Statistics for Students
Statistics for Students
1
AIM : FREE EDUCATION Measures
TO ofALLCentral Tendency and Dispersion
APUL
CHAPTER
Measures of Central
23 Tendency and Dispersion
Chapter Highlights
Measures of central tendency, Arithmetic mean, Geometric mean, Harmonic mean, Median, Quartiles, deciles
and percentiles, Mode, Symmetric distribution.
f x + f 2 x2 + ... + f n xn
∑ fi xi x1, x2, x3, ...., xn respectively, then the weighted average is
defined as:
i =1
x = 1 1 =
f1 + f 2 + .... + f n n w1 x1 + w2 x2 + ...... + wn xn
∑ fi Weighted A.M. =
w1 + w2 + ....... + wn
i =1
23.2 Chapter 23
AIM : FREE EDUCATION TO ALL APUL
Combined Mean
( )
1/ N
G.M. = x1f ⋅ x2f ....... xnf
1 2 n
,
If we are given the A.M. of two data sets and their sizes,
then the combined A.M. of two data sets can be obtained n
by the formula: where N= ∑ fi
i =1
n1 x1 + n2 x2
x12 = ⎛ n ⎞
⎜ ∑ fi log xi ⎟
n1 + n2
where, x12 = Combined mean of the two data sets 1 and 2 or G.M. = antilog ⎜
i =1 ⎟
x1 = Mean of the first data ⎜ N ⎟
⎜ ⎟
x2 = Mean of the second data ⎜⎝ ⎟⎠
n1 = Size of the first data
n2 = Size of the second data
I M P O R TA N T P O I N T S
Properties of A. M.
1. In a statistical data, the sum of the deviations of indi- In the case of continuous or grouped frequency distribution,
the values of the variate x are taken to be the values corre-
vidual values from A.M. is always zero, i.e.,
sponding to the mid-points of the class intervals.
n
∑ fi ( xi - x ) =0
Some Points About Geometric Mean
i =1
35 ∑ xi
1× x + 4 × x =
i =1
10 = 4 ; n
1+ 4
n
\ 50 = x + 35; \ x = 15. ⇒ ∑ xi = nx
i =1
\ Sixth number = 15.
23.5
AIM : FREE EDUCATION Measures
TO ofALLCentral Tendency and Dispersion
APUL
n n New Sum = 4900 – 110 + 210 = 5000
∑ ( xi + 2i) ∑ xi + 2 (1 + 2 + .... + n) \ New mean =
5000
= 50
i =1 i =1
\ = 100
n n
n ( n + 1) 14. The mean of n items is X . If the first item is increased
nx + 2 by 1, second by 2 and so on, then the new mean is
= 2 = x + (n + 1).
n n
(A) X + n (B) X +
11. The A.M. of n observations is M. If the sum of n – 4 2
n +1
observations is a, then the mean of remaining 4 obser- (C) X + (D) None of these
2
vations is
Solution: (C)
nM – a nM – a
(A) (B) Let x1, x2, ...xn be n items.
4 2
1
nM – a Then X = Sxi .
(C) (D) nM + a n
4
Solution: (A) Let y1 = x1 + 1, y2 = x2 + 2
Let the mean of the remaining 4 observations be X1 . y3 = x3 + 3, ..... , yn = xn + n
a + 4 X1 nM - a Then the mean of the new series is
Then, M = ⇒ X1 = .
( n - 4) + 4 4 1 1 1 1
Syi = S( xi + i ) = Sxi + (1 + 2 + 3 + ..... + n)
12. The weighted mean of first n natural numbers whose n n n n
weights are equal to the squares of corresponding 1 n( n + 1) n +1
numbers is = X+ ⋅ = X+ .
n 2 2
n +1 3n ( n + 1)
(A) (B) 15. The number of observations in a group is 40. If the
2 2 ( 2n + 1) average of first 10 is 4.5 and that of the remaining 30
( n + 1) ( 2n + 1) n ( n + 1) is 3.5, then the average of the whole group is
(C) (D)
6 2 15 1
(A) (B)
4 5
Solution: (B) (C) 8 (D) 4
1.12 + 2.22 + ...... + n.n2
Weighted Mean = Solution: (A)
12 + 22 + ....... + n2
x1 + x2 + ..... + x10
n ( n + 1) n ( n + 1) = 4.5
10
Sn3 2 2
= = ⇒ x1 + x2 + ..... + x10 = 45
Sn 2 n ( n + 1) ( 2n + 1)
6 x11 + x12 + ..... + x40
Also, = 3.5
3n ( n + 1) 30
= .
2 ( 2n + 1) ⇒ x11 + x12 + ..... + x40 = 105
13. Mean of 100 items is 49. It was discovered that three
items which should have been 60, 70, 80 were wrongly \ x1 + x2 + ..... + x40 = 150
read as 40, 20, 50 respectively. The correct mean is x1 + x2 + ..... + x40 150 15
1 \ = =
(A) 48 (B) 82 40 40 4
2
16. A person purchases one kg of tomatoes from each of
(C) 50 (D) 80
the 4 places at the rate of 1 kg, 2 kg, 3 kg, 4 kg per
Solution: (C) rupee respectively. On the average he has purchased
Sum of 100 items = 49 × 100 = 4900 x kg of tomatoes per rupee, then the value of x is
Sum of items added = 60 + 70 + 80 = 210 (A) 2 (B) 2.5
Sum of items replaced = 40 + 20 + 50 = 110 (C) 1.92 (D) None of these
23.6 Chapter 23
AIM : FREE EDUCATION TO ALL APUL
Solution: (C) 19. The mean of the values 0, 1, 2, 3, ... , n with the
Since we are given rate per rupee, harmonic mean will corresponding weights nC0, nC1,..., nCn respectively is
give the correct answer 2n 2n +1
4 4 × 12 48 (A) (B)
H.M. = = = ( n + 1) n( n + 1)
1 1 1 1 12 + 6 + 4 + 3 25
+ + + n +1 n
1 2 3 4 (C) (D)
= 1.92 kg per rupee. 2 2
Solution: (D)
17. The A.M. of 2nC0, 2nC2, 2nC4,..., 2nC2n is n
C1 + 2 ⋅ nC2 + 3 nC3 + ... + n nCn
2n 22n n
(A) (B) C0 + nC1 + nC2 + ... + nCn
( n + 1) ( n + 1)
n ⋅ 2n-1 n
22 n -1 2n -1 = n
=
(C) (D) 2 2
( n + 1) ( n + 1)
20. In a factory, workers work in three shifts say shift
Solution: (C) 1, shift 2 and shift 3 and they get wages in the ratio
(1 + x)2n = 2nC0 + 2nC1x + 2nC2x2 + 2nC3x3 +, ..., 4 : 5 : 6 depending on the shift 1, 2 and 3 respectively.
