Measures of Dispersion
–
Formulas
A measure of
dispersion is designed to state numerically the extent to which individual
observations vary around the average. There are several measures of dispersion
under the two broad categories as follows:
A.
Absolute
Measures:
1.
Range,
2. Quartile Deviation
(Also known as 'Semi-Interquartile Range')
3.
Mean
Deviation About Mean,
4.
Mean
Deviation About Median, and
5.
Standard
Deviation.
B.
Relative
Measures:
1.
Coefficient
of Range
2.
Coefficient
of Quartile Deviation,
3.
Coefficient
of M.D. About Mean,
4.
Coefficient
of M.D. About Median, and
5.
Coefficient
of Variation.
Note: Mean
Deviation (M.D.) is usually calculated about arithmetic mean, and hence if it’s
only ‘Mean Deviation’, it refers to M.D. About Mean only.
Formulas
1.
Range
=
Largest Value (L) – Smallest Value (S)
2.
Coefficient
of Range
=
[(L – S) ÷ (L + S)] × 100
3.
Quartile
Deviation = ½ (Q3 – Q1)
4. Coefficient
of Quartile Deviation (1st Formula)
=
[(Q3 – Q1) ÷ (Q3 + Q1)] × 100
5. Coefficient
of Quartile Deviation (2nd Formula)
=
[Quartile Deviation ÷ Median) × 100
6. Mean
Deviation About Mean (For Simple Distribution)
=
[∑Mod (xi – Mean)] ÷ n
7.
Mean
Deviation About Mean (For Frequency Distribution)
=
[∑fi {Mod (xi – Mean)}] ÷ ∑fi
8. Mean
Deviation About Median (For Simple Distribution)
=
[∑Mod (xi – Median)] ÷ n
9. Mean
Deviation About Median (For Frequency Distribution)
=
[∑fi {Mod (xi – Median)}] ÷ ∑fi
10.
Coefficient
of M.D. About Mean
=
[M.D. About Mean ÷ Mean] × 100
11.
Coefficient
of M.D. About Median
=
[M.D. About Median ÷ Median] × 100
Standard Deviation
12.
For
Simple Distribution (1st Formula)
= √
[∑ {(xi – Mean) ^2} ÷ n]
13.
For
Simple Distribution (2nd Formula)
= √
[{∑ (xi^2) ÷ n} – {(∑xi ÷ n) ^2}]
14. For
Simple Distribution (3rd Formula – Short-Cut Method)
= √
[{∑ (di^2) ÷ n} – {(∑di ÷ n) ^2}]
[Here,
di = xi – A; A = Assumed Mean (as near as possible to the
true mean)]
15. For
Simple Distribution (4th Formula – Step-Deviation Method)
= √ [{∑ (Di^2) ÷ n} – {(∑Di ÷ n) ^2}] × c
[Here,
Di = (xi – A)/c; A = Assumed Mean (as near as possible to
the true mean); c = Common factor]
16.
For
Frequency Distribution (1st Formula)
= √
[∑ {fi (xi – Mean) ^2} ÷ ∑fi]
17.
For
Frequency Distribution (2nd Formula)
= √
[{∑fi (xi^2) ÷ ∑fi} – {(∑fixi
÷ ∑fi) ^2}]
18.
For
Frequency Distribution (3rd Formula – Short-Cut Method)
= √
[{∑fi (di^2) ÷ ∑fi} – {(∑fidi
÷ ∑fi) ^2}]
[Here,
di = xi – A; A = Assumed Mean (value of that observation
or that mid-value which has the highest frequency)]
19.
For
Frequency Distribution (4th Formula – Step-Deviation Method)
= √ [{∑fi (Di^2) ÷ ∑fi} – {(∑fiDi ÷ ∑fi) ^2}] × c
[Here,
Di = (xi – A)/c; A = Assumed Mean (value of that
observation or that mid-value which has the highest frequency); c = Common
factor or common width]
20. Coefficient of Variation (i.e. Coefficient of SD)
= (SD ÷ AM) × 100
21. Variance
= SD^2
Important note:
Formula of SD
for simple frequency distribution and grouped frequency distribution under all
the above four methods are same. Only thing to be remembered is that in case of grouped frequency distribution xi
will be the mid-values of the class intervals.
Important Properties of Standard Deviation
1.
If
y = x ± c where 'c' is a constant,
SD
of y = SD of x
i.e.
σy = σx
2.
If
x = c + dy where 'c' and 'd' are constants,
σx
= (Mod 'd') × σy
3.
√
[1/n∑ (xi – Mean of x) ^2] ≤ √ [1/n∑ (xi – A) ^2] whatever be the value of A.
SD of Composite Group
Composite SD
= √ [(N1σ1^2 + N2σ2^2
+ N1d1^2 + N2d2^2) ÷ (N1
+ N2)]
[Here,
xi
= Observations of Group 1
yi
= Observations of Group 2
d1
= Mean of Group 1 – Composite Mean
d2
= Mean of Group 2 – Composite Mean
σ1
= SD of Group 1
σ2
= SD of Group 2
Composite mean
= [(∑fx) + (∑fy)]/ (N1 + N2)
= [(N1*Mean of
x) + (N2*Mean of y)]/ (N1 + N2)
n1 = Total
number of observations of Group- 1
n2 = Total
number of observations of Group- 2
N1 = Total
frequency of Group- 1
N2 = Total
frequency of Group- 2]
No comments:
Post a Comment