The past three decades we have observed a phenomenal pivoting towards Judgmental Forecasting (JF) which actually demonstrates the importance of “critical thinking” in providing accurate forecasts. Although there are still cases where judgmental forecast adjustments are not suggested (eg in the case of steady processes with lots of historical data), the belief that judgmental intervention forms a deteriorating factor in forecast formulation is no longer the ground rule.
The present research proposal falls under the general umbrella of forecasting, and in particular, Judgmental Forecasting (JF). Its main focus is to build on the findings of the Good Judgment Project (GJP). A wishable outcome of the proposed research is an enhanced methodology capable of serving as an indispensable instrument for providing accurate forecasts to decision makers.
The principle differentiation of the present research procedure is its focus in forecasting events with minimum to little historical data. This particular characteristic deprives the potential forecaster from using effectively a wide variety of tools, like extrapolation, time series analysis etc, thus making forecast accuracy an ambivalent outcome. In particular the current research will focus on the use of a modified version of “Structured Analogies”, in combination with the “Superforecasting” concept, as described by Philip Tetlock.
- Define an accurate – reproducible procedure to help identify potential Superforecasters in any environment (public, corporate, etc)
- Identify Superforecasters (elite forecasters)
- Effectively combine forecasts supplied, in order to provide accurate estimations to the decision making authorities.
The research protocol that I will follow, is depicted bellow, in the form of a flow chart:
All participants that conclude the 6 month research cycle (answer to all questions being set), are entitled to the bellow benefits, regardless of their scoring/performance:
- Certification from the University of Peloponnese/Department of Economics, stating that they successfully attended the corresponding training and completed a scientific tournament in judgmental forecasting,
- Participation free of charge, in a Project Management Professional [PMP(r)] 35 hour long certification preparation course, provided by the researcher and hosted by the University of Peloponnese/Department of Economics (Tripolis Campus).
Apply for participation
- Create an email account in the form of (letter or w>)(6 random numbers) @domain.com eg email@example.com (anonymity protection)
- The “u” letter in the email address will be used by university students, where as the “w” for everybody else around the world.
- Send request for participation to firstname.lastname@example.org
- Upon approval you will receive further guidance on how and when to access the forecasting platform, that was created in order to facilitate the purposes of the current research.
The experiment is open for everyone to participate, under the terms described by the Lærd Dissertation 2016:
“As a researcher, I need to:
- Obtain informed consent from potential research participants;
- Minimise the risk of harm to participants;
- Protect their anonymity and confidentiality;
- Avoid using deceptive practices; and
- Give participants the right to withdraw from the research”
For even more information on the research subject, you can have a look at the following video tutorial.