Something about doing experiment

1. Introduction to experimentation
2. Characteristics of experimental data
3. Graphical presentation of data
4. Dealing with uncertainties
 Introduction to experimentation
• Stages of a typical
experiment
The aim → The plan →
Preparation → Preliminary
experiment → Collecting
data → Repeatability →
Analysis of data → What do
the data tell you? →
Reporting the experiment
• Keeping a record of your
experiment (notebook)
Date → Title → Aim of the
experiment → Description
of the apparatus → Sketch
of apparatus →
Experimental method →
Measurements → Graphs
→ Calculations →
Conclusion
 Characteristics of experimental data
• Units of measurement
The SI system of units
Multiples and subdivisions of units
• Tabulation of data

Table I: Times for an object to fall 25m

Time of fall (s) 2.2 2.0 2.6 1.9 2.1 2.4 2.2 2.3 2.3 2.0

Table II: Times for an object to fall 25m

Time of fall /s 2.2 2.0 2.6 1.9 2.1 2.4 2.2 2.3 2.3 2.0

Table III: Measured values of the inductance of a coil of wire

Inductance (x10-3H) 9.5 9.6 9.3 9.7 9.7 9.9 9.2 9.5 9.6
• Uncertainties in measurements
Table I: Variation of electrical resistance with
temperature of a copper wire

Temperature (°C) ± 0.5 °C Electrical resistance of length of

copper wire (Ω) ± 0.001 Ω
8.0 0.208

16.5 0.213

23.5 0.222

32.0 0.229

40.5 0.232

54.5 0.243
• Significant figures
– An experiment value is recorded as 6.12, this
implies that the actual value lies between 6.11
and 6.13.
– If it is written as 6.124, then this is implies that
actual value lies between 6.123 and 6.125.
– 6.12 has three significant figures; 6.124 has
four significant figures.
– Based on the number of significant figures
presented, the readers will know the size of
uncertainty of the value.
How many significant figures appear in the following number?

The number Significant figures

1.654 4
0.00437 3 (zeros here are not regarded as
significant figures)
64000 2 (the zeros are not significant, unless
being told explicitly)
1.20 3 (if the zero is not significant then the
number should be written as 1.2)
0.10000738 8
3.500x103 4 (use scientific notation to avoid confusion
while changing units)
• Calculations and significant figures
– 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.
3.7 x 3.01 = 11.137, but the final result is 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.
11.24 + 13.1 = 24.34, but the final result is 24.3
• Rounding numbers: it is a good advice not to round numbers until all the
calculations have been completed.
1.3563342 → two significant figures 1.4 (≥5, round-up); → four
significant figures 1.356 (<5, left alone).
• Examples of numbers expressed in scientific notation:
Number In scientific notation
12.65 1.265x101
0.00023 2.3x10-4
342.5 3.425x102
34001 3.4001x104

• Orders of magnitude: we normally have a feel for the size of numerical values
that should emerge in an experiment through previous experiences or common
senses. For example, if we were to measure the speed of a car moving along a
street and found it to be 4x106m/s, we should suspect something to be wrong. In
many situations we can assess how sensible the numbers are at least to within a
factor of 10 of the ‘actual’ value, i.e. we speak of knowing the value to within an
order of magnitude. To estimate the expected measured value to with an order
of magnitude is very helpful for avoiding gross mistakes.
 Graphical presentation of data

• A most effective way of presenting the data in

order to reveal relationships and anomalies is to
plot a graph.
• A graph can indicate:
– The range of measurements made
– The uncertainty in each measurement
– The existence or absence of a trend in the data
gathered.
– Which data points do not follow the general
trend exhibited by the majority of data.
• x-y graphs (Cartesian coordinate graphs) are
used extensively to present experimental data.
A typical x-y graph
Each point in the graph has a coordinate (x, y)
Independent variable: the quantity is controlled or deliberately varied during an
experiment. It is plotted as the x-coordinate.
Dependent variable: the quantity varies in response to the changes in the
independent variable. It is plotted as the y-coordinate.
We say formally that the y is a function of x.
A useful experimental graph includes: a title indicating the relationship being
investigated and axes that are clearly labeled with the name of the quantities
under study and their units of measurement.
Sometimes, the
scale on the y-
axis is
proportional to
the magnitude of
the dependent
quantity, but the
absolute value of
the quantity is not
given. In such a
case, the y-axis
is labeled with
arbitrary units.
Temperature Time (s) This experiment is performed in which the time
(°C)
taken for a small metal sphere to fall a fixed
21 62 distance through a liquid is recorded as the
26 48 temperature of the liquid increases.
30 35
37 26
42 22 This graph contains four mistakes.
46 19 Can you find them?
51 17
Use large enough symbols to indicate data points in a graph.
Symbols for cyclists A and B are easy to see, but it is too small for C.
The origin of the graph
does not have to be
included on the graph.
Choose a suitable
display ranges or
scales for x-axis and y-
axis can show more
clearly the relationship
between y and x.
Uncertainty bars (error bars): we draw vertical and horizontal lines extending from
a data point to present uncertainty. The length of the line is drawn according to
the size of the uncertainty in the quantity.

