Propagating uncertainties through calculations

∆ ( 𝑎± 𝑏 ) =∆ 𝑎+ ∆ 𝑏

If data are to be added or subtracted, add the absolute uncertainty:

Propagating uncertainties through calculations
If data are to be multiplied or divided, add the fractional or percentage
uncertainty: 𝑎 · 𝑏
∆ 𝑦 ∆𝑎 ∆𝑏 ∆𝑐
= + +
𝑦 =
𝑐

𝑦 𝑎 𝑏 𝑐

𝐵=( 0 . 7 ± 0 . 3 )
EX: A cylinder has a radius of 1.60 ± 0.01 cm and a height of 11.5 ± 0.1 cm.
Find the volume.

V = π r2 h = π (1.60) 2 x 11.5 = 92.488 cm2 = 92 cm2

Absolute uncertainty in V:
0.02120 x 92.488 cm2 = 1.96075 cm2
V = 92 ± 2 cm2
The power (P) dissipated in a resistor of resistance R carrying a current I is equal to I2R. The value of I
has an uncertainty of ±2% and the value of R has an uncertainty of ±10%.
The value of the uncertainty in the calculated power dissipation is
A. ±8%. B. ±12%. C. ±14%. D. ±20%.

P= I2R
2% + 2% + 10% = 14%.

A student measures the length of a line with a wooden meter stick to be

11 mm  1 mm. What is the percentage error or uncertainty in her measurement?

Fractional error = 1 / 11 = 0.09.

Percentage error = (1 / 11) ·100% = 9%

▪ Thus 1 mm is the absolute error/uncertainty.

▪ 1 mm is also the precision.
A 9.51  0.15 meter rope ladder is hung from a roof that is
12.56  0.07 meters above the ground.
How far is the bottom of the ladder from the ground?

▪ y = a – b = 12.56 - 9.51 = 3.05 m

▪∆y = ∆a + ∆b = 0.15 + 0.07 = 0.22 m
▪Thus the bottom is 3.05  0.22 m from the ground.

A car travels 64.7  0.5 meters in 8.65  0.05 sec.

What is its speed?

▪ v= d/t = 64.7 / 8.65 = 7.48 m s-1

▪∆v/v = ∆d/d + ∆t/t = 0.5/64.7 + 0.05/8.65
▪∆v/v = 0.0135
▪ ∆v/7.48 = 0.0135
▪ ∆v = 7.48( 0.0135 ) = 0.10 m s-1.
▪ Thus, the car is traveling at 7.48  0.10 m s-1.
 ▪ ∆F / F = 0.2 / 10 = 0.02 = 2%.
▪ ∆m / m = 0.1 / 2 = 0.05 = 5%.
▪ ∆a / a = ∆F / F + ∆m / m = 2% + 5% = 7%.

▪ ∆r / r = 0.5 / 10 = 0.05 = 5%.

▪ A = r2.
▪ ∆A/A = ∆r/r + ∆r/r = 5% + 5% = 10%.
LINEARIZATION

The area where students struggle is linearizing data.

Linearizing a graph means to adjust the variables so that a curved graph turns into a straight-line
graph. This does not mean to fit the curved data points with a straight line. Rather, it means to
modify one of the variables in some manner such that when the data are graphed using this new
data set, the resulting data points will appear to lie in a straight line.

The significance of “linearizing” data

A linear best fit to the data can give information from the slope and the intercept open to the
physical interpretation for the students who are not in calculus.
LINEARIZATION

EX: The time period T of oscillation of a mass m suspended from a vertical spring is given by
the expression

𝑇 =2 𝜋
√ 𝑚
𝑘

Where k is a constant. Which one of the following will give rise to a straight-line graph?
A. against m B. against C. T against m D. against m
LINEARIZATION

EX: A particle is moving in a circular path of radius r. The time taken for one complete
revolution is T. The acceleration a of the particle is given by expression:

2
4𝜋
𝑎= 2
𝑇

A. against T B. against C. against D. against

EX: The data from an experiment of the spectral lines of the hydrogen atom
is given on the table below. Derive an equation in terms of the energy and the
quantum numbers for the hydrogen atom.
EX: The data from an experiment of the spectral lines of the hydrogen atom is given on the table below. Derive an equation in terms of
the energy and the quantum numbers for the hydrogen atom.

