Lab4_ErrorAnalysisII (1).

nb 1

ERROR ANALYSIS II
Lab Report
Names:
Section:
Date:

Purpose
To understand the nature of the Gaussian and Poisson distribution functions and their relationship to error
analysis.

Introduction
Often, repeated measurements of a physical quantity will not each give the same value. Rather, there will be a
distribution of measured values about some central value. Here we will consider two of the important types of
distributions.

First, there may be a single true value for the quantity being measured, but random measuring errors due to
the apparatus or techniques will cause the result to vary. In principle, these random errors can be reduced by
improving the apparatus or techniques, such that the width of the distribution of measured values will
decrease. The measured values in this case usually form a continuous distribution but in this lab we will work
with discrete values for the measured variable.

The second type of distribution arises when the measurement process has inherent statistical fluctuations
that cannot be reduced by improving techniques. For example, in the decay of a radioactive sample, we can
only give the probability that each nucleus will decay within a given time interval. Even if the counter works
perfectly, the number counted for the same time interval will vary. The measurements for this second case
have discrete values, e.g. counts of decays.

For the first case the distribution of measured values about the mean is frequently described by a bell shaped
curve called the Normal or Gaussian distribution which is characterized by two parameters -- the mean which
gives the location of the peak of the distribution and the standard deviation which gives the width of the
distribution.
For the second case of statistical fluctuations such as nuclear decay, the distribution about the mean is given
by the Poisson distribution which is characterized by one parameter – the mean. If the mean number decaying
per counting time interval is large compared to the size of the fluctuations, the Poisson distribution reduces to
a Gaussian distribution. Interestingly, even for a mean as small as 10 the Poisson distribution is very well
approximated by a Gaussian distribution.
Lab4_ErrorAnalysisII (1).nb 2

Readings
Books:
John R. Taylor, “An Introduction to Data Analysis”, University Science Books. Chapters 10 and 11, and Chapter
3.

You can read about Cs-137 and how it is used for radioactive dating in these Wiki pages,
Caesium-137, dating Jefferson wine bottles.
We will use Cs-137 sample in today’s lab.

Theory & Concepts in Short

A. Statistical Definitions
Your textbook gives a thorough discussion of distribution functions. Here we will only give a brief summary of
some important points. The mean and variance of a parent distribution in random variable x are defined by,

From a limited set of measurements {xi } (n is not infinite) we can estimate μ from the average value of {xi }.
Likewise, the variance is estimated from the “sample” of measurements and also uses the average value of the
xi .
1 2 1
μ ≈ μ = x =
N
∑Ni=1 xi and σ 2 ≈ σ = σ x 2 = N-1
∑Ni=1 ( xi - x )2 .

The factor (N-1) gives a better approximation of the variance than n because we are using only an estimate of
μ rather than its actual value. The square root of the variance is called the standard deviation.

For a Gaussian distribution there is 68% chance that a single measurement will fall between μ−σ and μ+σ.
There is a 95.4% chance that it will fall within 2σ of μ and a 99.7% chance that it will fall within 3σ of μ .

Next we would like to know how close our estimate of μ is to the true value, i.e. how close μ = x is to μ.
This is not given by σ x , which is characteristic of the parent distribution (x), but rather by,
σx = σx / N
where N is the number of measurements of the sample; this is characteristic of the distribution of x. Though
the distribution of x can be anything, the distribution of x almost always becomes a Gaussian, when N gets
large (central limit theorem).
Above relation shows that as more measurements are made the uncertainty of the estimate of the mean
decreases, but only as the square root of the number of measurements.

B. Distribution Functions
A distribution function P (x) describes the chance of a single measurement of a
certain quantity (with true value μ) that will give a specific value of x. Specifically,
Lab4_ErrorAnalysisII (1).nb 3

when x is a continuous variable, P (x) dx is the probability of the measured value lying in between x and x +

dx. The area under the curve of P (x), i.e. P (x) dx, equals 1. This simply means

that there is 100 × % probability of getting all possible values of x in a measurement.

