# Statistical Issues in Searches for New Physics

- Institution: CERN
- Department: EP Seminar
- Speaker: Louis Lyons
- Year: 2015
- Language: English
- Subjects: Physics
- Source: [CERN](https://videos.cern.ch/record/3008263)
- Licence: [CERN](http://copyright.web.cern.ch)
- Who can watch: Tweens profiles and older (sources give no ratings; placed by Invisible Network from the source's own words)

Given the cost, both financial and even more importantly in terms of human effort, in building High Energy Physics accelerators and detectors and running them, it is important to use good statistical techniques in analysing data. This talk covers some of the statistical issues that arise in searches for New Physics. They include topics such as:

Should we insist on the 5 sigma criterion for discovery claims?

What are the relative merits of a Raster Scan or a "2-D" approach?

P(A|B) is not the same as P(B|A)

The meaning of p-values

Example of a problematic likelihood

What is Wilks Theorem and when does it not apply?

How should we deal with the "Look Elsewhere Effect"?

Dealing with systematics such as background parametrisation

Coverage: What is it and does my method have the correct coverage?

The use of p0 vs. p1 plots

.

0:00:00 Slide 1

0:02:33 Slide 2

0:04:12 Slide 3

0:07:43 Slide 4

0:07:47 Slide 5

0:09:58 Slide 6

0:15:30 Slide 7

0:15:35 Slide 8

0:15:39 Slide 9

0:17:01 Slide 10

0:20:47 Slide 11

0:23:13 Slide 12

0:26:52 Slide 13

0:27:01 Slide 14

0:27:53 Slide 15

0:28:10 Slide 16

0:32:43 Slide 17

0:33:05 Slide 18

0:34:10 Slide 19

0:34:14 Slide 20

0:35:04 Slide 21

0:39:53 Slide 22

0:41:24 Slide 23

0:42:20 Slide 24

0:43:02 Slide 25

0:45:16 Slide 26

0:46:04 Slide 27

0:46:08 Slide 28

0:47:56 Slide 29

0:50:14 Slide 30

0:50:25 Slide 31

0:53:28 Slide 32

0:55:40 Slide 33

0:57:23 Slide 34

0:59:08 Slide 35

1:03:26 Slide 36

1:03:30 Slide 37

1:06:18 Slide 38

1:06:22 Slide 39

1:06:26 Slide 40

1:10:21 Slide 41

## Lectures

1. [Statistical Issues in Searches for New Physics](https://invisiblenetwork.ai/learn/cern/3008263/be2cbcc822cc), 1 h 10 min (CERN's player)