COMPUTATIONAL

GEOPHYSICS
GEOP 205

A Small Dose of Statistics

Instructor
Dr. A. Abokhodair
E-MAIL: akwahab@[Link]
Ext. 2625 / 3851

3/31/2015 A. Abokhodair 1
KFUPM, Earth Sciences
 What is Statistics?
Statistics is the body of mathematics that allows you to analyze your
data, assess their ERRORS and acquire meaningful information from
them.

What is Error
 Difference between a measured value and the “True value”
 Uncertainty in measurements.

Types of Errors
1. Random = Result from human, instrumental, and
environmental inconsistencies
2. Systematic = Due to human or instrumental malfunctions.

3/31/2015 A. Abokhodair 2
KFUPM, Earth Sciences
 How to Present Data
RAW DATA:
Experimental data consists of a large volume of numbers which are
difficult to comprehend. The aim of data presentation methods is to
present the data in graphical image which illuminates major trends.

EXAMPLES

Scatter Diagram Frequency Histogram

Relative Frequency (f )
F(X)

X X

3/31/2015 A. Abokhodair 3
KFUPM, Earth Sciences
 The Probability Distribution
D x => 0
Frequency Probability
Distribution Distribution

Measure of Dispersion
P(X)

P(X)

Measure of Centrality
X X

3/31/2015 A. Abokhodair 4
KFUPM, Earth Sciences
 Statistics & Parameters
Measures of Central Tendency
• Mean = Center of weight of data
• Median = Middle of data
• Mode = Most common datum

Measures of Dispersion
• Variance = Average Squared Residual
• Range = Data Extent
• Quartiles = Quarter Range

3/31/2015 A. Abokhodair 5
KFUPM, Earth Sciences
 Distribution Statistics
Measures of Centrality Measures of Dispersion
 Mean  Variance
 Median  Range
 Mode  Quartiles

3/31/2015 A. Abokhodair 6
KFUPM, Earth Sciences
 Probability Distribution
•P(z) is a mapping between probability values and realization values of a random
variable.

If P(z) is a probability distribution Then

 P( z ) dz  1


P(z)
b
p  a  z  b   P( z ) dz
a z
a b

3/31/2015 A. Abokhodair 7
KFUPM, Earth Sciences
 Special Distributions

The Normal
  z   
2
1
P( z )  exp   
 2  2 2 
 

P(z)
The Uniform
 1
 , azb
P( z )   b  a z


a b
0 elsewhere

3/31/2015 A. Abokhodair 8
KFUPM, Earth Sciences
 Statistical Functions
The Expectation
Continuous Discrete

z  E (z )   z n p (z n )
E (z )  z  z

P (z )dz
n

Properties
E (c)  c
E ( a x  b)  a E ( x )  b  a   b
E ( x  y )  E ( x)  E ( y )  x   y
E ( x y )  E ( x) E ( y )  x  y x,y Independent
3/31/2015 A. Abokhodair 9
KFUPM, Earth Sciences
 Statistical Functions
The Variance
V ar (z )  s  E  z       z    p (z )dz
 
2 2 2
z
 
Continuous z

N
V ar (z )  s    z n  z 
2
Discrete 2
z p (z n )
n 1

Properties
var(c )  0
var( a x  b)  a 2 x2
var(ax  by )  a 2 x2  b2 y2 x,y Independent

3/31/2015 A. Abokhodair 10
KFUPM, Earth Sciences