+ 2nC2nx2n Number of workers in the shifts are in the ratio 3 : 2
: 1. If total number of workers working is 1500 and
Put x = – 1 wages per worker in Ist shift is Rs 400. Then mean
2n
C0 – 2nC1 + 2nC2 – 2nC3 +, ..., + 2nC2n = 0 (1) wage of a worker is
(A) Rs. 467 (B) Rs. 500
Now put x = 1
(C) Rs. 600 (D) Rs. 400
2n
C0 + 2nC1 + 2nC2 + 2nC3 +, ..., + 2nC2n = 22n (2)
Solution: (A)
Adding (1) and (2), we get Workers in Ist shift = 750
2n
C0 + 2nC2 +, ..., + 2nC2n = 22n–1 Wages in Ist shift = Rs. 400
Workers is IInd shift = 500
22 n -1 Wages in IInd shift = Rs. 500
A.M. of 2nC0, 2nC2 +, ..., + 2nC2n =
( n + 1) Workers in IIIrd shift = 250
Wages in IIIrd shift = Rs. 600
18. The A.M. of 2n+1C0, 2n+1C1, 2n+1C2, ... , 2n+1Cn is 750 × 400 + 500 × 500 + 250 × 600
\ Mean =
2n 2n 1500
(A) (B)
n n +1 Rs. 467 per worker
22n 22n 21. If a variable takes values 1, 2, 3, ... , n with frequencies
(C) (D)
n ( n + 1) 12, 22 , ... , n2, then the mean is
Solution: (D) Sn3
(A) Sn (B)
2n+1
C0 + 2n+1C1 + 2n+1C2 +, ..., Sn 2
3
Sn
+ 2n+1C2n + 2n+1C2n+1 = 22n+1 (C) (D) None of these
Sn
2n+1 Solution: (B)
Now C0 = 2n+1C2n+1, 2n+1C1 = 2n+1C2n,...
2n+1 1.12 + 2 . 22 + ... + n . n2 Sn3
Cr = 2n+1C2n–r+1 =
12 + 22 + ... + n2 Sn 2
So sum of first (n + 1) terms = Sun of last (n + 1) terms
2n+1
22. Mean of n items is x. If these n items are increased by
or C0 + 2n+1C1 + 2n+1C2 +, ..., + 2n+1Cn = 22n 12, 22, 32, ..., n2 successively, then mean gets increased
2 n +1 2 n +1
C0 + C1 + 2 n +1C2 +, ..., + 2 n +1
Cn by
or ( n + 1) ( 2n + 1) n( n + 1) ( 2n + 1)
n +1 (A) (B)
6 6
22n n2
= (C) (D) Remains same
( n + 1) 2
23.7
AIM : FREE EDUCATION Measures
TO ofALLCentral Tendency and Dispersion
APUL
Solution: (A) 3. The median value is given by the formula
x1 + x2 + x3 + ... + ... + xn ⎛ ⎛ n⎞ ⎞
x =
n ⎜ ⎜⎝ 2 ⎟⎠ - c f ⎟
Median = l + ⎜ ⎟ × h, where
x + 1 + x2 + 22 + x3 + 32 + ... + xn + n2
2
⎜ f ⎟
x′ = 1 ⎜⎝ ⎟⎠
n
x1 + x2 + ... + xn 12 + 22 + ... + n2 l = lower limit of the median class
⇒ x′ = +
n n n = total frequency
( n + 1) ( 2n + 1) f = frequency of the median class
⇒ x′ = x + h = width of the median class
6
cf = cumulative frequency of the class preceding the
median class
MEDIAN
Some Points About Median
Median is the middle most or the central value of the variate
• It is an appropriate average in dealing with qualitative
in a set of observations, when the observations are arranged
data, like intelligence, wealth etc.
either in ascending or in descending order of their magni-
• The sum of the deviations of the items from median,
tudes. It divides the arranged series in two equal parts.
ignoring algebraic signs, is less than the sum from any
other point.
Calculation of Median
Median of an Individual Series
QUARTILES, DECILES AND PERCENTILES
Let n be the number of observations.
1. Arrange the data in ascending or descending order. Quartile
2. (A) If n is odd, then
1 Just as the median divides a set of observations (when
Median = value of the (n + 1)th observation arranged in ascending or descending order of magnitudes),
2
(B) If n is even, then into two equal parts, similarly Quartile divides the obser-
⎛ n⎞ ⎛n ⎞ vations into four equal parts. The value of the item midway,
Median = mean of the ⎜ ⎟ th and ⎜ + 1⎟ th
⎝ 2⎠ ⎝2 ⎠ between the first item and the median is known as first or
observation.
lower quartile and is denoted be Q1. The value of the item
midway between the last item and the median is known
Median of a Discrete Series Third or Upper Quartile and is denoted Q3. The median is
1. Arrange the values of the variate in ascending or known as the Second Quartile and is denoted by Q2. The
descending order. methods for finding the values of Q1 and Q3 are similar to
2. Prepare a cumulative frequency table. that of the median. In the case of ungrouped data, when
3. (A) If n is odd, then arranged in ascending or descending order of magnitudes
Q1, Q3 can be obtained as follows:
⎛ n + 1⎞
Median = size of the ⎜ th term. n +1 3( n + 1)
⎝ 2 ⎟⎠ Q1 = th item, Q3 = th item.
