Download as ppt, pdf, or txt
Download as ppt, pdf, or txt
You are on page 1of 27

Last Lecture

• Visual presentation of data (graphs)


• Shapes of distributions
• Examples and review

Today

• Measures of central tendency


– Mode
– Median
– Arithmetic mean
• Properties of each statistic
• Summation notation
• Relationships
Definition
“Central tendency”

• Summarizes all the scores in a sample


• “Most characteristic” score

0|
1 | 06
2 | 0345678
3 | 00012222345677799
4 | 011122444556666666667788899
5 | 000012334455555667788
6 | 00024555569
7|5
8 | 357
9|
Three common statistics
• Mode

• Median

• Arithmetic mean (average)


The mode
Definition: the mode of a sample is the most frequently occurring score.
Think of it as the “most likely” score.

Example: {4, 5, 6, 6, 7, 7, 7, 7, 8, 8, 9, 10}

The mode Mo = 7.
The frequency polygon
The mode is the point at which the frequency polygon is highest:
Properties
The mode is
• not always unique
• unstable
• the only appropriate measure of central tendency for nominal data.

{24, 34, 34, 48, 60, 73, 75, 77, 80, 81, 86, 86, 92, 96}
Properties
The mode is
• not always unique.
• unstable.
• the only appropriate measure of central tendency for nominal data.

Adding only a single observation to a sample can drastically change


the value of the mode.

{24, 34, 34, 48, 60, 73, 75, 77, 80, 81, 86, 86, 92, 96}

{24, 34, 34, 34, 48, 60, 73, 75, 77, 80, 81, 86, 86, 92, 96}
Mo = 34

{24, 34, 34, 48, 60, 73, 75, 77, 80, 81, 86, 86, 86, 92, 96}
Mo = 86
Properties
The mode is
• not always unique.
• unstable.
• the only appropriate measure of central tendency for nominal data.

Nominal data
• only identifies classes and
• does not imply magnitude so
• scores can’t be ordered.
The Median
Definition: the median is the 50th percentile, which can be computed
using the techniques we discussed last week.

Howell: Median location=(N+1)/2

Example: {24, 34, 34, 36, 48, 60, 73, 75, 77, 80, 81, 82, 92, 96}

N = 14
(N +1)/2 = 7.5
X7 = 73
X8 = 75
Md = 74
Properties
The median is

• very stable,

• not necessarily unique (like any other percentile),

• appropriate for ordinal or higher measurements, but

• mathematically difficult to work with.


The Arithmetic Mean
The arithmetic mean X (the average)

• Most common way to measure central tendency

N
1
X
N
X
i 1
i
The Arithmetic Mean
{4, 5, 6, 6, 7, 7, 7, 7, 8, 8, 9, 10}
1 12
X   Xi
12 i 1
4  5  6  6  7  7  7  7  8  8  9  10

12
84

12
7
The frequency distribution of this sample is symmetric and unimodal,
which means
Mo = Md = X ,
the mode, median and mean are all equal.
Properties of the Mean
The mean

• can be used for ordinal and higher scale data,

• is very sensitive to extreme scores (unstable), and

• is the geometric “balancing point” of a distribution.


Sensitivity to extreme scores
{4, 5, 6, 6, 7, 7, 7, 7, 8, 8, 9, 10}
X 7
Md = 7

{4, 5, 6, 6, 7, 7, 7, 7, 8, 8, 9, 50}
X  10.33
Md = 7

• In the presence of extreme scores,.


Summation Rules
• Rule 1: The Sum of a Variable.
n

V  V
i 1
i 1  V 2  ...  Vn

• Rule 2: The Sum of a Constant.


n

 c  c  c  ...  c  nc
i 1

• Rule 3: The Sum of a Product of a Constant and a Variable.


n n

 cV  cV
i 1
i 1  cV 2  ...  cVn  c(V 1  ...  Vn )  c Vi
i 1
and
n
Vi 1 n

i 1 c
  Vi
c i 1
• Rule 4: Distribution of Summation.
n

 (V  W ) (V
i 1
i i  W 1)  (V 2  W 2)  ...  (Vn  Wn)
1

 (V 1  ...  Vn)  (W 1  ...  Wn )


n n
  Vi   Wi, and
i 1 i 1
n n n

 (V  W )  V W
i 1
i i
i 1
i
i 1
i
The Balance Point
{2, 3, 3, 4, 4, 4, 5, 5, 6}

1 9 1 36
X   Xi  (2  3  3  ...  6)  4
9 i 1 9 9
The Balance Point
{1, 1, 2, 2, 2, 3, 3, 6, 7}

1 9 1 27
X   Xi  (1  1  2  ...  7)  3
9 i 1 9 9
Proof that the Mean is the Balance Point
Definition: For a sample of scores {X1, X2,…XN}, the deviation scores
{D1, D2,…DN} are all the differences between the scores (X) and their
mean ( X ).

Di  Xi  X
Proof that the Mean is the Balance Point
For the mean to “balance” the scores, for every – Di there must be
some positive Djs that zero it out.
Proof that the Mean is the Balance Point
For the mean to “balance” the scores, for every – Di there must be
some positive Djs that zero it out. That is,

D  0
i 1
i
Proof
N N

 D  ( X  X )
i 1
i
i 1
i

N N
  Xi   X
i 1 i 1

N
  Xi  NX
i 1

N N
1
  Xi  N X i
i 1 N i 1
N N
  Xi  Xi
i 1 i 1

0
The Relationship Between the Mode, Median, and Mean
Where the mode, median and mean are located relative to each
other depends on the shape of the frequency distribution.

• Symmetric: X = Md
– Unimodal: X = Md = Mo

• Skewed:
– Negatively: X < Md < Mo
– Positively: Mo < Md < X
Negatively Skewed
Positively Skewed
Bimodal
• Measures of central tendency
– Mode
– Median
– Arithmetic mean
• Properties of each statistic
• Summation notation
• Relationships

You might also like