Abstract
An ongoing controversy in philosophy of statistics is the so-called “likelihood principle”
essentially stating that all evidence which is obtained from an experiment about an
unknown quantity θ is contained in the likelihood function of θ. Common classical
statistical methodology, such as the use of significance tests, and confidence intervals,
depends on the experimental procedure and unrealized events and thus violates the
likelihood principle.
The likelihood principle was identified by that name and proved in a famous paper by Allan
Birnbaum in 1962. However, ever since both the principle itself as well as the proof has
been highly debated. This presentation will illustrate the debate of both the principle and
its proof, from 1962 and up to today. An often-used experiment to illustrate the controversy
between classical interpretation and evidential confirmation based on the likelihood
principle will also be presented and discussed
Original language | English |
---|---|
Title of host publication | Symposium i anvendt statistik 2015. |
Number of pages | 8 |
Publisher | Danmarks Statistik |
Publication date | 2015 |
ISBN (Print) | 978-87-501-2171-8 |
Publication status | Published - 2015 |
Event | 37th Symposium i Anvendt Statistik 2015 - Danmarks Tekniske Universitet, Kongens Lyngby, Denmark Duration: 26 Jan 2015 → 28 Jan 2015 Conference number: 37 http://www.statistiksymposium.dk/ |
Conference
Conference | 37th Symposium i Anvendt Statistik 2015 |
---|---|
Number | 37 |
Location | Danmarks Tekniske Universitet |
Country/Territory | Denmark |
City | Kongens Lyngby |
Period | 26/01/2015 → 28/01/2015 |
Internet address |