College of Science Health and Engineering

2016 PHY1SCA Laboratory

Experimental Methods

Uncertainties
Some uncertainty in a measured quantity always remains no matter how good the experimenter or
experimental design.

Estimates of the uncertainty in measurements should always accompany the measurement and
need to be recorded with your record of results.

Significant figures
When an experimental record is quoted as 6.12, by convention this implies that the true value lies
between 6.11 and 6.13.

Similarly, 6.124 implies 6.125 and 6.123.

Rules for Determining the Number of Significant Figures

• Count from the first non‐zero value to the last figure.

6.12 is to give the number to 3 significant figures.

6.125 is to give the number to 4 significant figures.
0.0010306 is to give the number to 5 significant figures.

• For whole numbers greater than 10 that end with one or more zeros, if it is not stated
explicitly, it is assumed that the number is known to the nearest 10.Trailing zeros are
insignificant and for example,
1670 has 3 significant figures.

Rounding Numbers
In a calculation involving a number of arithmetic steps, it is good advice not to round numbers until
the entire calculation has been completed.

Calculations and Significant Figures.

Consider the following calculation which determines the area of a circle.

A = πd2/4

Let the diameter equal to d = 8.9 mm

A = π(8.9 mm)2/4.0
= 62.21138852 mm2
(as determined by calculator)

How do you assign significant figures to the calculation?

Rules for Determining Number of Significant Figures:

College of Science Health and Engineering
2016 PHY1SCA Laboratory

• When multiplying or dividing numbers: identify the number in a calculation that is given to
the least number of significant figures. Give the result of the calculation to the same
number of significant figures.

eg. 3.7 x 3.01 = 11.137

3.7 has 2 significant figures

Answer : 11

• When adding and subtracting numbers, round the result of the calculation to the same
number of decimal places as the number in the calculation given to the least number of
decimal places.

Significant Figures and Scientific Notation

Often it is not clear how many figures in a quoted number are significant, with this often occurring
during units conversion.

For example, in converting the following time measurement from seconds to milli‐seconds,

365 s ‐> 365000 ms

Question : How do you convey the number of significant figures in the milli‐sec result?

Answer: Express the result in scientific notation.

Rules for Representing Numbers in Scientific Notation

• In scientific notation, the first non‐zero figure that appears is FOLLOWED by a decimal point.

• The number of significant figures is equal to the number of figures to the left of the
multiplication sign.

For example,
• 365 s (three significant figures) in scientific notation becomes:
3.65 x 102 s and
3.65 x 105 ms.
• 34001 which has 5 significant figures becomes 3.4001 x 104
• 0.00341 which has 3 significant figures becomes 3.41 x 10‐3.

Order of Magnitude.
In experimental work the habit of estimating the expected measured value to within an order of
magnitude is very helpful for avoiding gross errors.
College of Science Health and Engineering
2016 PHY1SCA Laboratory

Graphical Presentation of Data

• x,y plots are correctly referred to as called Cartesian Coordinate Plots.

• Independent Variable : That quantity which is controlled or deliberately varied during an

experiment is referred to as independent variable and is plotted as the x‐coordinate.

The independent variable is also known as the abscissa.

• Dependent Variable : The quantity that varies in response to changes in the independent
variable is referred to as the dependent variable and is plotted as the y‐coordinate .

The dependent variable is also referred to as the ordinate.

Titles, Labels and Units for Graphs.

The title of a graph indicates the relationship being investigated.

eg. Title : A versus B

where : A = dependent variable
B = independent variable

Labelling Axes
Clearly label each axis with the name of the quantity under study.
The unit of the measurement is given in brackets after the name of the quantity.

Scale of Graph.
In most situations it is sensible to choose scales so that the points that are plotted fill the available
graph paper.

Origin of graph
There is no rule to say that the origin of graphs must begin at (0.0, 0.0)

In some cases, a poor choice of origin may fail to reveal important trends in experimental data.

Error bars
It is important to indicate the size of the uncertainties in the x and y quantities by attaching error
bars to each data point.

Error bars are vertical and/or horizontal lines that extend from the data point.

The length of the error bar is a measure of the size of the uncertainty in the quantity.

In some experiments the uncertainties are not constant from one measurement to another and as a
consequence, the error bars vary from point to point on the graph.

If error bars are too small to plot clearly, then omit them.

When to plot a graph.

Possible choices:
(1) Plot results as the experiment proceeds
College of Science Health and Engineering
2016 PHY1SCA Laboratory

OR

(2) After the measurements have been completed.

Advantage of (1): Can be an efficient use of time with immediate indication of interesting or
unexpected features in the data.

Possible Disadvantage: A knowledge is required of the likely range of x and y otherwise scales cannot
be drawn in advance.

In these practicals, you will be required to present tabulated results so plot graphs after
measurements are taken.

Once data points are plotted they may appear to lie along a curve, a straight line or show no trends
at all.

The linear (or straight line) x‐y graph is of particular importance to the experimentalist.

Linear x‐y Graphs

Linear graphs have an important place in the analysis of experimental data for the following reasons.
• Two important quantities can be determined from a straight line plot ‐ gradient and
ordinate intersection.
• Departure from linearity can be easily observed
• Points far from the line can be identified and possibly rejected as being due to experimental
error.
• Can reliably predict values for the y‐quantities for a chosen value of x‐coordinate.

Line of Best fit – by hand

For plotted experimental data that we are sure a linear relationship exists between the x and y
quantities, it is very useful to be able to write down an equation that represents that relation.

ie y = mx + c. (1)

where m is the gradient and c the y axis intercept.

The first step in determining equation 1 and the values of m and c is to draw a line of best fit
through data points.

