Department of Physics & Astronomy Lab Manual

Undergraduate Labs
Averaging,ErrorsandUncertainty
TypesofError
Therearethreetypesoflimitationstomeasurements:
1) Instrumentallimitations
Anymeasuringdevicecanonlybeusedtomeasuretowithacertaindegreeoffineness.
Ourmeasurementsarenobetterthantheinstrumentsweusetomakethem.

2) Systematicerrorsandblunders
These are caused by a mistake which does not change during the measurement. For
example,iftheplatformbalanceyouusedtoweighsomethingwasnotcorrectlysetto
zero with no weight on the pan, all your subsequent measurements of mass would be
toolarge.Systematicerrorsdonotenterintotheuncertainty.Theyareeitheridentified
andeliminatedorlurkinthebackgroundproducingashiftfromthetruevalue.

3) Randomerrors
These arise from unnoticed variations in measurement technique, tiny changes in the
experimental environment, etc. Random variations affect precision. Truly random
effectsaverageoutiftheresultsofalargenumberoftrialsarecombined.

Precisionvs.Accuracy
A precise measurement is one where independent measurements of the same quantity
closelyclusteraboutasinglevaluethatmayormaynotbethecorrectvalue.
Anaccuratemeasurementisonewhereindependentmeasurementsclusteraboutthetrue
valueofthemeasuredquantity.
Systematicerrorsarenotrandomandthereforecannevercancelout.
Theyaffecttheaccuracybutnottheprecisionofameasurement.

A. Lowprecision,Lowaccuracy:
Theaverage(theX)isnotclosetothecenter
B. Lowprecision,Highaccuracy:
Theaverageisclosetothetruevalue
C. Highprecision,Lowaccuracy:
Theaverageisnotclosetothetruevalue
Department of Physics & Astronomy Lab Manual
Undergraduate Labs
WritingExperimentalNumbers
UncertaintyofMeasurements
Errors are quantified by associating an uncertainty with each measurement. For example, the
best estimate of a length I is 2.S9cm, but due to uncertainty, the length might be as small as
2.S7cmoraslargeas2.61cm.Icanbeexpressedwithitsuncertaintyintwodifferentways:
1. AbsoluteUncertainty
Expressedintheunitsofthemeasuredquantity:L = 2. 59 _ . 2cm

2. PercentageUncertainty
Expressedasapercentagewhichisindependentoftheunits
Above,sinceu.u22.S9 = 1%wewouldwriteL = 7. 7cm_ 1%

SignificantFigures
Experimentalnumbersmustbewritteninawayconsistentwiththeprecisiontowhichtheyare
known.Inthiscontextonespeaksofsignificantfiguresordigitsthathavephysicalmeaning.
1. Alldefinitedigitsandthefirstdoubtfuldigitareconsideredsignificant.

2. Leadingzerosarenotsignificantfigures.
Example: I = 2.S1cm has 3 significant figures. For I = u.u2S1m, the zeros serve to
movethedecimalpointtothecorrectposition.Leadingzerosarenotsignificantfigures.

3. Trailingzerosaresignificantfigures:theyindicatethenumbersprecision.

4. One significant figure should be used to report the uncertainty or occasionally two,
especiallyifthesecondfigureisafive.

RoundingNumbers
Tokeepthecorrectnumberofsignificantfigures,numbersmustberoundedoff.Thediscarded
digitiscalledtheremainder.Therearethreerulesforrounding:
Rule1:Iftheremainderislessthan5,dropthelastdigit.
Roundingtoonedecimalplace:S.S46 S.S

Rule2:Iftheremainderisgreaterthan5,increasethefinaldigitby1.
Roundingtoonedecimalplace:S.798 S.8

Rule3:Iftheremainderisexactly5thenroundthelastdigittotheclosestevennumber.
This is to prevent rounding bias. Remainders from 1 to 5 are rounded down half the
timeandremaindersfrom6to10areroundeduptheotherhalf.
Roundingtoonedecimalplace:S.SS S.6,alsoS.6S S.6
Department of Physics & Astronomy Lab Manual
Undergraduate Labs
Examples
TheperiodofapendulumisgivenbyI = 2nlg.
Here,l = u.24misthependulumlengthandg = 9.81ms
2
istheaccelerationduetogravity.
WRONG:I = u.98S2692SS922s
RIGHT: I = u.98s
Yourcalculatormayreportthefirstnumber,butthereisnowayyouknowItothatlevelof
precision.Whennouncertaintiesaregiven,reportyourvaluewiththesamenumberof
significantfiguresasthevaluewiththesmallestnumberofsignificantfigures.