Gaussian Distribution
The most commonly encountered distribution is the Normal or Gaussian distribution,
1 -
(x-μ) 2

PG (x) =  2σ 2

2π σ
This curve is peaked at x = μ and is symmetric about μ.
For a set of measurements x is a good estimate of μ. The normal distribution
PG (x) is encountered very often. This is due to the central limit theorem. Roughly,
the theorem states that if the uncertainty (or random error) in the measured
value of x are caused by adding several (K) independent factors,
each with its own (not necessarily Normal) distribution,
then in the large K limit the overall distribution of x approaches a normal distribution PG (x).

Poisson Distribution
A second common distribution function for integer variables is the Poisson distribution,
-μ μk
PP (k) =
k!
where k is a non - negative integer variable and μ is the expected value of k or the rate.
This distribution is a good model to understand the randomness
of a phenomena that presumably happens with a constant rate μ in nature.
Note that μ is not necessarily an integer, i.e. it can be a fractional number.
PP (k) is an asymmetric, peaked function whose maximum value does not coincide with μ.
For a set of measurements of k, the average k is again a good estimate of μ.

Notice that while Gaussian PG (x) appears to involve two parameters μ and σ,
Poisson PP (k) involves only μ. Both the peak
position and width of Poisson PP (k ) are solely determined by μ.
The standard deviation of Poisson PP (k) is σ = μ .
For large μ, Poisson PP (k) becomes almost symmetric peak centered at its
mean and can be quite accurately approximated by Gaussian PP (k) with σ = μ;
we will use this property to study the Gaussian distribution.

Data and report

In your lab report, write up each section separately in a manner complete enough so that your results can be
followed by someone who does not have a copy of this lab handout. Give a brief description of the purpose of
the section, describe the procedure, list the results in tables, describe the analysis, and give conclusions.

Significant figures
Be sure to read section on significant figures and quote all results accordingly. In general, the uncertainty is
quoted to only one significant figure unless that is a 1, in which case it is sometimes quoted to two (i.e., 0.3,
Lab4_ErrorAnalysisII (1).nb 4

not 0.33, but 0.1 or 0.13 would be acceptable). The value should be quoted to the same precision as the error,
or at most one more. For example, give 2.5 ± 0.4 not 2.543 ± 0.4 and not 2.5 ± 0.007, although 2.54 ± 0.1 or 2.54
± 0.14 would be acceptable.

Procedure
Apparatus
Radioactive source (5 mC 137 Cs)
Geiger tube
Electronic counter
LabPro Interface
Computer

Experiment
In this experiment you will measure the activity of a radioactive 137 Cs source. The half life of the 137 Cs (30.1
yr.) is long compared to the length of the experiment so that its activity can be considered constant. You will
use a computer to acquire data on the sample’s activity and then compare the data to the theoretical Gaus-
sian and Poisson distributions in order to understand the role that distribution functions play in your data.

The nuclear decay is detected by a Geiger tube. Each time a gamma or beta ray from a decaying nucleus
enters the Geiger tube a current pulse is generated. The tube is connected to a counter circuit which converts
it into a 5 V, 1 ms wide voltage pulse which is available at the phone jack on the back of the counter. The
counter also counts the pulses and displays the count on the front panel. However, you will not use this
feature. Instead the voltage pulse from the counter is fed into a LabPro interface. The LabPro interface trans-
mits the decay data to the computer through its input port. Using the program LoggerPro you will repeatedly
record the number of counts in a time interval (whose length you have chosen) and make a plot of the distribu-
tion of counts about the average value. You will then export the data to Mathematica to compare to the
theoretical distribution functions and plot your results.