4 4
(B) If n is even, then
For a frequency distribution, Q1 and Q3 are given by,
⎛ ⎛ n⎞ ⎛ n ⎞ ⎞
⎜ ⎜⎝ 2 ⎟⎠ + ⎜⎝ 2 + 1⎟⎠ ⎟ [( n / 4) - C f ]
Median = size of the ⎜ ⎟ th term. Q1 = l + × h,
2 f
⎜ ⎟
⎜⎝ ⎟⎠
[(3n / 4) - C f ]
Q3 = l + × h,
f
Median of a Continuous Series
where l = lower limit of the class in which a particular
1. Prepare the cumulative frequency table. quartile lies,
⎛ n⎞ f = frequency of the class-interval in which a
2. Find the median class, i.e., the class in which the ⎜ ⎟
⎝ 2⎠ particular quartile lies,
th observation lies.
23.8 Chapter 23
AIM : FREE EDUCATION TO ALL APUL
i = class-interval of the class in which a particular Solution: (A)
quartile lies, 7 5 1
Arrange the data as a – , a – 3, a – , a – 2, a – ,
cf = cumulatively frequency of the class preceding 2 2 2
1
the class in which the particular quartile lies. a + , a + 4, a + 5.
2
[( nh / 4) - c f ] 1 5
In general, Qi = l + × h, i = 1, 2, 3, 4 a -2+a - 2a -
f Median = 2 = 2 = a – 5.
2 2 4
Decile
24. Median of 2nC0, 2n
C1, 2n
C2, 2n
C3,..., 2n
Cn (when n is
The value of the variable which divides the series, when even) is
arranged in ascending or descending order, into 10 equal (A) 2nCn/2 (B) 2nC(n+1)/2
parts is called decile. There are 9 deciles denoted by D1, D2 (C) 2nC(n–1)/2 (D) None of these
... D9. When the series is ungrouped the deciles are calcu-
lated as follows: Solution: (A)
2n
n×h C0, 2nC1, 2nC2, ... , 2nCn is odd number of binomial
Di = , i = 1, 2, ..., 9 coefficients (when n is even) and middle binomial
10
coefficient is 2nCn/2.
When the data is classified or grouped,
25. Median of 2nC0, 2nC1, 2nC2, 2nCn (when n is odd) is
[( nh / 10) - c f ]
Di = l + ×h
f (A)
1 2n
2
(C( n -1)/ 2 + 2n
C( n +1)/ 2 )
where symbols have their usual meaning.
(B) 2nCn/2
(C) 2nCn
Percentile (D) None of these
The value of the variable which divides the series, when
Solution: (A)
arranged in ascending or descending order, into 100 equal 2n
parts is called percentile. There are 99 percentiles denoted C0, 2nC1, 2nC2, ... , 2nCn is even number of bino-
by P1, P2, P3, P4, ..., P99 respectively. When the series is mial coefficients (when n is odd), and then mid-
ungrouped the percentiles are calculated by the following dle terms are 2nCn–1/2 and 2nCn+1/2. So median is
2n 2n
formula: Cn -1/ 2 + Cn +1/ 2
.
n×h 2
Pi = , h = 1, 2, ..., 99. 5
100 26. If a variable takes the discrete values a + 4, a -
2
When the data is classified or grouped, the percentiles are 7 1 1
a - , a – 3, a – 2, a + , a - , a + 5 (a > 0),
calculated by the formula 2 2 2
[( nh / 100) - c f ] then the median is
Pi = l + × h, i = 1, 2, ...., 99.
f 1 5
(A) a - (B) a -
where symbols have their usual meanings. 2 4
5
(C) a – 2 (D) a +
4
SOLVED EXAMPLES Solution: (B)
1⎛ 1⎞
7
23. If a variable takes the discrete values a + 4, a – , Median = ⎜⎝ a - 2 + a - ⎟⎠
2 2 2
5 1 1
a – ←, a – 3, a – 2, a + , a – , a +5 (a > 0), =a-
5
2 2 2 4
then the median is
5 1 MODE
(A) a – (B) a –
4 2
Mode is that value in a series which occurs most frequently.
5 In a frequency distribution, mode is that variate
(C) a – 2 (D) a +
4 which has the maximum frequency.
23.9
AIM : FREE EDUCATION Measures
TO ofALLCentral Tendency and Dispersion
APUL
Computation of Mode The commonly used measures of dispersion are:
Mode of Individual Series 1. Range
In the case of individual series, the value which is repeated 2. Quartile Deviation or Semi-interquartile range
maximum number of times is the mode of the series. 3. Mean Deviation
4. Standard Deviation
Mode of Discrete Series
Range
In the case of discrete frequency distribution, mode is
the value of the variate corresponding to the maximum It is the difference between the greatest and the smallest
frequency. observations of the distribution.
If L is the largest and S is the smallest observation in
Mode of Continuous Series a distribution, then its Range = L – S. Also,
1. Find the modal class, i.e., the class which has maxi- L-S
mum frequency. The modal class can be determined Coefficient of range = .
L+S
either by inspection or with the help of grouping table.
2. The mode is given by the formula Quartile deviation
f m - f m -1 Quartile deviation or semi-interquartile range is given by
Mode = l + × h,
2 f m - f m -1 - f m + 1 1
Q.D. = (Q3 – Q1)
where l = the lower limit of the modal class 2
h = the width of the modal class Q - Q1
fm – 1 = the frequency of the class preceding Coefficient of Q.D. = 3
Q3 + Q1
modal class
fm = the frequency of the modal class Mean deviation
fm +1 = the frequency of the class succeeding
For a frequency distribution, the mean deviation from an
modal class
average (median, or arithmetic mean) is given by,
In case, the modal value lies in a class other than the one n
containing maximum frequency, we take the help of the fol- ∑ fi xi - x
lowing formula; i =1
f m +1 M.D. = n
Mode = l = ×h ,
f m -1 + f m + 1 ∑ fi
i =1
where symbols have usual meaning.
Mean deviation
Coefficient of M.D. =
Some Points about Mode Corresponding average
• It is not based on all items of the series
• It is not necessary that a distribution has unique mode. Standard deviation
• As compared to other averages mode is affected to a The standard deviation of a statistical data is defined as the
large extent by fluctuations of sampling. positive square root of the squared deviations of observa-
• It is not suitable in a case where the relative impor- tions from the A.M. of the series under consideration.
tance of items have to be considered.