The uncertainty in the x-axis The uncertainty in the y-axis

time (s) ± 5s Temperature (°C) ± 4°C
 A linear x - y graph

Points do not lie exactly along a straight line. This is likely due to the
experimental uncertainty. It is reasonable to conclude that there is a linear
relationship between magnetic field and current.

Two important constants can be deduced from a straight line through the
points. These are the gradient and the intercept of the line.
The line of ‘best fit’ through a set of data points: the data points on the
graph appear to be scattered evenly above and below the line.

Line 1 is a good effort at the best straight line, while line 2 is less
satisfactory than line 1.
 How to draw the best fit line through a set
of x – y data ‘by eye’

Step 1. Position a clear plastic rule along the plotted

data points.
Step 2. Move the rule until the points are scattered as
equally as possible above and below the line.
Step 3. The origin is not a special point so do not force
the line through it.
Step 4. Using the rule, draw a fine line through the
points with a sharp pencil.
 The gradient and intercept of the best straight line

y = mx + c
m – gradient; c – intercept
y1 = mx1 + c
y2 = mx2 + c
rise
m = (y2 – y1)/ (x2 – x1)
m = rise / run
c = mx - y run
When finding the gradient:
1. Only choose points that lie on the best line. This normally means that
data points cannot be used for determination of the gradient.
2. Choose points on the line that are well separated as this reduces the
effect of any inaccuracy arising from measuring the run and rise from
the graph
The units must be included when calculating the
gradient and intercept of a line on a graph.

m=
(25 − 4 )mm
= 2.1mmN −1
(11 − 1)N
 Estimate uncertainties in gradient and intercept
1. Draw the best straight line; 2. Draw the steepest line, which passes through
the top of the error bars for the data points at the extreme right, and the bottom
of the error bars for the points at the extreme left. 3. Draw the flattest line, which
passes through the bottom of the error bars for the data points on the extreme
right, and the top of the error bars for the data on the extreme left.

The intercept = (0.5 ± 0.5)mm

best 0.5mm 1.6mm h-1

upper 1.1mm
lower 0.1mm
1.9mm h-1
The gradient = (1.9±0.3)mm h-1
2.2mm h-1
Linearising
equations

Not a strainght in M

M
T = 2π
k
2π
T= M
k

y=mx +c
A straight line in M1/2
Logarithmic graphs
current in linear scale

I = AeBV
logI = BV + logA
y=mx+c

Log-log graphs current in log10 scale

I = Ad n
logI = logA + nlogd
y=nx+c
 Dealing with uncertainties
• Uncertainties (errors of measurement: no measurement
is perfectly accurate or exact. Many instrumental,
physical, and human limitations cause measurements to
deviate from the “true” values of the quantities being
measured. These deviations are called “experimental
uncertainties”, but more commonly “experimental error”.
• The “true value” of a measured quantity is a concept
induced for purposes of discussion. We can never hope
to measure true values.
• When we specify the “error” in a quantity or result, we
are giving an estimate of how much that measurement is
likely to deviate from the true value of the quantity.
• A measurement or experimental result is of little use if
nothing is known about the probable size of its error.
 Determinate and indeterminate errors