The parent isotope Cs-137 with a half-life of 30.1 years beta decays (94.6%) to the metastable state of
Ba-137m. This further decays by gamma emission (662 keV) with a half-life of 2.6 min. to the stable Ba-137
element. A diagram of this decay process is shown below.
Lab4_ErrorAnalysisII (1).nb 5

The Geiger tube should be connected to the BNC connector on the back of the counter, and the DIG/SONIC1
input of the LabPro interface to the jack on the back of the counter.The LabPro interface is connected to the
USB port of the computer.

Make sure the counter is turned on. (The LabPro interface is always on.)

On the menu go to Experiment > Set Up Sensors > Show All Interfaces (or click on the Vernier icon )

and set the DIG/SONIC1 to radiation.

Choose suitable values for Collection LENGTH and count interval (SAMPLING RATE). Click button COLLECT to
collect data.
To save your data, choose FILE : EXPORT DATA AS .csv ... (not SAVE).

You will now do a series of “runs” each of which measures the decay counts N in a set of n consecutive inter-
vals making up the total collection time of the run. The decay rates R = N/Δt (Δt is the time interval of one
measurement of N).

A. Background 50 sec measurement.

Remove the radioactive source from the sample holder.
Collect data for 50 sec at 20 Hz, 1000 = 50x20 data points, each collected for 1/20=0.05 sec
The count rate in a given interval is the number of counts divided by the count interval.
Export data to a file BG_50.csv

B. γ-decay 50 sec measurement.

Place the radioactive source looking down on the top shelf in the sample holder under the Geiger tube.
Record the count (decay) rate of the source penetrating through the Al plate for 50 sec at 20 Hz.
Export data to a file down_50.csv
Lab4_ErrorAnalysisII (1).nb 6

C. γ+β-decay 50 sec measurement.

Flip the radioactive source looking up, keep it on the same shelf from above in the sample holder under
the Geiger tube.
Record the count (decay) rate of the source for 50 sec at 20 Hz.
Export data to a file up_50.csv

D. γ+β-decay 10 sec measurement.

Without touching the radioactive source repeat step C but collect data for 10 sec at 20 Hz, 200 = 10x20
data points.
Export data to a file up_10_20Hz.csv

E. γ+β-decay 10 sec measurement at 5 Hz.

Without touching the radioactive source repeat step D but collect data for 10 sec at 5 Hz, 50 = 10x5 data
points.
Export data to a file up_10_5Hz.csv

Dialog:
Step1.
Import the *.csv data into Mathematica file.
Use the command Import[“Insert/filepath/filename”] from a file or {InputString[]} from a dialog box.
(See the example below.)

Import imports data from a *.csv file into Mathematica file.

An example of the code is given below.
- Part[expr, All, {2}] gives the second column of the data.
- Rest removes the title line.
- Flatten flattens out external nested lists, result is a vector of counts.

bg50 =
Flatten[Rest[Part[Import["/Users/Girsh/Documents/275_f14/Lab4/BG_50.csv"], All, {2}]]];
down50 =
Flatten[Rest[Part[Import["/Users/Girsh/Documents/275_f14/Lab4/down_50.csv"], All, {2}]]];

...

Import all your data files and assign logical names.

Lab4_ErrorAnalysisII (1).nb 7

Step2.
- For all the cases calculate # of measurements, each measurement time, total counts, mean, Sqrt[mean], σ,
uncertainty, decay rate with uncertainty.
- Present results in organized tables with proper labels.
- Plot # of counts distribution and the probability distribution for each case.
(See the example below.)
- Put clear explanations for each data set, explain all the table columns, like how Rate and its error is
calculated, and report the table columns with correct number of significant figures if needed.

