You are on page 1of 31

Measures of Central Tendency

and Dispersion
Why Describe Central Tendency?
• Data often cluster around a central value that
lies between the two extremes. This single
number can describe the value of scores in
the entire data set.
• There are three major measures of central
tendency.
1) Mean
2) Median
3) Mode
The Mode
• The mode is the most frequently occurring
number in a set of data.
• E.g., Find the mode of the following numbers…
• 15, 20, 21, 23, 23, 23, 25, 27, 30
• Also, if there are two modes, the data set is
bimodal.
• If there are more than two modes, the data set is
said to be multimodal.
The Median
• The middle score when all scores in the data set
are arranged in order.
• Half the scores lie above and half lie below the
median.
• E.g., Find the median of the following numbers…

10, 12, 14, 15, 17, 18, 20.


• When there are an even number of
scores, you must take the average of the
middle two scores.

Eg., 10, 12, 14, 15, 17, 18


(14 + 15)/2 = 14.5.
• The median can also be calculated from a
frequency distribution.
• E.g., A stats class received the following marks out
of 20 on their first exam.

X freq Cumulative freq


20 1 15
19 2 14
16 2 12
14 1 10
12 4 9
11 2 5
10 3 3

What is the median grade?


• Step 1 - Multiply 0.5 times N + 1 to obtain
the location of the middle frequency.
0.5(15 + 1) = 8
• Step 2 - Locate this score on your
frequency distribution.
12

Example in the book on pp. 77-78.


The Mean
• The is the sum of all the scores data set
divided by the number of scores in the set.
E.g., What’s the mean of the
following test scores?
x
x = 56, 65, 75, 83, 92
n
x = 371/5 = 74.2
• The mean can also be calculated using a frequency
distribution.
• The following scores were obtained on a stats exam
marked out of 20.

X freq
20 1
19 2
16 2
14 1
12 4
11 2
10 3
Find the mean of the exam scores.
• Multiply each score by the frequency. Add
them together and divide by N

X freq fX
20 1 20
19 2 38 X = fX/N
16 2 32
14 1 14
12 4 48 = 204/15
11 2 22
10 3 30 = 13.6

N = 15 fX = 204
Characteristics of the Mean
• Summed deviations about the mean equal 0.

Score X-X
2 2 - 5 = -3
3 3 - 5 = -2
5 5-5=0
7 7-5=2
__8__ 8-5=3
 X = 25 (x - x) = 0
X=5
• The mean is sensitive to extreme scores.

Score Score
2 2 Note, the median
3 3 remains the same in
5 5 both cases.
7 7
__8__ __33__
 X = 25  X = 50
X=5 X = 10
• The sum of squared deviations is least
about the mean (pp. 82-83).

Score (X - X)2
2 (2 - 5)2 = 9
3 (3 - 5)2 = 4
5 (5 - 5)2 = 0
7 (7 - 5)2 = 4
__8__ (8 - 5)2 = 9
 X = 25 (x - x) 2
= 26

X=5
The Weighted Mean
• Used when a single mean must be
calculated for two or more groups of
different sizes.
• The different sizes of the groups is
accounted for or “weighted”.
• E.g., The means of five different stats classes are
as follows. Calculate the weighted mean.

Sample Mean f
58 45
64 42
77 83
62 38
52 45
The larger the sample, the greater its weight in
determining the overall mean.
• Step 1: Multiply each sample mean by the
N of that group.
• Step 2: Add these products together.
• Step 3: Divide by the overall N.
Sample Mean f fX
58 45 2610
64 42 2688
77 83 6391
62 38 2356
52 45 2340
N = 253  fX = 16385

X =  fX/Nt

= 16385/253 = 64.76
Comparison of the Mean,
Median, and Mode
• The mode is the roughest measure of central
tendency and is rarely used in behavioral statistics.
• Mean and median are generally more appropriate.
• If a distribution is skewed, the mean is pulled in the
direction of the skew. In such cases, the median is a
better measure of central tendency.
Skewness of Distribution
 Comparing the mean and the median

Normal Negative
Positive Skew Skew
Distribution

Mean & Median Mean Mean Median


Median the
same
Why Measure Dispersion?
• Measures of dispersion tell us how spread out
the scores in a data set are. Surely all scores
will not be equal to the mean.
• There are four major measures of dispersion
we will look at:
• Range (crude range)
• Semi-Interquartile Range
• Variance
• Standard Deviation
The Range
• The simplest measure of variability. Simply
the highest score minus the lowest score.
• Limited by extreme scores or outliers.

E.g., Find the range in the following test scores.


100, 74, 68, 68, 57, 56

Range = H - L = 100 - 56 = 44
Semi-Interquartile Range
• A measure of variability obtained by
subtracting the score at the 25th percentile
(i.e., first quartile) from the score at the
75th percentile (i.e.,third quartile) and
dividing by 2.

• If the distribution is normal, this cuts off the


middle 50% of all cases.
Normal Distribution

Q1 Q2 Q3

• Cuts off approximately 50% of all cases if the


distribution is skewed (best measure in a
skewed distribution).
• E.g., An organic chemistry class obtains
the following distribution on their first
exam...
X f cf What is the semi-
91 1 32 interquartile
61 1 31
range?
59 4 30
56 6 26
54 7 20
52 5 13
50 4 8
47 3 4
45 1 1
• Step 1 - 25th percentile
0.25(32) = 8
The eighth score is at the 1st quartile.
50
• Step 2 - 75th percentile
0.75(32) = 24
The 24th score is at the 3rd quartile.
56
• Step 3 - Calculate the median.
0.5(N + 1)
0.5(32 + 1) = 16.5 ~ 17.
The median is the 17th score.
54

• Step 4 - Subtract the 1st quartile from the


3rd quartile and divide by 2.
(56 - 50)/2 = 3
• Step 4 - Add and subtract 3 from the
median.
54 + 3 = 57
54 - 3 = 51

50% of the scores lie between 51 and 57.


54 + 3.
The Variance
• The sum of the squared deviations from
the mean divided by N.

 (x - x)
2

s 2
=
N
Calculating Variance (Deviation Formula)
X X-X (X -
X)2
12 3 9
11 2 4
10 1 1
9 0 0
9 0 0
9 0 0
8 -1 1
7 -2 4
6 -3 9
 x = 81  (x - x) = 0  (x - x) = 28
2

x=9
S2 =  (x - x)2 = 28 = 3.11
n 9
A Simpler Formula(raw score method)
X X2 SS =  x2 - ( x)2
12 144 N

11 121
10 100 = 757 - (81)2 = 28
9 81 9
9 81
9 81 s 2 = SS

8 64 N
7 49
6 36 = 28 = 3.11
 x = 81  x2 = 757 9
Calculating Standard Deviation
• Simply calculate the square root of the
variance.

• So if s2 from the previous example was


3.11, the standard deviation (denoted by s)
is 1.76.

You might also like