Quantum what's? I don't know what the heck they are talking about!
It doesn't matter, you may have no clue what they are talking about, however, this
problem can be done without knowing any information about atomic physics, simply
analyze the data using a graph.
EX: The data from an experiment of the spectral lines of the hydrogen atom is given on the table below. Derive an equation in terms of
the energy and the quantum numbers for the hydrogen atom.

The graph suggests that the energy varies inversely with the quantum number,
therefore, we can try to linearize the graph using...
E vs. 1/n.
But I know that this will not lead to linear graph. So after few unsuccessful attempts
( )
I realized that the graph of E vs.1/n2 should yield a straight line
EX: The data from an experiment of the spectral lines of the hydrogen atom is given on the table below. Derive an equation in terms of the
energy and the quantum numbers for the hydrogen atom.

We can now use the graph to find the slope:

13 .6 − 0 . 8 13 .6
𝑠𝑙𝑜𝑝𝑒= =13 .6 𝑒𝑉 𝐸= 2 𝑒𝑉
1− 0 . 06 𝑛

We have just derived the equation for

the energy of any electron state in the H atom!
DRAWING A GRAPH

In many cases, the best way to present and analyze data is to make a
graph. A wonderful tool of communication. Graphs also let you display
uncertainties nicely.
 DRAWING A GRAPH

When making graphs:

1. The independent variable is on the x-axis and the dependent variable is on the y-
axis.

2. Every graph should have a title that this concise but descriptive,
in the form ‘Graph of (dependent variable) vs. (independent variable)’.

3. The scales of the axes should suit the data ranges.

4. The axes should be labeled with the variable, units, and instrument uncertainties.

5. The data points should be clear.

 DRAWING A GRAPH

When making graphs:

7. Error bars represent the uncertainty range.

Error bars should be shown correctly (using a straight-edge).

8. Data points (average value ONLY!) should not be connected dot-to-dot fashion.
A line of best fit should be drawn instead. The best fit line is not necessarily the
straight line and should pass through all of the crosses created by the error bars.
Approximately the same number of data points should be above your line as below it.

9. Each point that does not fit with the best fit line should be identified.
If you have an outlier, leave it as it is, discuss it, and the draw
a second graph omitting the outlier and discuss it again.

10. Think about whether the origin should be included in your graph (what
is the physical significance of that point?) Do not assume that the line
should pass through the origin.
In principle, the size of the error bar could well be different for every single
point and so they should be individually worked out.
In practice, it would often take too much time to add much time to add all the
correct error bars, so some (or all) of the following short cut could be
considered.
▪ Rather than working out error bars for each point – use the
worst value and assume that all of the other errors bars are the same.
In the report include explanation, a sentence like: “Taking the highest uncertainty
we are reasonably sure that result is ….”
You should discuss the experiment results at the end of the lab report.
What type of dependence you discovered,….
For example when graphing experimental data, you can see immediately if
you are dealing with random or systematic errors (if you can compare with
theoretical or expected results).
 WHAT ARE MAX AND MIN LINES and what to do with them?
Remember that most Physics experiments lead to
graphical analysis of data. We can use graphs to
express and find uncertainties. You must include
graphs of your data.
Let’s say you have found the line of best fit and the
slope of this line (linearization)
 WHAT ARE MAX AND MIN LINES and what to do with them?
Therefore, you have found a value of the slope that
corresponds to some physical quantity (13.6 eV in
previous example which is maximum energy electron
can have in H atom: n = 1). Now you must use the
maximum and minimum best-fit lines to determine
the final uncertainty in the stated value of the slope of
your best-fit line.
Here’s how:
 WHAT ARE MAX AND MIN LINES and what to do with them?