Step Action
(1) Position a clear plastic ruler along the plotted data points.
(2) Move the ruler until the points are scattered as equally as possible
above and below the line.
(3) The origin is not a special point so do not force the line through it.
(4) Using the ruler, draw a line through the points with a sharp pencil

When finding the gradient (using rise over run) you must :
• Only choose points that lie on the best line. This normally means that data points cannot be
used for determination of the gradient.
• Choose points on the line that are well separated as this lessens the effect of any inaccuracy
arising from measuring the run and rise from the graph.
College of Science Health and Engineering
2016 PHY1SCA Laboratory

In situations where the x axis does not begin at zero, the value of c cannot be obtained directly from
the graph. In such cases c can be calculated by first finding m and then rearranging the line
equation so that
c = y ‐ mx

When quoting the value of m and c DO NOT forget to include their corresponding units.

Uncertainties in Gradient and Intercept.

As there is an uncertainty associated with every experimental measurement, there must also be an
uncertainty in the gradient and intercept of the line drawn through the data.
If error bars have been attached to each point, we can use these to assist in estimating the
uncertainty.

In order to estimate the uncertainties we draw three lines through the data.

Line 1 : Line of best fit.

Line 2 and 3 : Lines drawn so that they give the maximum and
minimum gradient consistent with the error bars.

eg. Best fit gradient = 1.9 mm.h‐1

maximum gradient = 2.2 mm.h‐1
minimum gradient = 1.6 mm.h‐1

gradient = (1.9 ±0.3) mm.h‐1

A similar approach is used for the determining the uncertainty in the y intercept.

If error bars are not attached to the points or are insignificant, we estimate the uncertainty through
the scatter of points around the line of best fit. In this case we draw two lines.
Line 1 : Line of best fit (remember this is centred through the scatter of points)

Line 2 : Uncertainty line. This is a line above or below the line of best fit that
indicates the worst possible line that could conceivably represent the
results.

eg. Best fit gradient = 2.3 N.m‐1

Uncertainty gradient = 2.4 N.m‐1

gradient = (2.3 ±0.1) N.m‐1

7.10 Linearising of Equations

In many experiments we have prior knowledge about the relationship between the quantities being
investigated.

If this is the case, and if the relationship can be expressed as an equation, then we can often choose
what to plot on each axis to produce a straight line graph.
College of Science Health and Engineering
2016 PHY1SCA Laboratory

This is referred to as linearising the equation

For example consider the equation:

ଵ
‫ ܭ‬ൌ ݉‫ ݒ‬ଶ
ଶ

If we wish to determine the mass (m) of an object from its kinetic energy (K) and velocity (v), then
one approach is to plot the Kinetic Energy versus velocity. This will result in a parabola from which
there is no simple way of determining the object's mass.
If, however, we plot K verses ‫ ݒ‬ଶ it will produce a straight line with a gradient of m/2 (kg)

Graphical Interpolation and Extrapolation:

Interpolation: Determination of a value y at an x value that lies
within the range of experimental data points.

Extrapolation: When an x‐value lies outside the measurement range, the corresponding y‐value is
found by extrapolation

Calculations
How do I do calculations with values that have an error?
When you do calculations with values that have uncertainties, you will need to calculate the
uncertainty in the derived results. The simplest way to do this is to compute that largest
possible error the calculation could possibly have. Examples:

Addition: 3.118 ± 0.008 g + 3.131 ± 0.013 g = ?

Sum of values = 3.118 + 3.131 = 6.249 g
Largest possible error = 0.008 + 0.013 = 0.021 g
Result is 6.249 ± 0.021 g

Subtraction: 3.131 ± 0.013 g ‐ 3.118 ± 0.008 g = ?

Difference in values = 3.131 – 3.118 = 0.013 g
Largest possible error = 0.008 + 0.013 = 0.021 g
Result is 0.013 ± 0.021 g
Notice that zero is included in the range indicating that it is possible there is no difference in
the two results.

Multiplication: 3.131 ± 0.013 g x 6.1± 0.2 cm = ?

Product of values = 3.131 g x 6.1 cm = 19.1 g.cm
Maximum possible value = 3.144 x 6.3 = 19.8 g.cm
Maximum deviation = 19.8 g.cm – 19.1 g.cm = 0.7 g.cm
Result is 19.1 ± 0.7 g.cm

Division: 3.131 ± 0.013 g ÷ 3.118 ± 0.008 g = ?

Quotient of values = 3.131 ÷ 3.118 = 1.0042
Maximum possible quotient = 3.144 ÷ 3.110 = 1.0109
Maximum deviation = 1.0109 – 1.0042 = 0.0067
Result is 1.004 ± 0.007
College of Science Health and Engineering
2016 PHY1SCA Laboratory

Also for multiplication and division, the above result can be achieved by adding the
percentage errors in each component – and is often a faster method of achieving the result.
It is important to remember however that the final result of a calculation should be expressed
with the actual error, not the percentage error. In summary, errors in calculations involving
the four basic mathematical operations can be determined from:

ADDITION of values ADD actual errors

SUBTRACTION of values ADD actual errors
MULTIPLICATION of values ADD percentage errors
DIVISION of values ADD percentage errors

Calculations, where several operations are combined, may become confused. Setting out the
calculation in tabular form will assist considerably.

For example: V=(x1‐x2).y2

By measurement it is found that x1= 1.26 ± 0.02 m, x2= 0.32 ± 0.02 m, and y = 3.12 ± 0.01 m

Quantity Value Uncertainty % Uncertainty

x1 1.26 .02
x2 0.32 .02
x1 – x2 0.94 .04 4.26
y 3.12 .01 0.32
y2 9.73 0.64
V=(x1‐x2).y2 9.146 0.45 4.90

V = 9.1 ± 0.4 m3
College of Science Health and Engineering
2016 PHY1SCA Laboratory