ThemassofanobjectwasfoundtobeS.S6gwithanuncertaintyofu.uS2g.
WRONG:m = S.S6 _ u.uS2g
RIGHT:m = S.S6 _ u.uSg
Thefirstwayiswrongbecausetheuncertaintyshouldbereportedwithonesignificantfigure

Thelengthofanobjectwasfoundtobe2.S9Scmwithanuncertaintyofu.uScm.
WRONG:I = 2.S9S _ u.uScm
RIGHT:I = 2.S9 _ u.uScm
Thefirstwayiswrongbecauseitisimpossibleforthethirddecimalpointtobemeaningful.

Thevelocitywasfoundtobe2. 4Smswithanuncertaintyofu.6ms.
WRONG:: = 2.S _ u.6ms
RIGHT:: = 2.4 _u.6ms
Thefirstwayiswrongbecausethefirstdiscardeddigitisa5.Inthiscase,thefinaldigitis
roundedtotheclosestevennumber(i.e.4)

Thedistancewasfoundtobe4S6uumwithanuncertaintyaround1m
WRONG:J = 4S6uum
RIGHT:J = 4.S6uu 1u
4
m
Thefirstwayiswrongbecauseittellsusnothingabouttheuncertainty.Usingscientific
notationemphasizesthatweknowthedistancetowithin1m.

Department of Physics & Astronomy Lab Manual
Undergraduate Labs
StatisticalAnalysisofSmallDataSets
Repeatedmeasurementsallowyoutonotonlyobtainabetterideaoftheactualvalue,butalso
enable you to characterize the uncertainty of your measurement. Below are a number of
quantities that are very useful in data analysis. The value obtained from a particular
measurementisx.ThemeasurementisrepeatedNtimes.OftentimesinlabNissmall,usually
nomorethan5to10.Inthiscaseweusetheformulaebelow:

Mean(x
avg
) Theaverageofallvaluesofx(thebestvalueofx)
x
avg
=
x
1
+ x
2
++x
N
N

Range(R)
The spread of the data set. This is the difference
betweenthemaximumandminimumvaluesofx
R = x
max
- x
min

Uncertaintyina
measurement
(x)
Uncertainty in a single measurement of x. You
determine this uncertainty by making multiple
measurements. You know from your data that x lies
somewherebetweenx
max
andx
min
.
x =
R
2
=
x
max
- x
min
2

Uncertainty
intheMean
(x
avg
)
Uncertainty in the mean value of x. The actual value
of x will be somewhere in a neighborhood around
x
avg
. This neighborhood of values is the uncertainty
inthemean.
x
avg
=
x
N
=
R
2N

MeasuredValue
(x
m
)
The final reported value of a measurement of x
contains both the average value and the uncertainty
inthemean.
x
m
= x
avg
_ x
avg

The average value becomes more and more precise as the number of measurements N
increases.Althoughtheuncertaintyofanysinglemeasurementisalwaysx,theuncertaintyin
themeanx
avg
becomessmaller(byafactorof1N)asmoremeasurementsaremade.

Department of Physics & Astronomy Lab Manual
Undergraduate Labs
Example
Youmeasurethelengthofanobjectfivetimes.
Youperformthesemeasurementstwiceandobtainthetwodatasetsbelow.

Measurement DataSet1(cm DataSet2(cm)

x
1
72 80
x
2
77 81
x
3
8 81
x
4
85 81
x
5
88 82

Quantity DataSet1(cm) DataSet2(cm)

x
avg
81 81
R 16 2
x 8 1
x
avg
4 0.4

ForDataSet1,tofindthebestvalue,youcalculatethemean(i.e.averagevalue):
x
avg
=
72cm+ 77cm+ 82cm+86cm+ 88cm
S
= 81cm

Therange,uncertaintyanduncertaintyinthemeanforDataSet1arethen:
R = 88cm- 72cm=16cm
x =
R
2
= 8cm
x
avg
=
R
2S
= 4cm

DataSet2yieldsthesameaveragebuthasamuchsmallerrange.