1. Draw a straight line with the least slope possible

(minimum best-fit line) that connects corners of your
first and last error boxes.
2. Draw a straight line with the greatest slope possible
(maximum best-fit line) that connects corners of your
first and last error boxes.
3. Determine the slopes and y-intercepts of these two lines.
4. Best fit line has following uncertainties:
( 𝑚𝑚 𝑎𝑥 −𝑚 𝑚 𝑖𝑛 )
Uncertainty in the slope: ∆ 𝑚=
2
Uncertainty in the y-intercept: ( 𝑏𝑚 𝑎𝑥 −𝑏𝑚 𝑖𝑛 )
∆ 𝑏=
2
Visual Help

Lines of Best Fit and Max/Min Lines for ‘Graph of Quantity a vs. Quantity b’

Note that by using this technique, you may get max and min lines that do not go
through the error boxes of every data point. This is ok and you will not be penalized for
it.
·Look for the line to lie within all horizontal and vertical error bars.
·Only graph B satisfies this requirement.
IB has a requirement that when you conduct an experiment of your own design, you must
have five variations in your independent variable.
And for each variation of your independent variable you must conduct five trials to gather
the values of the dependent variable.
The five values for each dependent variable will then be averaged.
。

Here is an example of perfect student’s work.

▪ This is a well designed table containing all of the information and data points required by IB:
The “good” header

Table: Distance between magnet and compass needed to change direction of compass needle vs. time
the magnet spent in the freezer (measurements)
▪ This is a well designed table containing all of the information and data points required by IB:
Table: Distance between magnet and compass needed to change direction of compass needle vs. time
the magnet spent in the freezer (measurements)
Time in Distance between magnet and compass to change direction of compass needle
freezer (±.05 cm)
(± 0.2* sec)
Trial 1 Trial 2 Trial 3 Trial 4 Trial 5

0 14.7 14.9 14.6 14.3 14.3

10 14.9 14.7 14.5 15.0 14.8
20 14.9 15.1 14.8 14.5 15.0
30 15.2 14.9 15.1 15.0 15.2
40 15.6 14.8 14.7 15.3 15.3
*although digital stopwatch went to the hundredth’s decimal place, the uncertainty is still much greater
because there was no way to ensure that the time the magnet was in the freezer was exact. More
reasonable uncertainty would be 2 (uncertainty in opening and closing the door and taking out the
magnet) .
▪ This is a well designed table containing all of the information and data points required by IB:
Table: Distance between magnet and compass needed to change direction of compass needle vs. time
the magnet spent in the freezer (measurements)
Time in Distance between magnet and compass to change direction of compass needle
freezer (±.05 cm)
(± 0.2* sec)
Trial 1 Trial 2 Trial 3 Trial 4 Trial 5

0 14.7 14.9 14.6 14.3 14.3

10 14.9 14.7 14.5 15.0 14.8
20 14.9 15.1 14.8 14.5 15.0
30 15.2 14.9 15.1 15.0 15.2
40 15.6 14.8 14.7 15.3 15.3
Each measurement of distance has instrument uncertainty ½ smallest increment on meter stick
(±.05 cm).
▪ This is a well designed table containing all of the information and data points required by IB:
Table: Distance between magnet and compass needed to change direction of compass needle vs. time
the magnet spent in the freezer (measurements)
Time in Distance between magnet and compass to change direction of compass needle
freezer (±.05 cm)
(± 0.2* sec)
Trial 1 Trial 2 Trial 3 Trial 4 Trial 5

0 14.7 14.9 14.6 14.3 14.3

10 14.9 14.7 14.5 15.0 14.8
20 14.9 15.1 14.8 14.5 15.0
30 15.2 14.9 15.1 15.0 15.2
40 15.6 14.8 14.7 15.3 15.3
*although digital stopwatch went to the hundredth’s decimal place, the uncertainty is still much greater
because there was no way to ensure that the time the magnet was in the freezer was exact. More
reasonable uncertainty would be 2 (uncertainty in opening and closing the door and taking out the
magnet) .

