Scientific Measurement
SPA4103
Dr Alston J Misquitta, Prof William Gillin,
Dr Kostya Trachenko & Prof David Dunstan
Lecture 1 - Introduction
Module Overview
Scientific Measurement: Module Information
Module Organisers: Dr Alston J. Misquitta
Prof William Gillin
Dr Kostya Trachenko
Prof David Dunstan
(GOJ 216: office hour Mon 100 - 200)
(GOJ 121: office hour ??)
(GOJ 226: 11am to 1pm on Tuesdays)
(GOJ 124: 10am to 11am on Thursdays)
All information is online at QMplus
2 lectures per week (see QM+ for announcements):
Tuesday 1200 - 1300
Friday
0900 - 1000
1-2 lab sessions per week: either Mon+Tue or Thu+Fri
You have been assigned to these automatically!
Weeks 1-4
Weeks 5-6
Week 7
Weeks 8
Weeks 9-11
Week 12
1 lab/week
2 labs/week
lectures/lab
lectures
2 labs/week
no lab/lectures
Complete 4 lab experiments
Complete 2 lab experiments
Write up experiment 4 as formal report
Obtain formative assessment of report in your lab session
Complete one longer experiment
Write-up experiment 5 as second formal report
All deadlines are listed on QM+
Dr Alston Misquitta
No exam: 100% coursework
Scientific Measurement - Lecture 1
2
Module Overview
Labs is located on 2nd floor of Physics building:
GOJ 204 : 2-5pm
Groups: Mon/Tue or Thu/Fri
Choose lab partner or lab technician will assign for you
Read the lab manual!!!
Late work will not be marked!
Watch the SCM QM+ for changes, info, guidelines etc.
Type
Percentage
In-lab assessment 1
10%
First formal report
15%
Oscilloscope (in lab bk)
10%
In-lab assessment 2
20%
Lab-book
20%
Second formal report
25%
Late Penalties
5% per 24 hour period, changing to a mark of 0 after 120 hours (5 days).
If you have a genuine problem meeting the deadline, fill in an EC form (see office) before the deadline.
Dr Alston Misquitta
Scientific Measurement - Lecture 1
3
Module Overview
Laboratory Organisers
Dr Alston J Misquitta
Prof William Gillin
Dr Kostya Trachenko
Prof David Dunstan
Lab Technicians
Also 3/4 postgraduate
student demonstrators in
each lab session
Mr Peter Crew
Dr Alston Misquitta
Mr Saqib Qureshi
Scientific Measurement - Lecture 1
4
Module Overview
Texts available in library short loan collection - no need to buy
Good guide to
laboratory practice
£26
Good statistics reference
£26
Adrian Bevan
QC174.8 BEV
Guide to writing
reports
£17
Another good stats ref.
£15
Dr Alston Misquitta
Scientific Measurement - Lecture 1
5
Module Aims
Aims of this course:
1. To teach experimental technique
*** knowledge of underlying physics will not be tested
*** experiments not of world quality
can get high precision none the less
2. Error analysis:
*** how accurate? → purpose of these lectures!
*** develop a quantitative sense of uncertainity
3. How to obtain, manipulate, present and interpret experimental data
4. Computer control of equipment: basic programming skills
5. Automated data acquisition using microcontrollers
Plagiarism will be treated very seriously
Dr Alston Misquitta
Scientific Measurement - Lecture 1
6
Scientific Methodology - The Scientific Method
Whats the difference between these lists?
Luminferous aether
Plum pudding atomic model
Aristotelian Gravity
Dr Alston Misquitta
Quantum mechanics
Special relativity
Newtonian gravity
Thermodynamics
Scientific Measurement - Lecture 1
7
Scientific Methodology - The Scientific Method
Whats the difference between these lists?
Quantum mechanics
Special relativity
Newtonian gravity
Thermodynamics
Luminferous aether
Plum pudding atomic model
Aristotelian Gravity
Nice ideas, but flawed - proven wrong
Dr Alston Misquitta
Great pillars of modern physics!
Scientific Measurement - Lecture 1
8
Scientific Methodology - The Scientific Method
Science is driven by experiment and data
• Only experiment can distinguish between rival theories
• Only experiment can determine fundamental constants of nature
c, ħ , G, k are all derived from experiment NOT theory
Experiment is the final arbiter of Truth
Thus experimenters have a HUGE responsibility
Honesty and Integrity are paramount
Open mindedness: do not presume to know the “answer”
Do not “fiddle” results to get 'correct' answer
Provided your method is ok - Experiment is correct (almost) BY DEFINITION!
If data & theory disagree, the theory is WRONG!
Experiment tells us what the TRUTH is - Theory tells us why
Dr Alston Misquitta
Scientific Measurement - Lecture 1
9
Gravity at Large Distances
Gillies, Meas.Sci.Technol. 10(1999)421–425
How about Gravity?
After ~320 years how well
do we know G ?
Reich: Nature 466, 1030 (2010)
Summary of measurements of G
1969-1999
Many large discrepancies...
6.52
Dr Alston Misquitta
6.79
Scientific Measurement - Lecture 1
20
10
Measurement Uncertainty
Measurement is important
Lets measure the same object many times:
measure a chair several times & plot results
Why is there a spread of results?
ruler is flimsy?
40
some people can't read a ruler?
space-time is fluctuating changing the size of the chair?