1. Standard deviation (also denoted by s ) for ungrouped
set of observations is given by
SYMMETRIC DISTRIBUTION
n
A distribution is a symmetric distribution if the values of ∑ (xi - x )2
mean, mode and median coincide. In a symmetric distri- i =1
s =
bution frequencies are symmetrically distributed on both n
sides of the centre point of the frequency curve. 2. Standard deviation for frequency distribution is given
by,
Measures of Dispersion n
⎛1 ⎞
1/ 2 Computation of Standard Deviation
s = ⎜ ( n2 - 1)⎟
⎝ 12 ⎠ Direct Method
Standard deviation shows the limits of variability by which 2
the individual observation in a distribution will vary from
s=
∑ x2 ⎛ ∑ x⎞
-⎜
the mean. For a symmetrical distribution with mean x , the ⎟
n ⎝ n ⎠
following area relationship holds good:
x ± s covers 68.27 % observations. Short-cut Method
x ± 2s covers 95.45 % observations. 2
s=
∑ d2 ⎛ ∑d⎞
-⎜
x ± 3s covers 99.73 % observations. ⎟ , for ungrouped data
n ⎝ n ⎠
These limits are illustrated by the following curve known
as Normal Curve. where A is assumed mean and d = x – A.
2
s=
∑ fd 2 ⎛ ∑ fd ⎞
-⎜ ⎟ , for grouped data
N ⎝ N ⎠
68.27% where N = ∑f.
95.44%
99.73% Step-deviation Method
x – 3σ x – 2σ x – 1σ x x + 1σ x + 2σ x + 3σ
2
Fig. 23.1 s =h
∑ fd ′ 2 ⎛ ∑ fd ′ ⎞
-⎜
x-A
⎟ ; d′ =
N ⎝ N ⎠ h
Empirical relationships
If the data is moderately non-symmetrical, then the follow- Combined Standard Deviation
ing empirical relationships hold: Let A1 and A2 be two series having n1 and n2 observations
4 respectively. Let their A.M. be x1 and x2 , and standard
Mean deviation = s
5 deviations be s1 and s2. Then the combined standard devi-
2 ation s or s12 of A1 and A2 is given by
Semi-Inter-quartile range = s
3 2
Probable error of standard deviation = s n1s12 + n2s 22 + n1d12 + n2 d22
3 s12 or s =
= Semi-inter-quartile range. n1 + n2
5
Quartile deviation = M.D.
6 n1 (s12 + d12 ) + n2 (s 22 + d22 )
From these relationships, we have =
n1 + n2
4 S.D. = 5 M.D. = 6 Q.D
where d1 = ( x1 - x12 ), d2 = ( x2 - x12 ),
Coefficient of S.D. (C.V.)
n1 x1 + n2 x2
For comparing two or more series for variability, the relative and x12 = is the combined mean.
n1 + n2
measure, called coefficient of variation (C.V.) is used. This
measure is defined as
23.11
AIM : FREE EDUCATIONMeasures
TO ofALL
Central Tendency and Dispersion
APUL
29. The mean deviation from the mean for the set of
I M P O R TA N T P O I N T S observations – 1, 0, 4 is
(A) less than 3 (B) less than 4
■ Quartile deviation is less affected by extreme values of
the series.
(C) greater than 2.5 (D) greater than 4.9
■ Mean deviation is based on all the items of series. It is Solution: (A and B)
therefore more representative than the range or quartile -1 + 0 + 4
deviation. x = = 1.
3
■ Mean deviation from the median is less than that mea-
1
sured from any other mean. \ Mean Deviation = (| – 1 – 1| + | 0 – 1 | + | 4 – 1 | )
■ Standard deviation ≤ Range i.e., variance ≤ (Range)2. 3
=2
n2 - 1
■ S.D. of first n natural numbers is .
12 30. If the S.D of a set of observations is 4 and if each
observation is divided by 4, the S.D of the new set of
observations will be
(A) 4 (B) 3
SOLVED EXAMPLES (C) 2 (D) 1
Solution: (D)
27. The coefficient of variation of two series are 58% and
We know that if y = x/h when s y = sx /| h |.
69%. If their standard deviations are 21.2 and 15.6,
\ The S.D. of new set of observations will be 4/4 = 1.
then their A.Ms are
(A) 36.6, 22.6 (B) 34.8, 22.6 31. A sample of 35 observations has the mean 80 and S.D.
(C) 36.6, 24.4 (D) None of these as 4. A second sample of 65 observations from the
same population has mean 70 and S.D. 3. The S.D. of
Solution: (A) the combined sample is
We know that (A) 5.85 (B) 5.58
s × 100 (C) 3.42 (D) None of these
C.V. =
x
Solution: (A)
s
or x = × 100
C.V. Here n1 = 35, x1 = 80, s1 = 4,
21.2 × 100
\ Mean of first series = = 36.6 n2 = 65, x2 = 70, s 2 = 3.
58
35 × 80 + 65 × 70
15.6 × 100 \ x12 = = 73.5.
Mean of second series = = 22.6 35 + 65
69
s12 = ⎡ 35(16 + 42 × 25) + 65(9 + 12 × 25) ⎤
28. Mean deviation of the series a, a + d, a + 2d, a + 2nd ⎢ ⎥
from its mean is ⎣ 100 ⎦
( n + 1) d nd = 34.21 = 5.85
(A) (B)
( 2n + 1) 2n + 1
32. If μ is the mean of a distribution, then
n ( n + 1) d ( 2n + 1) d
(C) (D) ∑ fi ( yi - μ ) is equal to
( 2n + 1) n ( n + 1)
(A) M.D. (B) S.D.
Solution: (C) (C) 0 (D) None of these
2n + 1
.( a + a + 2nd ) Solution: (C)
x = 2 = a + nd Sfi yi
( 2n + 1) We have, μ=
Sf i
S x - x = 2d (1 + 2 + ......+ n) = n (n + 1) d
⇒ Sfi yi – S fi μ = 0
n ( n + 1)d
\ M.D. =
2n + 1 ⇒ S fi (yi – μ) = 0
23.12 Chapter 23
AIM : FREE EDUCATION TO ALL APUL
33. The means of five observations is 4 and their variance 34. If 25 % of the items are less than 20 and 25 % are more
is 5.2. If three of these observations are 1, 2, and 6, then 40, the quartile deviation is
then the other two are (A) 20 (B) 30
(A) 2 and 9 (B) 3 and 8 (C) 40 (D) 10
(C) 4 and 7 (D) 5 and 6
Solution: (D)
Solution: (C) 40 - 20
Q.D. = = 10.
x1 = 4, N = 5 2
35. The sum of squares of deviations for 10 observations
S(x - x ) 2
and = 5.2 taken from mean 50 is 250. The coefficient of variation
N is
⇒ S(x - x ) 2 = (5.2) 5 (A) 10 % (B) 40 %
(C) 50 % (D) None of these
\ S(x - x ) 2 = 26
Solution: (A)
2 2 2
\ (1 – 4) + (2 – 4) + (6 – 4) + Co-efficient of variation
(a – 4)2 + (b – 4)2 = 26 s s
= × 100 = × 100 (Q x = 50) = 2s
x 50
where a, b are the other two observatios.