Each measurement of distance has instrument uncertainty ½

smallest increment on meter stick (±.05 cm).
▪ This is a well designed table containing all of the information and data points required by IB:

Table: Distance between magnet and compass needed to change direction of compass needle vs. time
the magnet spent in the freezer (measurements)

Time in Distance between magnet and compass to change direction of compass needle
freezer (±.05 cm)
(± 0.2* sec) Dependant variable 5 Trials
Trial 1 Trial 2 Trial 3 Trial 4 Trial 5
( manipulated )
Independent

0 14.7 14.9 14.6 14.3 14.3

10 14.9 14.7 14.5 15.0 14.8
20 14.9 15.1 14.8 14.5 15.0
*although digital30
stopwatch
to the 15.2
went 14.9
hundredth’s 15.1
decimal place,15.0
is still much greater the 15.2
uncertainty
40 15.6 14.8 14.7 15.3 15.3
because there was no way to ensure that the time the magnet was in the freezer was exact. More
reasonable uncertainty would be 2 (uncertainty in opening and closing the door and taking out the
magnet) .
Each measurement of distance has instrument uncertainty ½
smallest increment on meter stick (±.05 cm).
Table: Distance between magnet and compass needed to change direction of compass
needle vs. time the magnet spent in the freezer (processed data)

Time in freezer (± Average Distance Percentage

Distance (cm)
0.2* seconds) (cm) error

0 14.56 14.6 ±0.3 2%

10 14.78 14.8±0.3 2%
20 14.86 14.9±0.3 2%
30 15.08 15.1±0.2 1%
40 15.30 15.3±0.3 2%
Uncertainty has
Measurement uncertainty in distance was calculated as range/2 1 sig. fig.

The value of best estimate must be expressed to the same

precision as the uncertainty
Graph: Temperature and Magnetic Field Strength

15.4

Distance to switch compass needle

15.2 f(x) = 0.017 x + 14.6
R² = 0.98972602739726
15

direction (cm)
14.8

14.6

14.4

14.2
0 5 10 15 20 25 30 35 40 45
Time (s)
The size of the vertical error bars in the graph is 0.3 (maximum measurement
uncertainty) up and down at each point in the graph. The size of horizontal error
bars are instrument uncertainty: 0.2 sec
Uncertainty of slope/gradient and y-intercept
▪To determine the uncertainty in the gradient and intercepts of a best fit line
we look only at the first and last error bars, as illustrated here:
15.4
Distance to switch compass (39.8, 15.6)
15.2
needle direction (cm)
x
m ma
15 m best
(-0.2, 14.9)
14.8
mmin (40.2, 15.0)
14.6
(0.2, 14.3)
14.4

14.2
0 5 10 15 20 25 30 35 40 45
Time (s)

Minimum gradient line: (-0.2, 14.9) and (40.2, 15.0) 𝑦 𝑚𝑖𝑛=0.00247525 𝑥+14 .9005
Maximum gradient line: (0.2, 14.3) and (39.8, 15.6) 𝑦 𝑚𝑎𝑥 =0.0328283 𝑥+14.2934
Uncertainty of slope/gradient and y-intercept
Graph: Temperature and Magnetic Field Strength
15.4

Distance to switch compass

15.2
𝑦 𝑚𝑖𝑛=0.00247525 𝑥+14 .9005

needle direction (cm)

15
14.8 𝑦 𝑚𝑎𝑥 =0.0328283 𝑥+14.2934
14.6
14.4
14.2
0 5 10 15 20 25 30 35 40 45
Time (s)
( 𝑚𝑚𝑎𝑥 −𝑚 𝑚𝑖𝑛 )
Uncertainty in the slope: ∆ 𝑚=
2
Uncertainty in the y-intercept: ( 𝑏𝑚𝑎𝑥 − 𝑏𝑚𝑖𝑛 )
∆ 𝑏=
2

y = 0.017x + 14.6 3