45
50
Measurement of our chair should reflect the uncertainty!
write it as: 45 ± 2 cm
means TRUE value lies in range 43 - 47 cm
UNLIKELY to be >51 (45 + 3x2)
UNLIKELY to be <39 (45 - 3x2)
On a graph we show this as an error bar:
Dr Alston Misquitta
Scientific Measurement - Lecture 1
11
Dr Alston Misquitta
Scientific Measurement - Lecture 1
12
Systematic
error
How about Gravity?
After ~320 years how well
do we know G ?
Fig 2: Accuracy indicates proximity of measurem
results to the true value, precision to the repeatabilit
reproducibility of the measurement
Random
error
of G
Percentage Error
Percentage error measures the accuracy of a m
experimental value E and a true or accepted v
s...
Dr Alston Misquitta
Percentage Difference
Percentage difference measures precision of t
experimental values E1 and E2 expressed as a
13
Scientific Measurement - Lecture 1
calculate the percent difference is:
New standards for
temperature
measurements
done nearby
at the
National Physical
Laboratory
(NPL)
A low uncertainty measurement of the Boltzmann constant, de Podesta et al., Metrologia 50, 354 (2013)
Dr Alston Misquitta
Scientific Measurement - Lecture 1
14
Measurement Uncertainty
Any measurement has an uncertainty or error due to:
• equipment
• definition of measurement
• sight of observer
• angle of viewing the ruler & object
• calibration of instrument
How we deal with this is the subject of these lectures!
Physicists only measure 5 fundamental quantities
Length:
Time:
Current:
Temp.:
Mass:
Dr Alston Misquitta
Distance travelled by light in some time interval
Number of periods of specific wavelength radiation
Force between two conductors
Triple point of water
Lump of metal in Paris!
Scientific Measurement - Lecture 1
15
Measurement Uncertainty
Measurement of our chair should reflect the uncertainty!
write it as: 45 ± 2 cm
means TRUE value lies in range 43 - 47 cm
UNLIKELY to be >51
UNLIKELY to be <39
On a graph we show this as an error bar:
Any measurement is WORTHLESS
unless you estimate it's uncertainty
This is the only way of comparing measurements
and deciding on compatibility
Real measurement published in 2011
http://arxiv.org/abs/1109.4897v1
and corrected in 2012.
There was a problem with their fibre
optic timing equipment.
https://en.wikipedia.org/wiki/Faster-than-light_neutrino_anomaly
Dr Alston Misquitta
Scientific Measurement - Lecture 1
16
Measurement Uncertainty
Dr Alston Misquitta
Scientific Measurement - Lecture 1
17
Data Analysis
Dr Alston Misquitta
x
0.41
-0.01
2.41
2.79
-0.48
0.82
1.85
3.10
0.93
2.97
1.40
-0.07
2.11
2.40
0.25
2.33
1.13
-0.09
0.92
2.37
2.86
1.03
1.28
2.02
0.79
3.06
0.72
y
0.17
0.00
5.81
7.77
0.23
0.68
3.42
9.59
0.86
8.83
1.97
0.01
4.44
5.75
0.06
5.44
1.29
0.01
0.84
5.63
8.17
1.06
1.63
4.06
0.63
9.38
0.51
In an expt. x was varied and y was measured.
Is there a relationship between them?
What is the relationship between the two data?
Scientific Measurement - Lecture 1
18
Data Analysis
x
0.41
-0.01
2.41
2.79
-0.48
0.82
1.85
3.10
0.93
2.97
1.40
-0.07
2.11
2.40
0.25
2.33
1.13
-0.09
0.92
2.37
2.86
1.03
1.28
2.02
0.79
3.06
0.72
Dr Alston Misquitta
y
0.17
0.00
5.81
7.77
0.23
0.68
3.42
9.59
0.86
8.83
1.97
0.01
4.44
5.75
0.06
5.44
1.29
0.01
0.84
5.63
8.17
1.06
1.63
4.06
0.63
9.38
0.51
Humans are visual animals - brains recognise visual patterns very well
Plotting graphs of data is a powerful technique in discovering patterns
Scientific Measurement - Lecture 1
19
http://www.tylervigen.com/spurious-correlations
As visual creatures, we may see correlations where there are none!
Dr Alston Misquitta
Scientific Measurement - Lecture 1
20
Presenting Data
What is the difference between these numbers?
3 x 102
314
314.159 265
314.159 26535 89793 23846 26433 83279 50288 41971 69399 37510
π x 100
All are representations of the same number.
Number of sig.figs implies precision of that number.
Only in rare cases will you know a number to more than 3-4 sig figs!
Do not use more than 3 sig figs unless you can justify it!!!
Quantum electrodynamics: gyromagnetic ratio of the electron: g
Theory
Experiment
: ½(gth - 2) = 1159652140(28) x 10-12
: ½(gexp- 2) = 1159652186.9(4.1) x 10-12
note: 1159652140(28) x 10-12 is same as (1159652140 ± 28) x 10-12
Dr Alston Misquitta
Scientific Measurement - Lecture 1
21
Summary
Only experiment can determine the truth
Measurement ALWAYS has uncertainty
Never quote a measurement without its uncertainty
Plotting data graphically is very useful
Never plot graphs without error bars
... ever!
Never quote more sig figs than necessary
Dr Alston Misquitta
Scientific Measurement - Lecture 1
22