\ 9 + 4 + 4 + (a – 4)2 + (b – 4)2 = 26 S ( xi - 50) 2 250
Also, s= = = 25 = 5
n 10
\ (a – 4)2 + (b – 4)4 = 9 \ Co.efficient of variation = 2 × 5 = 10%.
1+ 2 + 6 + a + b
Also, =4
5
\ a + b = 20 – 9 = 11
Clearly 4, 7 only satisfy the above equation in a, b.
Hence reqd. numbers are 4, 7.
EXERCISES
1. The average of n numbers x1, x2, x3,..., xn is M. If xn is 4. The mean weight of 9 items is 51. If one more item is
replaced by x′, then new average is added to the series the mean becomes 16. The value of
nM - xn + x ′ the 10th item is
(A) M – xn + x′ (B)
n (A) 35 (B) 30
( n - 1) M + x ′ M - xn + x ′ (C) 25 (D) 20
(C) (D)
m n
5. The mean and S.D. of the marks of 200 candidates
2. The standard deviation of 25 numbers is 40. If each of were found to be 40 and 15 respectively. Later, it was
the numbers is increased by 5, then the new standard discovered that a score of 40 was wrongly read as 50.
deviation will be The correct mean and S.D. respectively are
(A) 40 (B) 45 (A) 14.98, 39.95 (B) 39.95, 14.98
21
(C) 40 + (D) None of these (C) 39. 95, 224.5 (D) None of these
25
3. If M.D. is 12, the value of S.D. will be 6. If Q.D. is 16, the most likely value of S.D. will be
(A) 15 (B) 12 (A) 24 (B) 42
(C) 24 (D) None of these (C) 10 (D) None of these
23.13
AIM : FREE EDUCATIONMeasures
TO ofALL
Central Tendency and Dispersion
APUL
7. If a variable x takes values 0, 1, 2, ..., n with fre- 14. Consider any set of observations x1, x2, x3,..., x101; it
quencies proportional to the binomial coefficients nC0, being given that x1 < x2 < x3 < ... < x100 < x101; then
n
C1, nC2,..., nCn, then the Var (x) is the mean deviation of this set of observations about a
point k is minimum when k equals
n2 - 1 n
(A) (B) (A) x1 (B) x51
12 2
x + x2 + ... + x101
n (C) 1 (D) x50
(C) (D) None of these 101
4
8. The sum of squares of deviations for 10 observations 15. The mean of the numbers
50 50
taken from mean 50 is 250. The coefficient of variation C0 50 C2 50
C4 C50
is , , ..., equals
1 3 5 51
(A) 50% (B) 10%
250 249
(C) 40% (D) None of these (A) (B)
51 51
9. If the standard deviation of n observations x1, x2,..., 249
xn is 4 and another set of n observations y1, y2,..., yn (C) (D) None of these
39 × 17
is 3. The standard deviation of n observations x1 – y1,
x2 – y2,..., xn – yn is 16. The standard deviation of a distribution is 30 and each
2 item is raised by 3, then new S.D. is
(A) 1 (B) (A) 32 (B) 28
3
(C) 27 (D) None of these
(C) 5 (D) Data insufficient
17. For three numbers a, b, c product of the average of the
1 n
10. Let r be the range and S = 2
∑ ( x - x)2 be the
n - 1i =1 i
1 1 1
numbers a2, b2, c2 and 2 , 2 , 2 cannot be less than
a b c
S.D. of a set of observations x1, x2,..., xn, then
(A) 1 (B) 3
n n (C) 9 (D) None of these
(A) S ≤ r (B) S = r
n -1 n -1 18. The variance of a, b and g is 9, then variance of 5a, 5b
n and 5g is
(C) S ≥ r (D) None of these
n -1 (A) 45 (B) 9/5 (C) 5/9 (D) 225
11. The A.M. of n numbers of a series is x . If the sum of 19. Mean of the numbers 1, 2, 3,..., n with respective
the first (n – 1) term is k, them the nth number is weights 12 + 1, 22 + 2, 32 + 3,..., n2 + n is
(A) x - k (B) nx - k 3n( n + 1) 2n + 1
(A) (B)
(B) x - nk (D) nx - nk 2( 2n + 1) 3
3n + 1 3n + 1
12. If a variable takes values 0, 1, 2,..., n with frequencies (C) (D)
4 2
n n( n - 1) n - 2 2
qn, q n -1 p , q p ,..., pn, where p + q = 1, 20. The G.M. of the number 3, 32, 33,..., 33n is
1 1. 2
n 3n
then the mean is
(A) 32 (B) 32
(A) np (B) nq
3n+1 n+1
(C) n(p + q) (D) None of these
(C) 3 2 (D) 32
13. The S.D. of a variate x is s. The S.D. of the variate
21. The reciprocal of the weighted mean of first n natural
ax + b
where a, b, c are constants, is numbers whose weights are equal to the squares of the
c corresponding numbers is
⎛ a⎞ a 2( 2n + 1) 3n( n + 1)
(A) ⎜ ⎟ s (B) s (A) (B)
⎝ c⎠ c 3n( n + 1) n( 2n + 1)
⎛ a2 ⎞ 3n( n + 1)
(C) ⎜ 2 ⎟ s (D) None of these (C) (D) None of these
⎝c ⎠ 2n + 1
23.14 Chapter 23
AIM : FREE EDUCATION TO ALL APUL
22. The A.M. of a set of 50 numbers is 38. If two numbers
3n +1 (3n - 1)
of the set, namely 55 and 45 are discarded, the A.M. of (A) x + (B) x + 3
the remaining set of numbers is n 2n
(A) 38.5 (B) 37.5 3n (3n - 1)
(C) x + (D) x + 3
(C) 36.5 (D) 36 n 2n
23. The mean weight per student in a group of seven stu- 30. A sequence of odd positive integers is written as
dents is 55 kg If the individual weights of 6 students 1
are 52, 58, 55, 53, 56 and 54; then weight of the sev- 3 5 7 9
enth student is 11, 13, 15, 17, 19, 21, 23, 25, 27
(A) 55 kg (B) 60 kg The mean of the nth row is
(C) 57 kg (D) 50 kg
n3 ( 2n2 + 1) n3 ( 4 n2 + 2)
24. If the mean of a set of observations x1, x2, x3,..., xn is (A) (B)
3 6
x , then mean of observations xi + 3i ∀ i = 1, 2, 3,... n
equals n( n - 1) ( 2n - 1) n( 2n2 + 1)
(C) (D)
6 3
3( n + 1)
(A) x + 3(n + 1) (B) x +
2 31. The arithmetic mean of a set of observation is . If each
n +1 observation is divided by a and then is increased by
(C) x + (D) None of these 10, the means of the new series is
2n
x x + 10
(A) (B)
25. The weighted mean of the square of 1st n natural a a
