1

02.1 Descriptive Statistics

ANNA MARIA LOURDES S. LATONIO, PhD

CENTRAL LUZON STATE UNIVERSITY

 2

STATISTICAL METHODS

Descriptive Inferential

Tabular Estimation

Graphical Tests of hypothesis

Numerical
 3

Descriptive Statistics
ETHOD
U L A R M
TAB HOD
AL MET
AP H IC THOD
Frequency GR ICAL ME
Bar graph NUMER
Percentage
Line graph Central Tendencies
Cumulative Dot/Scatter plot Measures of Variability

Crosstabs Circle graph Measures of Position

Boxplot Skewness
Stem and leaf plot Kurtosis
Describing the
Distribution
with Numbers
 MEASURES OF
CENTRAL TENDENCY
Mean, Median and Mode
 ASLatonio

The Mean
!"# $% &'' $(!)*+&,-$.!
• MEAN = ."#()* $% $(!)*+&,-$.!
5
1
Population Mean: 4 = 2 /-
5
-31
Sample Mean: .
1
/ = 2 /-
.
-31
 02. DESCRIBING DATA WITH NUMBERS 7

The Mean
Some properties of the Mean:
1. It always exists
2. It is unique
3. It reflects the magnitude of every observation
4. It is easily affected by extreme values
5. The mean of the subgroups can be combined into the
overall mean of all the data, called the weighted
mean.
 8

• MEDIAN
The Median is the midpoint of a distribution, the number such
that half the observations are smaller, and the other half are
larger.

Two ways of finding the median:

Order the data from lowest to highest

if n is odd, ! middle value in the array

"=
if n is even, ! average of the two middle
"=
values
 02. DESCRIBING DATA WITH NUMBERS 9

The Median
Some properties of the Median:
1. It is a positional value
2. Extreme values do not affect the median
as strongly as they do the mean.
 ASLatonio

The Mode
• The Mode is the most common observation in
a data set.

• Does the distribution have one or several

major peaks. Call modes?

• A distribution with one major peak is called

unimodal.
• If there are two major peaks, the distribution is
bimodal.
 ASLatonio

Example: Mean, Median and Mode

• Length of time (hrs) spent performing a specific task in
the laboratory by nine research assistants:
ID Name SEX Hours
1 Adrielle M 3.5
2 Marvin M 4.2
3 Arianne F 5.8
4 Sarah F 1.3
5 Kit M 4.1
6 Princess F 4.5
7 Angelee F 3.2
8 Barbz F 3.0
9 Valen F 4.7
 ASLatonio

Example: Mean, Median and Mode

ID Name SEX Hours
Mean 1 Adrielle M 3.5
2 Marvin M 4.2
" $" $ …$"' (.* $+.% $ …$+.,
! = # % = 3 Arianne F 5.8
' -
(+.( 4 Sarah F 1.3
= = 3.81 hrs
- 5 Kit M 4.1
6 Princess F 4.5
7 Angelee F 3.2
5
1 8 Barbz F 3.0

! = 1 62 9 Valen F 4.7
0
234
 ASLatonio

Example: Mean, Median and Mode

Median Hour
Order
s
Ordered list of observations:
1 1.3
Since n= 9 is odd, the center observation 2 3.0
is the 5th observation. The position of 3 3.2
the center observation is at 4 3.5
5 4.1
(n+1)/2 = (9+1)/2 = 5
6 4.2
The median length of hours spent performing 7 4.5
a specific task by the 9 research assistants
is 4.1 hours. 8 4.7
9 5.8
Suppose there are 10 research assistants in the list. The ordered
list will have 10 observations. The median will be
the average of the 5th and 6th observations in the ordered list.
 ASLatonio

Example: Mean, Median and Mode

• The Mode ID Name SEX Hours
1 Adrielle M 3.5
2 Marvin M 4.2
The length of time performing
the specific task has no mode, 3 Arianne F 5.8
since all the observations 4 Sarah F 1.3
has frequency equal to only 1. 5 Kit M 4.1
6 Princess F 4.5
With regards to the qualitative variable
7 Angelee F 3.2
Sex, the modal Sex is Female. 8 Barbz F 3.0
There are 6 female and 3 male 9 Valen F 4.7
research assistants.

Mode for Sex: Female (F)

 MEASURES OF VARIABILITY
Measures of variability describe the
spread or scatter of the observations in
the data.
Range
Variance
Standard Deviation
Standard Error
Coefficient of Variation
 ASLatonio

The Range

Range
= highest value – lowest value
= HV – LV
 ASLatonio

Variance
Variance is the mean of the squared
deviations of the observations from the mean
*
Population variance: 1
!" = & +' − + -
%
'()

0
-
1 -
Sample variance: . = & +' − +
/−1
'()
 ASLatonio

Computation formula
for Sample Variance
Sample variance (computation formula):

+ + "
1 1
!" = "
' ,( − ' ,(
%−1 %
()* ()*
 ASLatonio

Standard Deviation (SD)

The Standard Deviation is the positive

square root of the variance

! = !#
• Population SD:

• Sample SD: $ = %#
 20

Measures of Variability
Interpretation of Standard Deviation:
- the greater the value of the standard
deviation, the more the observations scatter
from the mean.

s = 2.3 On the average, the

observations deviate from the
mean by 2.3 units.
 ASLatonio

Standard Error
The Standard Error is the standard deviation
of the sample mean

Population SE: !
"

