k 1 2 3 4 5 6 7 8 9 10
2k 2 4 8 16 32 64 128 256 512 1,024
Sturge's formula
57-61 2
52-56 6
47-51 7
42-46 13
37-41 5
32-36 7
27-31 5
2 to k rule
Essentially we would look to construct k
classes for our frequency distribution, when
the value of 2k first exceeds the number of
observations in our sample. So, if we had a
sample with 45 observations, we would first
consider constructing 6 classes, because 26
= 64, the first power of 2 with a value larger
than the sample size of 45.
A guide, not a dictator.
Strictly speaking the 2k rule is a guide, not a
rule. If the 2k rule suggests you need 6
classes, also consider using 5 or 7 classes
... but certainly not 3 or 9.
The following scores were the results on a 60-item test in a certain
class in statistics. Construct a grouped frequency distribution for the
data using 7 classes.
33 41 28 54 59 44 49 46 53
35 39 27 57 44 43 50 45 56
34 43 31 51 43 42 49 46 52
33 36 30 52 39 44 45 47 51
34 44 32 53 31 38 42 37 48
Class interval or class width
H L
w
k
H : the highest value, L: the smallest value
Class interval can also be estimated
based on # of observations
H L
w
1 3.322ln n
Select the lower limit of the first class and
set the limits of each class
It could be L or any value smaller than L.
It should be an even multiple of the class
interval.
There should be between 5 and 20 classes.
The classes must be continuous.
The classes must be exhaustive.
The classes must be mutually exclusive.
The classes must be equal in width.
Relative frequency
• o Relative frequency of a class is the
frequency of that class divided by to total
number of frequency.
f
RF
n
Example
The following scores were the results on a
60-item test in a certain class in statistics.
Construct a grouped frequency distribution for
the data using 7 classes.
33 41 28 54 59 44 49 46 53
35 39 27 57 44 43 50 45 56
34 43 31 51 43 42 49 46 52
33 36 30 52 39 44 45 47 51
34 44 32 53 31 38 42 37 48
Relative Class Marks Class Cumulative Cumulative
Class limits Frequency frequency or Midpoint boundaries frequency frequency
(C) (f) (RF) % (Mpt) (CB) ≤ ≥
Good Luck!