numbers whose weights are corresponding numbers,
x + 10a
equals (C) (D) a x + 10
a
( n + 1) ( 2n + 1) n( n + 1)
(A) (B) 32. The average salary of male employees in a firm was
2 2
Rs. 520 and that of females was Rs. 420. The mean
n +1 salary of all the employees was Rs. 500. The percent-
(C) (D) None of these
2 age of male employees is
(A) 80 (B) 60 (B) 40 (D) 20
26. If the variate of a distribution takes the values 1, 2,
3, ...n with frequencies n, n – 1, n – 2, ... 3, 2, 1, then 33. The average weight of students in a class of 35 stu-
mean value of the distribution is dents is 40 kg. If the weight of the teacher be included,
1
n( n + 2) n( n + 1) ( n + 2) the average rises by kg; the weight of the teacher is
(A) (B) 2
3 6 (A) 40.5 kg (B) 50 kg
n+2 ( n + 1) ) ( n + 2) (C) 41 kg (D) 58 kg
(C) (D)
3 6 34. An automobile driver travels from plane to a hill station
120 km distant at an average speed of 30 km per hour.
27. The means of five observations is 4 and their variance
He then makes the return trip at an average speed of
is 5.2. If three of these observations are 1, 2 and 6, then
25 km per hour. He covers another 120 km distance on
the other two are
plane at an average speed of 50 km per hour. His aver-
(A) 2 and 9 (B) 3 and 8 age speed over the entire distance of 360 km will be
(C) 4 and 7 (D) 5 and 6 1
30 + 25 + 50
28. If the variate takes the values 0, 2, 4, 8, ... 2n with fre- (A) km/h (B) (30 ⋅ 25 ⋅ 50) 3
3
91 × 8
quencies nC0, nC1, nCn and if the mean is , then 3
n equals 2n (C)
1 1 1
km/h (D) None of these
+ +
(A) 4 (B) 6 30 25 50
(C) 5 (D) None of these
35. If the mean deviation about the median of the numbers
29. The mean of n items is x . If each item is successively a, 2a, ..., 50a is 50, then | a | equals
increased by 3, 32, 33,... 3n, then new mean equals (A) 5 (B) 2 (C) 3 (D) 4
23.15
AIM : FREE EDUCATIONMeasures
TO ofALL
Central Tendency and Dispersion
APUL
(A) 10.0 (B) 20.0
36. Let x1, x2, ..., xn be n observations, and let x be their
(C) 10.1 (D) 20.2
arithmetic mean and s2 be the variance
Statement-1: Variance of 2x1, 2x2, ..., 2xn is 4s2. 40. Statement-1: The variance of first n even natural
Statement-2: Arithmetic mean 2x1, 2x2, ..., 2xn is 4x. n2 - 1
numbers is
(A) Statement-1 is false, Statement-2 is true. 4
(B) Statement-1 is true, statement-2 is true; state- Statement-2: The sum of first n natural numbers is
ment-2 is a correct explanation for Statement-1. n( n + 1)
(C) Statement-1 is true, statement-2 is true; and the sum of squares of first n natural
2
statement-2 is not a correct explanation for n( n + 1) ( 2n + 1)
Statement-1. numbers is
6
(D) Statement-1 is true, statement-2 is false.
(A) Statement-1 is true, Statement-2 is true;
37. If the mean of a set of observations x1, x2, ..., x10 is 20 Statement-2 is a correct explanation for
then the mean of x1 + 4, x2 + 8, x3 + 12, ..., x10 + 40 is Statement-1.
(A) 34 (B) 42 (C) 38 (D) 40 (B) Statement-1 is true, Statement-2 is true;
Statement-2 is not a correct explanation for
38. The mean of the numbers a, b, 8, 5, 10 is 6 and the
Statement-1.
variance is 6.80. Then which one of the following
(C) Statement-1 is true, Statement-2 is false.
gives possible values of a and b?
(D) Statement-1 is false, Statement-2 is true.
(A) a = 0, b = 7 (B) a = 5, b = 2
(C) a = 1, b = 6 (D) a = 3, b = 4
39. If the mean deviation of number 1, 1 + d, 1 + 2d, ..., 1
+ 100d from their mean is 255, then the d is equal to
41. The median of a set of 9 distinct observations is 44. In a series of 2n observations, half of them equal a and
20.5. If each of the largest 4 observations of the set is remaining half equal -a. If the standard deviation of
increased by 2, then the median of the new set [2003] the observations is 2, then |a| equals [2004]
(A) is increased by 2 1
(A) (B) 2
(B) is decreased by 2 n
(C) is two times the original median 2
(C) 2 (D)
(D) remains the same as that of the original set n
45. If in a frequency distribution, the mean and median are
42. Let two numbers have arithmetic mean 9 and geomet-
21 and 22 respectively, then its mode is approximately
ric mean 4. Then these numbers are the roots of the
[2005]
quadratic equation [2004]
(A) 22.0 (B) 20.5
(A) x2 + 18x + 16 = 0
(C) 25.5 (D) 24.0
(B) x2 - 18x - 16 = 0
(C) x2 + 18x - 16 = 0 46. Let x1, x2, …, xn be n observations such that
(D) x2 - 18x + 16 = 0 ∑ xi2 = 400 and ∑ xi = 80. Then a possible value of
n among the following is [2005]
43. Consider the following statements [2004]
(A) 15 (B) 18
(A) Mode can be computed from histogram (C) 9 (D) 12
(B) Median is not independent of change of scale
(C) Variance is independent of change of origin and 47. Suppose a population A has 100 observations 101,
scale. 102, … , 200, and another population B has 100 obser-
Which of these is/are correct? vations 151, 152, … , 250. If VA and VB represent the
(A) only (A) (B) only (B) variances of the two populations, respectively, then
(C) only (A) and (B) (D) (A), (B) and (C) VA
is [2006]
VB
23.16 Chapter 23
AIM : FREE EDUCATION TO ALL APUL
(A) 1 (B) 9/4
(C) 4/9 (D) 2/3 Statement-1: Variance of 2x1, 2x2 . . . 2xn is 4 s 2 .