#
Sample SE: "
 ASLatonio

Coefficient of Variation
!"#$%#&% %'()#")*$
CV = ×-..%
+'#$
0
×100%
=1
 ASLatonio

Example: Range
ID Name SEX Hours
Range 1 Adrielle M 3.5
2 Marvin M 4.2
R = HV – LV 3 Arianne F HV =
= 5.8 – 1.3 5.8
= 4.5 hrs 4 Sarah F LV =
1.3
5 Kit M 4.1
The length of hours spent performing a 6 Princess F 4.5
specific task ranges from 1.3 to 5.8 hrs. 7 Angelee F 3.2
8 Barbz F 3.0
9 Valen F 4.7
 ASLatonio
ID Hours (X i) X i2

Example: Sample Variance

1 3.5 12.25

Sample Variance 2 4.2 17.64

+ + "
1 1 3 5.8 33.64
!" = "
' ,( − ' ,(
%−1 %
()* ()*
+
4 1.3 1.69
+

' , ( " = 143.61 ' , ( = 34.3 16.81

%=9 5 4.1
()* ()*

1 1 6 4.5 20.25
!" = 143.61 − 34.3 "
9−1 9
1 1 7 3.2 10.24
= 143.61 − 1176.49
8 9
8 3.0 9.00
1
= 143.61 − 130.72
8
9 4.7 22.09
1
= 12.89
8
= 1.61 squared hrs Total 34.3 143.61
 ASLatonio

Example: Sample Standard Deviation

ID Name SEX Hours
Sample SD 1 Adrielle M 3.5

! = !# 2 Marvin M 4.2
3 Arianne F 5.8
= 1.61 4 Sarah F 1.3

= 1.27 ℎ* 5 Kit M 4.1

6 Princess F 4.5
7 Angelee F 3.2

It is estimated that the length of time 8 Barbz F 3.0

performing a specific task by a 9 Valen F 4.7
research assistant on the average
deviates from the true mean by 1.27
hours.
 ASLatonio

Example: Standard Error (SE)

ID Name SEX Hours
SE 1 Adrielle M 3.5
"
s.e. = #
2 Marvin M 4.2
3 Arianne F 5.8
$.&'
= (
4 Sarah F 1.3
5 Kit M 4.1
$.&' 6 Princess F 4.5
= ) 7 Angelee F 3.2
8 Barbz F 3.0
= 0.42 hr 9 Valen F 4.7

It is estimated that the mean length of

time performing a specific task for
samples of size 9 research assistants
on the average deviates from the true
mean by 0.42 hours.
 ASLatonio

Example: Coefficient of Variation

ID Name SEX Hours
Coefficient of Variation 1 Adrielle M 3.5
$ 2 Marvin M 4.2
!" = ×100%
% 3 Arianne F 5.8
4 Sarah F 1.3
1.27 5 Kit M 4.1
= ×100%
3.81 6 Princess F 4.5
7 Angelee F 3.2
= 33.33% 8 Barbz F 3.0
9 Valen F 4.7

Using a sample size of 9, it is estimated

that the standard deviation of the
observations is 33.33% of the value of the
true mean length of time performing a
specific task.
MEASURES OF POSITION

Percentiles
Deciles
Quartiles
 29

Measures of Position
Percentile – is a value below which a
certain percentage of observations lie.

The 25th percentile of a set of data is

the value below which 25 % of the data
lie.
 30

Measures of Position

Percentile

The position p of the observation which

defines the pth percentile in a set of data
consisting of n observations:

!(#$%)
Position of the Pth percentile =
%''
 31

Example: the position p of the 25th

percentile in an array consisting of n = 22
observations is at …
th !" ( !!$%)
Position of the 25 percentile = %''
= 5.75

The 25th percentile is in the 5.75th position.

• Arrange first the data from lowest to highest. Assign X 1 as the lowest
value and X 22 as the highest value. Identify the 5 th and 6 th values to
solve for the value of the 25 th percentile as follows.

The 25th percentile = X5 + 0.75(X6 - X5 ) = _______

 32

Measures of Position
Decile

First Decile D1 = P10

Second Decile D2 = P20
:
:
Ninth Decile D9 = P90
 33

Measures of Position
Quartile

First Quartile Q1 = P25

Second Quartile Q2 = P50
Third Quartile Q3 = P75
 34

Interquartile Range

is the length of the interval between

P25 and P75 and describes the range of
the middle half of the distribution.
FURTHER DESCRIPTIONS

Measure of Skewness
Measure of Kurtosis
 36

Further descriptions

Skewness is a measure of asymmetry of

a distribution.

Kurtosis is a measure of the degree of

peakness or flatness of a
distribution.
 37

Skewness
Mode Mean
Median
Mean Mode
Mean Median
Median

Coefficient of Skewness Mode

'
Cs @ 0, approaches normality
! ∑ #$ %& Cs > 0, positively skewed
Cs = !%( (!%*), ' Cs < 0, negatively skewed

Pearson Coefficient of skewness

. Sk @ 0, approaches normality
-(&%&) Sk > 0, positively skewed
Sk = , Sk < 0, negatively skewed
 38

Kurtosis
– is a measure of a distribution’s peakedness
(or flatness)

Coefficient of Kurtosis
C k @ 0, mesokurtic
!" ∑ $% &' (
Ck = !&) (!&+)(!&-). (
−3 C k > 0, leptokurtic
C k < 0, platykurtic

Leptokurtic

Mesokurtic
platykurtic
 39
DEPARTMENT OF
STATISTICS

En j oy
d y i n g
Stu
t i s t i c s J
Sta