu sopiva henkilö tähän?
    • Aikataulun ja linkkilistan aihe-/vaihejaottelu eivät ole aivan yhdenmukaiset. Sikaflunssastoorituokioiden rajauksia ja niiden sijoittelua aikatauluun voisi varmaankin parantaa/tarkentaa nykyisestä. Jouni & Marko, jos ehditte?
    • Edellisen yhteydessä voisi myös miettiä, jos sikaflunssalinkkejä nostelisi sivun lopusta aikataulun yhteyteen sopiviin kohtiin.
  5. Kunhan alkaa selviämään kuinka monta ryhmää ja yksilöä tulee harkat tekemään, voidaan miettiä tarvitaanko kahta päivää loppuseminaariin. Nyt seminaarivaraus on aikataulussa sekä 11.4. että 12.4.
  6. En vielä kovin hyvin tiedä miten RM osio tulee kurssin loppupuolella järjestymään, mutta olisiko ideaa varata 1-2 päivää lopusta koko kurssin kokonaiskuvan kertaamiseen ja esille nousevien/nousseiden erityiskysymysten lisätarkasteluun? Esim. DA summary & specific topics 7.4. ja RM summary & specific topics 8.4. Tai jotain sinne päin.
    • --# : Ehkä yksikin sessio riittäisi tähän. 8.4. tai 11.4.? --Mikko Pohjola 13:16, 4 March 2011 (EET)

--Mikko Pohjola 13:08, 4 March 2011 (EET)

Draft synopsis

For a draft synopsis of the course, see a previous version.

Case study exercise

Case study exercise description and instructions can be found at: Category:DARM exercise.

Lectures related to uncertainty

I was briefly discussing with Jouni from the uncertainty, Monte-Carlo and Value of Information path of the lectures.

I think this “path” has three main pieces:

  • What is uncertainty, how to identify uncertainty etc. This could include all kind of issues like expert elicitation, data analysis, literature search, expert estimates, modelers estimate etc.
  • How to implement uncertainty to the calculations? Thus, how to actually combine uncertainties from different sources and/or how to create uncertainty functions. This could be divided to two different parts, Bayes and Monte-Carlo, since these two are the two most common ways of implementing uncertainty to the assessment.
  • Third phase consist sensitivity analysis and value of information. Thus, these are methods that are applied after the model is done and after the uncertainties have been implemented.

I think the basic idea of this is already included in the course structure, but we could try to make it clearer in the lecturing structure?


Responsibilities among organizers

The course basically consists of three temporal/contentual sections:

  1. Introduction 28.2.-4.3.
  2. DA 8.3.- 29.3.
  3. RM 30.3.-12.4.

The two main teaching/studying methods are employed in parallel more or less throughout the course:

  • Theory lectures and exercises (classroom)
  • Case discussions and group/independent work (classroom and outside)

In order to allow us to concentrate on certain things below is suggestion for distribution of (primary) responsibility on certain issues along the course

Intro

Jouni and Mikko to agree on who leads which parts, and prepare corresponding teaching materials. Basically exercise introduction and instruction prepared by Mikko, general course arrangement basically a responsibility of Jouni. Marko to participate and help as available/needed.

DA

Marko and Jouni to agree on who leads which parts, and prepare corresponding teaching materials. Also developing the model for demonstrations a responsibility of Jouni and Marko (Can the study plans also be executed with the demonstrator model (updated according to needs of the plans)?). Jouni to create and keep up the linkage between case study/exercise and theory. Mikko will be back to participate in the exercise part 1 presentations and discussions. Also Mikko is available if assessment performance is wanted to be taken up in DM under uncertainty.

RM

Mikko to plan the story-/timelines for both theory and case for the last section. More detailed distribution of lead responsibilities to be agreed upon by Jouni and Mikko. Follow-up and assistance for post-seminar improvement work on exercises needs to arranged. Also exercise output checks and student evaluations need to be agreed upon.

Other

In addition there are certain general issues that also need to be agreed upon and taken care of, e.g:

  • collection and preparation of study and background materials for both theory and case to be done by whom?
  • classroom reservations to be done by whom? (us or UEF?)
    • needs for specific equipment or space needs must be clearly indicated in the course schedule
  • video recording / streaming of lectures? (which lectures are suitable?)
  • is there a need for an increased continuous Opasnet up-keeping or help desk -service during the course?
  • requirements/needs for keeping up participation lists? how is this to be done?

theory vs. story vs. case study exercise

An idea for creating and keeping up a tight linkage between the swine-flu story and the theory lectures as well as case study exercises: Let us make the detailed theory lecture/exercise plans, so that in every possible theory topic there is a connection (at least a weak one) to the swine-flu story. It can be done e.g. by:

  • taking lecture/exercise examples from the case to demonstrate the theory
  • after presenting the theory, making students work out what the particular part of theory would mean in the case context
  • making students explore and discuss a certain aspect(s) of the case as an introduction to the particular part of theory, presented afterwards.

The theory and the exercise are now implemented in the schedule to progress in unison. If a good linkage between the case story and theory lectures/exercises is achieved, it also guarantees contentual support for progressing the case study exercises. It is not necessary to be strictly bound to the chronological order of events in the story (the story parts will probably naturally fall into a nearly chronological order if in line with the theory and cases study exercise plans).