Statement-2: Arithmetic mean of 2x1, 2x2 . . . 2xn is 4
48. The average marks of boys in a class is 52 and that of x. [2012]
girls is 42. The average marks of boys and girls com- (A) Statement-1 is false, statement-2 is true
bined is 50. The percentage of boys in the class is (B) Statement-1 is true, statement-2 is true; statement
[2007] 2 is a correct explanation for statement 1
(A) 40 (B) 20 (C) Statement-1 is true, statement-2 is true; statement
(C) 80 (D) 60 2 is not a correct explanation for statement 1
(D) Statement-1 is true, statement-2 is false
49. The mean of the numbers a, b, 8, 5, 10 is 6 and the
variance is 6.80. Then which one of the following 54. All the students of a class performed poorly in
gives possible values of a and b? [2008] Mathematics. The teacher decided to give grace marks
(A) a = 0, b = 7 of 10 to entire class. Which of the following statistical
(B) a = 5, b = 2 measures will not change even after the grace marks
(C) a = 1, b = 6 were given? [2013]
(D) a = 3, b = 4 (A) median (B) mode
(C) variance (D) mean
50. If the mean deviation of number 1, 1 + d, 1 + 2d, ….. ,
1 + 100d from their mean is 255, then the d is equal to 55. The variance of the first 50 even natural numbers is
[2009] [2014]
(A) 10.0 (B) 20.0 833
(A) (B) 833
(C) 10.1 (D) 20.2 4
437
51. For two data sets, each with size 5, the variances are (C) 437 (D)
4
given to be 4 and 5 and the corresponding means are
given to be 2 and 4, respectively. The variance of the 56. The mean of the data set comprising of 16 observations
combined data set is [2010] is 16. If one of the observation valued 16 is deleted and
11 three new observations valued 3, 4 and 5 are added to
(A) (B) 6
2 the data, then the mean of the resultant data, is
13 5 [2015]
(C) (D) (A) 16.0 (B) 15.8
2 2
(C) 14.0 (D) 16.8
52. If the mean deviation about the median of the numbers
a, 2a . . . 50a is 50, then |a| equals [2011] 57. If the standard deviation of the numbers 2, 3, a and 11
(A) 3 (B) 4 is 3.5, then which of the following is true?
(C) 5 (D) 2 [2016]
2 2
(A) 3a - 23a + 44 = 0 (B) 3a - 26 a + 55 = 0
53. Let x1, x2 . . . xn be n observations, and let x be their
arithmetic mean and s 2 be their variance. (C) 3a 2 - 32a + 84 = 0 (D) 3a 2 - 34 a + 91 = 0
ANSWER KEYS
r=0
n
The correct option is (B) n
n n n
2. If each item of a data is increased or decreased by the
⇒ x =
2 n ∑ n -1Cr -1 =
2 n
2n -1 =
2
r =1
same constant, the standard deviation of the data remains
unchanged. ⎡ n n -1 ⎤
⎢Q ∑ Cr -1 = 2n -1 ⎥
The correct option is (A) ⎢⎣ r = 1 ⎥⎦
3. We know that n
Q.D. =
5
× M.D. =
5
× 12 = 10 ∑ r 2nCr
1 ∑ fi xi2
6 6 and ∑ fi xi2 = = r=0
3 3 n ∑ fi n
\ S.D. =
2
× Q.D. =
2
× 10 ⇒ S.D. = 15 ∑ n Cr
r=0
The correct option is (A)
n
n
4. Let the values of 9 items be x1, x2, ..., x9
1
∑ r2 r
n -1
Cr -1
r=0
Therefore, mean of x1, x2, ..., x9 is
x + x2 + ... + x9
⇒
n
∑ fi xi2 = 2n
15 = 1
9 1 n n
⇒ x1 + x2 + ... + x9 = 15 × 9 = 135 ⇒
n
∑ i i 2n ∑ ( r - 1 + 1)
f x 2
=
n -1
Cr -1
r=0
Let x10 be the 10th item. The mean of x1, x2, ... x9, x10 is 16.
x1 + x2 + ... + x9 + x10 1 n ⎡ n n ⎤
⇒
10
= 16 ⇒
n
∑ fi xi2 = n ⎢ ∑ ( r - 1) n -1Cr -1 + ∑ n -1Cr -1 ⎥
2 ⎢⎣ r = 0 r=0 ⎥⎦
⇒ x1 + x2 + ... + x9 + x10 = 160 ⇒ 135 + x10 = 160
\ x10 = 160 – 135 = 25 n ⎡ n ( n - 1) ⎤
= ⎢ ∑ ( r - 1) n-2
Cr - 2 + 2n -1 ⎥
The correct option is (C) 2n ⎢⎣ r = 0 ( r - 1) ⎥⎦
5. Corrected Sx = 40 × 200 – 50 + 40 = 7990 1 n
\ Corrected x = 7990 / 200 = 39.95 ⇒
n
∑ fi xi2 = 2n [(n – 1)2n–2 + 2n–1]
2
Incorrect Sx2 = n[s 2 + x ] = 200[152 + 402] = 365000 n n
= n [(n – 1 + 2)2n–2] = n (n + 1)
Correct Sx2 = 365000 – 2500 + 1600 = 364100 2 2
1
\ Corrected s =
364100
- (39.95) 2 Now, Var (x) = ∑ fi xi - x
2 2
200 n
= (1820.5 - 1596) n( n + 1) n2 n
\ Var (x) = - =
4 4 4
= 224.5 = 14.98 The correct option is (C)
The correct option is (B)
250
3 8. S.D. (s) = = 25 = 5
6. We know that, S.D. = Q.D. 10
2 s
3 Hence, coefficient of variation = × 100
\ S.D. = × 16 = 24 mean
2
The correct option is (A)
23.18 Chapter 23
AIM : FREE EDUCATION TO ALL APUL
=
5
× 100 = 10%. + ... + n . p n
⇒ x =
50 n n -1 n( n - 1) n - 2 2
qn + q p+ q p + ... + p n
The correct option is (B) 1 2!