Example:
var = bg50;
time = 0.05; (* measurement time, in sec.*)

(* Length[var] is total # of measurements *)

(* Total[var] is Total # of Counts *)
(* Mean[var] is (Total # of Counts)/(# of measurements) *)
(* StandardDeviation[var] is σ for total # of measurements *)
(* Sqrt[Total[var]]/Length[var] is Uncertainty *)
(* Rate is Mean/time *)
(* Uncertainty for Rate is (Uncertainty for counts)/time *)
(* Tally counts the number of occurences getting # counts *)
(* Table[{N[Tally[var]〚i,2〛/Length[var]]}, {i,Length[Tally[var]]}]
is to calculate the probability *)
(* PoissonDistribution[Mean[var]] calculates  -μ μx /x! *)

Grid{Text["γ-decay, 50 sec"], SpanFromLeft}, "# of measurements", "Time, sec",

"Total Counts", "Mean, counts", " Mean ", "σ", "Rate, #/sec", {Length[var], time,
Total[var], PlusMinus[N[Mean[var]], NumberForm[N[Sqrt[Total[var]] / Length[var]], 2]],
N[Sqrt[Mean[var]]], N[StandardDeviation[var]], PlusMinus[N[Mean[var] / time],
NumberForm[Sqrt[Total[var]] / Length[var] / time, 2]]}, Frame  All
(* Grid[Transpose[
Prepend[Sort[Table[{Tally[var]〚i,1〛,Tally[var]〚i,2〛,N[Tally[var]〚i,2〛/Length[var]]},
{i,Length[Tally[var]]}]], {"# of counts",
"Number of occurences getting # counts","Probability Density"}]], Frame  All] *)
ListPlot[Tally[var], PlotStyle  {Red}, PlotMarkers  {"◆", 18}, Filling  Axis,
Frame  True, FrameLabel  {"ν →" , "Number of ν counts sec occurences"}, PlotRange  All]
Show[ListPlot[
Table[{Tally[var]〚i, 1〛, N[Tally[var]〚i, 2〛 / Length[var]]}, {i, Length[Tally[var]]}],
PlotStyle  {Red}, PlotMarkers  {"◆", 18}, Filling  Axis, PlotRange  All],
DiscretePlot[PDF[PoissonDistribution[Mean[var]], k], {k, 0, 40},
ExtentSize  {Scaled[.25], Scaled[.25]}, PlotRange  All],
Frame  True, FrameLabel  {"ν →" , "Probability Density"}]

Step3.
- When does the histogram resemble a Gaussian curve centered on μ?
- On these histograms that can be approximated by a Gaussian distribution, overlay a Gaussian curve fit
with proper parameters μ and σ on top of the histogram and Poisson distribution function.
Lab4_ErrorAnalysisII (1).nb 8

If needed, check ‘Lab 1. Error Analysis I’ to see how to define and plot a Gaussian distribution. Include all the
plotting functions inside one Show[...] function to overlay them on top of each other.
- Draw a vertical line on the histograms at the position of the mean. Show the range μ ± σ range on the
histograms.
- Explain the relation of the counts observed in the experiment to the Poisson distribution. Also compare
square roots of the means and standard deviations.
- Explain why the Poisson distribution can model the random behavior of the Cs-137 radioactive decay.

Step4.
- Collect and present all the calculations in one cumulative table with proper labels. Use Grid function. Write
one row for each data set bg50, down50, .... To be able to show the numbers in correct format and with
correct number of significant figures, use the functions, PlusMinus[a, b] which is a ± b, or Interval[a, b] which
is the interval [a, b], and also check the functions ScientificForm, EngineeringForm, or NumberForm to write
numbers with correct number of significant figures.
- Explain the difference between decay rates and their uncertainties in different runs. Explain statistical
differences between results of the runs B and C; also between runs C, D, and E.
- Use error propagation rules for the errors and calculate the γ-decay rate and the β-decay rate of the
radioactive source [do background subtraction]. Compare to the decay diagram above. Explain the
discrepancy.

Rutgers 275 Classical Physics Lab

“ERROR ANALYSIS II”
Contributed by Maryam Taherinejad and Girsh Blumberg ©2014