Experimental errors are of two types: (1) indeterminate (or random)

and determinate (or systematic) errors.
Indeterminate errors: indeterminate errors are present in all
experimental measurements.
¾ There is no way to determine the size or sign of the error in any
individual measurement.
¾ The errors cause that the repeated measurements of a quantity
give different results, usually with as many positive as negative
fluctuations.
¾ The errors can have many causes, including operator errors or
biases, fluctuating experimental conditions, varying environmental
conditions, and inherent variability of measuring instruments.
¾ The effect of indeterminate errors on the results can be somewhat
reduced by taking repeated measurements then calculating their
average, because errors of positive and negative sign tend to
compensate each other in the averaging process.
Determinate (or systematic) errors: systematic error has
the same size and sign for every measurement when it is
repeated many times.
¾ A common cause of systematic error is instrumental
or procedural bias and an experimental blunder. For
example: using a miscalibrated scale or instrument,
using an incorrect value of a constant in the
equations, using the wrong units, reading a scale
incorrectly.
¾ There is no sure method for discovering and
identifying them just by looking at the experimental
data.
¾ Their effects can not be reduced by averaging
repeated measurements.
 An Example

Indeterminate error may occur in aligning one end of the

pencil with the scale marking on the ruler.

Systematic error may occur if the ruler is wrongly marked

or calibrated.
 Indeterminate and systematic errors in target practice
(a) Because all shots arrived close to
one another, we can tell the
indeterminate errors are small.
Because the distribution of shots is
centered on the center of the
target, the systematic errors are
also small.
(b) The indeterminate errors are still
small, but the systematic errors are
much larger – the shots are
“systematically” off center toward
the right.
(c) Here the indeterminate errors are
large, but the systematic ones are
small – the shots are widely
scattered but not systematically off
center.
(d) Here, both indeterminate and
systematic errors are large.
The same experiment as
before but without
showing the position of
the target. This situation
closely corresponds to
the one in most real
experiments, in which we
do not know the true
value of the quantity
being measured. Here,
we can still assess the
indeterminate errors
easily but cannot tell
anything about the
systematic ones.
Precision and accuracy: a measurement with relatively small indeterminate
error is said to have high precision. A measurement with small indeterminate
error and small determinate error is said to have high accuracy. Precision
does not necessarily imply accuracy. A precise measurement may be
inaccurate if it has a determinate error.
Deviation: the difference between each measurement and the average (mean)
of the entire set of measurement is called the deviation of the measurement
from the mean. Use the word deviation when an individual measurement of a
set is being compared with a quantity which is representative of the entire set.
Deviations can be expressed as absolute amounts, or as percents.
Difference: the absolute or percent difference between the two independent
and equally reliable results. This implies no judgment about which is more
reliable.
Experimental discrepancy: when a measurement or result is compared with
another which is assumed or known to be more reliable, we call the difference
between the two the experimental discrepancy. Discrepancies may be
expressed as absolute discrepancies or as percent discrepancies. It is
customary to calculate the percent by dividing the discrepancy by the more
reliable quantity. However, if the discrepancy is only a few percent, it makes
no practical difference which of the two is in the denominator.
 Measures of error

Limits of error: error limits may be expressed in the form

Q ± ∆Q
where Q is the measured quantity and ∆Q is the magnitude of its limit of
error. This expresses the experimenter's judgment that the "true" value of
Q lies between Q - ∆Q and Q + ∆Q. This entire interval within which the
measurement lies is called the range of error.

Average deviation of the mean: for a set of n measurements Qi whose

mean value is <Q>, the average deviation of the mean (A.D.M.) is:
n n
∑ Qi − Q ∑ Qi
n −1
where Q = i =1
(n − 1)(n 2 ) n
This is an estimate of how far the mean value of Q is likely to deviate from
the "true" value of Q. As a rough rule of thumb, the probability of <Q>exp
being within three A.D.M. (on either side) of the true value is nearly 100%
(actually 98%). This is a useful relation for converting (or comparing)
A.D.M. to limits of error.
Standard Deviation of the mean: the standard deviation is a well known,
widely used, and statistically well-founded measure of error. For a set of n
measurements Qi whose mean value is <Q>, the standard deviation of the
mean is found from:
1 1
 n
2
2  n
2
2
 ∑ (Qi − Q )   ∑ Qi − n Q 
2
 i =1   
 n(n − 1)  =  i =1 
n(n − 1)
   
   

The use of the standard deviation is hardly justified unless the

experimenter has taken a large number of repeated measurements of
each experimentally determined quantity.
It can be shown that when the measurements are distributed according
the "normal" ("Gaussian") distribution, average deviations and standard
deviations are related by a simple formula:
[average deviation] = 0.80 [standard deviation]
 Standard methods for expressing error
Absolute Error: uncertainties may be expressed as absolute measures,
giving the size of the a quantity's uncertainty in the same units in the quantity
itself. For example, a piece of metal is weighed a number of times, and the
average value obtained is: M = 34.6 gm. By analysis of the scatter of the
measurements, the uncertainty is determined to be ∆M = 0.07 gm. This
absolute uncertainty may be included with the measurement in this manner:
M = 34.6 ± 0.07 gm
The value 0.07 after the ± sign in this example is the estimated absolute
error in the value 3.86.