Wereportthemeasuredlengthsx
m
as:
DataSet1:x
m
= 81 _ 4cm
DataSet2:x
m
= 81. _ . 4cm

NoticethatforDataSet2,x
avg
issosmallwehadtoaddanothersignificantfiguretox
m
.
Department of Physics & Astronomy Lab Manual
Undergraduate Labs
StatisticalAnalysisofLargeDataSets
If only random errors affect a measurement, it
canbeshownmathematicallythatinthelimitof
an infinite number of measurements (N ),
the distribution of values follows a normal
distribution(i.e.thebellcurveontheright).This
distribution has a peak at the mean value x
avg

andawidthgivenbythestandarddeviationo.
Obviously, we never take an infinite number of
measurements. However, for a large number of
measurements, say, N~1u - 1u
2
or more,
measurements may be approximately normally
distributed. In that event we use the formulae
below:

Mean(x
avg
)
Theaverageofallvaluesofx(thebestvalueof
x).Thisisthesameasforsmalldatasets.
x
avg
=
x

N
=1
N

Uncertaintyina
measurement
(x)
Uncertainty in a single measurement of x. The
vast majority of your data lies in the range
x
avg
_ o
x = o =
_
(x

- x
avg
)
2 N
=1
N

Uncertainty
intheMean
(x
avg
)
Uncertainty in the mean value of x. The actual
value of x will be somewhere in a neighborhood
around x
avg
. This neighborhood of values is the
uncertaintyinthemean.
x
avg
=
o
N

MeasuredValue
(x
m
)
The final reported value of a measurement of x
contains both the average value and the
uncertaintyinthemean.
x
m
= x
avg
_ x
avg

Most of the time we will be using the formulae for small data sets. However, occasionally we
perform experiments with enough data to compute a meaningful standard deviation. In those
cases we can take advantage of software that has programmed algorithms for computing x
avg

ando.
x
avg

x
F
i
e
q
u
e
n
c
y

o
Department of Physics & Astronomy Lab Manual
Undergraduate Labs
PropagationofUncertainties
Oftentimes we combine multiple values, each of which has an uncertainty, into a single
equation.Infact,wedothiseverytimewemeasuresomethingwitharuler.Take,forexample,
measuring the distance from a grasshoppers front legs to his hind legs. For rulers, we will
assumethattheuncertaintyinallmeasurementsisonehalfofthesmallestspacing.

The measured distance is J

m
= J _ J where J = 4.6Scm- 1.ucm = S.6Scm. What is the
uncertainty in J
m
? You might think that it is the sum of the uncertainties in x and y (i.e.
J = x + y = u.1cm). However, statistics tells us that if the uncertainties are independent
of one another, the uncertainty in a sum or difference of two numbers is obtained by
quadrature: J = (x)
2
+ (y)
2
= u.u7cm. The way these uncertainties combine depends
onhowthemeasuredquantityisrelatedtoeachvalue.Rulesforhowuncertaintiespropagate
aregivenbelow.

Addition/Subtraction z = x _ y
z = (x)
2
+ (y)
2

Multiplication z = xy
z = |xy|
_
_
x
x
]
2
+ _
y
y
]
2

Division
z =
x
y

z = _
x
y
_
_
_
x
x
]
2
+ _
y
y
]
2

Power z = x
n
z = |n|x
n-1
x
Multiplication
byaConstant
z = cx z = |c|x
Function z = (x, y)
z =
_
_
o
ox
]
2
(x)
2
+_
o
oy
]
2
(y)
2

y = 4.6Scm_u.uScm x = 1.ucm_ u.uScm

Department of Physics & Astronomy Lab Manual
Undergraduate Labs
Examples

Addition
Thesidesofafencearemeasuredwithatapemeasuretobe124.2cm,222.5cm,151.1cmand
164.2cm. Each measurement has an uncertainty of 0.07cm. Calculate the measured perimeter
P
m
includingitsuncertainty.
P = 124.2cm+222.Scm+ 1S1.1cm+ 164.2cm = 662.ucm
P = (u.u7cm)
2
+ (u.u7cm)
2
+ (u.u7cm)
2
+ (u.u7cm)
2
= u.14cm

P
m
= 662.u _ u.1cm

Multiplication
Thesidesofarectanglearemeasuredtobe15.3cmand9.6cm.Eachlengthhasanuncertainty
of0.07cm.CalculatethemeasuredareaoftherectangleA
m
includingitsuncertainty.
A = 1S.Scm 9.6cm = 146.88cm
2

A = 1S.Scm 9.6cm
_
_
u.u7
1S.S
]
2
+ _
u.u7
9.6
]
2
= 1.Scm
2

A
m
= 147 _ 1cm
2

Power/MultiplicationbyConstant
Aballdropsfromrestfromanunknownheightb.Thetimeittakesfortheballtohittheground
is measured to be t = 1.S _ u.2s. The height is related to this time by the equationb =
1
2
gt
2

where g = 9.81ms
2
. Assume that the value for g carries no uncertainty and calculate the
heightbincludingitsuncertainty.
b =
1
2
(9.8ms
2
)(1.Ss)
2
= 8.281m
b =
1
2
(9.8ms
2
)(2 1.Ss u.2s) = 2.Sm
b
m
= 8 _ Sm