9. S.D = s x2- y 0 ⋅ nC0 q n p0 + 1⋅ nC1q n-1 p + ... + n ⋅ nCn q0 p n
⇒ x =
1 n
n
C0 q n p0 + nC1 q n-1 p1 + ... + nCn q n-n p n
= ∑ ( xi - yi - x + y)2
n i =1 n
1 n 1 n
∑ r ⋅ nCr qn-r pr
= ∑
n i =1
( xi - x ) 2 + ∑ ( yi - y )
n i =1
⇒ x =
r=0
n
∑ n Cr q n- r p r
2 n r=0
- ∑ ( xi - x ) ( yi - y )
n i =1 n
n
∑ r⋅ r n-1
Cr -1 q n-r p ⋅ p r -1
= s x2 + s 2y - 2 cov ( x, y ) =
r=0
n
As cov (x, y) is not known, therefore we cannot ∑ n Cr q n- r p r
r=0
find s x2- y or s x - y . Hence data is insufficient.
The correct option is (D) ⎛ n ⎞
np ⎜ ∑ n -1Cr -1 p r -1q( n -1) - ( r -1) ⎟
10. We have r = max | xi – xj | ⎝r =1 ⎠
i≠ j
⇒ x =
n
1 n ∑ n Cr q n - r p r
and S2 = ∑ ( xi - x)2
n - 1i =1
r=0
np( q + p) n -1
⎛ x + x2 + ... + xn ⎞
2 ⇒ x =
Now, ( xi - x ) = ⎜ xi - 1
2
⎟⎠ ( q + p) n
⎝ n
\ x = np (Q q + p = 1)
1
= [(xi – x1) + (xi – x2) + ... + (xi – xi – 1) The correct option is (A)
n2
ax + b a b
+ (xi – xi + 1) + ... + (xi – xn)] 13. Let y = i.e., y = x +
1 c c c
≤ [( n - 1)r ]2 (Q | xi – xj | ≤ r) b
n2 a
i.e., y = Ax + B, where A = , B =
n c c
⇒ ( xi - x ) ≤ r2 ⇒ ∑ ( xi - x)2 ≤ nr 2 \ y = Ax + B
i =1
\ y - y = A( x - x ) ⇒ ( y - y ) 2 = A2 ( x - x ) 2
1 n nr 2 nr 2
⇒ ∑
n - 1i =1
( xi - x ) 2 ≤
( n - 1)
⇒ S2 ≤
( n - 1) ⇒ S( y - y ) 2 = A2 S( x - x ) 2
n ⇒ n ⋅s y2 = A2 ⋅ ns x2
⇒ S≤r
n -1 ⇒ s 2y = A2s x2
The correct option is (A) a
⇒ s y = | A |sx ⇒ s y = sx
11. Let the numbers be x1, x2,..., xn. Then, c
1 n a
x = ∑ xi
n i =1
Thus, new S.D. =
c
s.
51 ⎤1
1 ⎡ (1 + x ) 51
(1 - x ) n( n + 1) ⎡ n( n + 1) 2n + 1⎤
= ⎢ - ⎥ ⎢ 2 +
2 ⎢⎣ 51 51 ⎥⎦ 2 ⎣ 3 ⎥⎦
0 =
n( n + 1) ⎡ 2n + 1 ⎤
50
C2 C50 5051
C4 250 50
⎢ 3 + 1⎥
\ 50
C0 + + = 12 = + ... + 2 ⎣ ⎦
3 51 5
2 51 51
3n2 + 7n + 2 (3n + 1) ( n + 2) 3n + 1
50 50
Now number of terms from C0 to C50 are 26 (items) = = =
2( 2n + 4) 4 ( n + 2) 4
\ Required mean
50
C0 50
C2 50
C4 50
C50 The correct option is (C)
+ + + ... + 20. G.M. of 31, 32, 33,..., 33n
1 3 5 51
1
26 1
= (3.32 ⋅ 33...33n ) 3n =
50 49 (31+ 2 + 3 + ...3n ) 3n
2 1 2
= × = 3n( 3n +1)
51 26 39 × 17 = 3 3n ⋅ 2
=
3n+1
= ( n + 1)( n - 1) = n - 1
2
1 2 2
= 2 (1 + 22 + ... + n2 ) - ( n + 1) 2
n 3 3
4 n( n + 1)( 2n + 1) \ Statement 1 is false.
= - ( n + 1) 2
n 6 The correct option is (D)
∑ xi = 2 ∑ x-x
2
5
∑ xi = 10; ∑ yi = 20 s=
n
s x2
⎛1
⎝2
⎞
= ⎜ ∑ xi2 ⎟ - ( x ) 2 =
⎠
1
5
(∑ ) - 16
yi2 Therefore, y = x + 10 ⇒ y + x + 10
∑ xi2 = 40 ∑ y + 10 - y - 10 ∑ y-y
2 2
s1 = = =s .
n n
∑ yi2 = 105 The correct option is (C)
2
⎛ x + y⎞
s z2 =
1
10
( )
∑ xi2 + ∑ yi2 - ⎜⎝ 2 ⎟⎠ ⎛ ∑ xi2 ⎞
55. s 2 = ⎜ ⎟-x
2
⎝ n ⎠
1 145 - 90
= ( 40 + 105) - 9 = 50
10 10
55 11 ∑ 2r
= = x= r =1
= 51
10 2 50
The correct option is (A) 50
1
∑ 4r 2
52.
n
∑ xi - A s2 = r -1
50
- (51) = 833
2