Relative (or Fractional) Error: uncertainties may be expressed as relative

measures, giving the ratio of the quantity's uncertainty to the quantity itself.
In general:
absolute error in a measurement
relative error =
size of the measurement
For example, in the previous example, the uncertainty in M = 34.6 gm was
∆M = 0.07 gm. The relative uncertainty is therefore:
∆M/M=0.07gm/34.6gm=0.002 or 0.2%
 Combination of uncertainties
A simple method: obtain the best estimate value R from a formula, then each
quantity in the formula is modified by an amount equal to the uncertainty in the
quantity to produce the largest value Rmax and smallest value Rmin. The
uncertainty is calculated as ∆R = (Rmax – Rmin)/2

Example: in an electrical experiment, the current through a resistor was found

to (2.5 ± 0.1) mA and the voltage across the resistor (5.5 ± 0.3) V. Calculate
the resistance of the resistor using R=V/I and the uncertainty in R, ∆R.
Solution: R = V/I = 5.5V / 2.5x10-3A = 2.20x103Ω
The maximum Rmax will occur when the numerator V is increased by an
amount equal to the uncertainty and the denominator I is reduced by an
amount equal to the uncertainty, Rmax = 5.8V / 2.4x10-3A = 2.42x103Ω
The minimum Rmin is found by using the minimum value of V and the
maximum value of I, Rmin = 5.2V / 2.6x10-3A = 2.00x103Ω
So DR = (Rmax – Rmin)/2 = 0.210x103Ω
Rounding to the appropriate number of significant figures,
R = (2.2 ± 0.2)x103 Ω
• If independent measurements ( X ± ∆X ) and ( Y ± ∆Y ) are combined to
give a result R, the associated uncertainty ∆R will be given by :
– Sum: R = X + Y; ∆R = ∆X + ∆Y
– Difference: R = X – Y; ∆R = ∆X + ∆Y
– Product: R = X · Y; ∆R/R = ∆X/X + ∆Y/Y
– Quotient: R = X/Y; ∆R/R = ∆X/X + ∆Y/Y

Example: the temperature of (3.0 ± 0.2) x 102 g of water is raised by (5.5 ±

0.5)°C by a heating element placed in the water. Calculate the amount of
heat transferred to the water to cause this temperature rise. Also calculate
the uncertainty in the amount of heat transferred to the water.
Solution: the formula that relates heat input Q to temperature T is: Q = mcT
where m is the mass of the water and c is its specific heat capacity
(4186 J/kg°C )
Q = 0.3 kg x 4186 J/kg°C x 5.5 °C =6907J
∆Q = Q (∆m/m + ∆T/T) = cT∆m + cm∆T = 1088.4J
We can now quote the heat transferred as Q = (6.9 ± 1.1) x 103 J.
 A few useful terms

• Instrument: a device for determine the value or magnitude of a quantity or

variable.
• Accuracy: closeness with which an instrument reading approaches the true
value of the variable being measured.
• Precision: a measure of the reproducibility of the measurements, i.e., given
a fixed value of a variable, precision is a measure of the degree to which
successive measurements differ from one another.
• Sensitivity: the ratio of output signal or response of the instrument to a
change of input or measured variable.
• Resolution: the smallest change in measured value to which the instrument
will respond.
• Accepted values will be found in standard data texts, for example,
densities, conductivities etc. Copies of these texts are available in the
laboratory.
• Nominal values are found stamped on pieces of equipment, such as
standard resistors, capacitors, klystrons etc. In general these values will
always have a non-specified tolerance which may well be 10% or more.
• Discrepancy is often useful to describe the deviation of one measurement
of the same quantity from another, or of the measurement from the
prediction of a theory. The use of the term "discrepancy" in such cases
is less confusing than the term "error".