PHAS0051 Data Analysis Problem Sheet 2022/23
Hello, dear friend, you can consult us at any time if you have any questions, add WeChat: daixieit
PHAS0051 Data Analysis Problem Sheet 2022/23
1. Mean, Variance, Standard Deviation, Standard Error on the Mean and Confidence Limits [8 Marks]
A student wishes to measure the density ρ, of an unknown metal by measuring both the weight and volume increase when immersing a sample in a graded water container on scales. The following values were measured for the mass m, and volume V.
m (g) |
7.28 |
7.29 |
7.29 |
7.30 |
7.28 |
7.31 |
7.30 |
7.27 |
7.29 |
7.27 |
7.26 |
7.30 |
V (cm3) |
3.40 |
3.41 |
3.43 |
3.39 |
3.42 |
3.45 |
3.44 |
3.42 |
3.42 |
3.43 |
3.43 |
3.46 |
ρ(g/cm3 |
|
|
|
|
|
|
|
|
|
|
|
|
i. Calculate the individual values of ρ and then the mean values for m, V and ρ from the values in the table. [2]
iii. Calculate the mean value of ρ from the mean values of m and V. [1]
iv. Calculate the uncertainty on the mean value found in (iii) by propagation of the uncertainties on the mean values of m and V. [1]
v. Calculate the 95% confidence level limits on ρ using the uncertainty on the mean estimated in (iv.) and assuming that the measurements are Gaussian. [1]
2. Linear fit and X2 [10 Marks]
An experiment is performed to test whether the current through a semiconductor device I, depends linearly on the potential drop across it, V. The following data are obtained.
V/volts |
0.20 |
0.40 |
0.60 |
0.80 |
1.00 |
1.20 |
1.40 |
1.60 |
1.80 |
2.00 |
2.20 |
I/mA |
4.09 |
5.05 |
6.11 |
7.15 |
8.50 |
9.78 |
11.22 |
12.77 |
14.40 |
15.87 |
17.52 |
The current measurements are estimated to have the same uncertainty (dI = ±0.25mA), whilst the potential drop (V) is estimated to be without significant uncertainty.
i. Using an appropriate program, least squares fit the data to a straight line. Make sure you fully label the graph. [3]
ii. Calculate the X2 probability for the fit. [3]
iii. On the same graph, plot the data with uncertainty bars and the best fit line and comment on the outcome of the experiment in light of the X 2 probability calculated in part ii and the graphical comparison of the best fit line and the data. [4]
3. Poisson Distribution and X2 [12 Marks]
The following are results from an experiment in which a Geiger counter is used to record the number of particles emitted per second by a weak radioactive source. In 481 repeated observations for a 1 second interval, counts are recorded with the following frequency:
Counts |
0 |
1 |
2 |
3 |
4 |
5 |
6 |
7 |
8 |
9 |
10 |
11 |
12 |
13 |
≥14 |
Frequency |
0 |
2 |
10 |
41 |
50 |
82 |
76 |
66 |
60 |
38 |
25 |
13 |
9 |
6 |
3 |
i. Calculate the mean number of counts per second (μ), the variance and the standard deviation. [3]
ii. Using the value for the mean found in (I) calculate the theoretically expected Poisson frequencies for each number of counts. [3]
iii. Calculate the χ2 probability P(χ2) that the frequency distribution follows the expected Poisson distribution with the mean you have determined. You can approximate the error on each frequency N, by a Gaussian with standard deviation √N (See Note 2 & 3 at the end of this question). [4]
iv. Plot the experimental data and theoretical Poisson distributions. [2]
The Poisson distribution is given by:
P(r) =
Here r is the measured number of counts/s- 1 and μ is the mean number of counts/s-1 .
Note 2: the approximation that the distribution of each Poisson frequency is a Gaussian with standard deviation √N is a poor one for small N (less than 5 - see Hughes and Hase, p111 Ch8.6). In order to make the approximation reasonable when comparing the above distribution with theory, the contents of adjacent bins may be summed to give sufficiently large frequencies.
Note 3: the probability P(χ2) is such that 1 − P(χ2) is the probability that the value of χ2 is as low, or lower than the calculated value of χ2 . So a value of P(χ2) very close to unity means an improbably good fit – perhaps too many parameters have been used in the fit, or the uncertainties are underestimated.
4. Best fit by Minimising X2 [10 Marks]
An experiment is carried out where the voltage, VC, across the capacitance in a resonant series LCR circuit is measured as the frequency f, (ω is the angular frequency) is varied through resonance. The following estimates of the voltage versus frequency are obtained.
f (kHz) |
Vc (volts) |
f (kHz) |
Vc (volts) |
f (kHz) |
Vc (volts) |
15 |
16.00 |
25 |
36.53 |
35 |
28.46 |
16 |
17.05 |
26 |
43.58 |
36 |
24.11 |
17 |
17.60 |
27 |
51.90 |
37 |
20.91 |
18 |
18.77 |
28 |
61.00 |
38 |
18.42 |
19 |
20.01 |
29 |
69.45 |
39 |
16.26 |
20 |
21.75 |
30 |
70.51 |
40 |
14.53 |
21 |
23.48 |
31 |
62.28 |
41 |
13.23 |
22 |
25.42 |
32 |
50.86 |
42 |
11.95 |
23 |
28.60 |
33 |
41.29 |
43 |
11.09 |
24 |
32.21 |
34 |
33.74 |
44 |
9.89 |
The uncertainty in the voltage is +/- 0.15V, the frequencies are without significant uncertainty. According to AC theory the voltage is predicted to vary as:
E0
Vc = 1
C(业2R2 + L2 (业2 − 业0(2))2)2
Here E0 is the amplitude of the AC voltage applied across the circuit which is measured to be 12 volts and L and C are the inductance and capacitance in the circuit having the values 30mH and 950pf. The inductance and capacitance are high precision components and the uncertainties on their values are negligible as is the value of the uncertainty on E0 . ω0 is the natural frequency of the freely oscillating circuit with zero damping such that;
R = 0, 业0 =
R in the circuit is only nominally known since it is made up of a resistance box set to 800Ω and an additional smaller contribution due to hysteresis losses in the core of the inductor.
i. By minimising the X2 fit of the data to the theoretical form, estimate the best value for the total resistance in the circuit and the additional resistance due to hysteresis. [4]
ii. What is the X2 probability for the best fit and does this value support the AC theory given above? [3]
iii. Plot a fully labelled graph and comment on any agreement / disagreement between the data and the fitted curve. [3]
2022-10-27