01_INTRODUCTION.ppt

DW200705 Introduction 1

INTRODUCTION OF INTRODUCTION OF STATISTICSSTATISTICS

DW200705 Introduction 2

STATISTICSSTATISTICS

The Data it self

The science which study about the collection, organization, analysis, and made conclusions of data in the numerical forms.

DW200705 Introduction 3

DATA QUALITYDATA QUALITY

Data has an important role, as in the term GIGO – garbage in garbage out .Geological hypothesis: (a) Is there a significance difference in composition between granite from two intrusion bodies? (b) Is there a regular cyclicities properties in a sequence of limestone thin layers? © What can we estimate the permeability of sandstone layers from a well bore log data (wireline logging)?

DW200705 Introduction 4

PRECISION and ACURATIONPRECISION and ACURATION

PRECISION: a measure is called precise if we repeat it on a object will give the same value.

ACCURATION: a measure is called accurate if their results close to a right given value.

There are a combination form between low/high precision with low/high accuration .

DW200705 Introduction 5

PRECISION and ACCURATIONPRECISION and ACCURATION

Precision

High Low

Accuracy

(suppose true value is 50)

High 49, 50, 50, 52, 50, 49, 51, 50

55, 47, 50, 52, 44, 53, 57, 47

Low 54, 55, 55, 57, 55, 54, 56, 55

60, 52, 55, 57, 49, 58, 62, 52

DW200705 Introduction 6

TYPES OF GEOLOGICAL DATA (1)TYPES OF GEOLOGICAL DATA (1)

RATIO SCALE DATA

The ratio scale data is a common measure such as length, weight of matter, it is a good and without problem data.

INTERVAL SCALE DATA

Zero point is not data termination, such as the temperature in Celsius or Fahrenheit degrees, the zero point not an end, however the Kelvin’s degree is a ratio scale data .

DW200705 Introduction 7

TYPE OF GEOLOGICAL DATA (2)TYPE OF GEOLOGICAL DATA (2)

CLOSED DATA

These data in the form of percentages or ppm (part per million), or other form which indicates a proportion to a fixed value. It must be careful in bivariates or multivariate analyses.

DIRECTIONAL DATA

The data are expressed in the form of angle, for example azimuth, it need special process because 0 o = 360 o

DW200705 Introduction 8

TYPE GEOLOGICAL DATA (3)TYPE GEOLOGICAL DATA (3)

ORDINAL SCALE DATA

It is a poor quality data, interval scale is not regular and just shows the rank position then it can not be used for added, subtracted or dividing processes. For example the Mohs’s hardness scale, Richter’s earthquake intensity scale that based on the degree of damages. Analyzing of ordinal data generally use the non-parametric statistics

which analyze the data based on rank order.

DISCRETE DATA

All those above data has continue properties. The discrete data, commonly in the form of integer, for examples: counting some objects such as the number of fossils in 1 cm2 of outcrop.

DW200705 Introduction 9

TYPE OF GEOLOGICAL DATA (4)TYPE OF GEOLOGICAL DATA (4)

NOMINAL OR CATEGORIAL DATA

These data are in the forms such as name, for example minerals, fossils which must be converted to numerical data. It was sometimes analysis in binary forms, e.q. if there is a fossil we note as 1, if there is no fossil we note as 0. Analysis commonly made by multivariable method.

DW200705 Introduction 10

RATIO SCALE

TYPE OF GEOLOGICAL DATA (5)TYPE OF GEOLOGICAL DATA (5)

1 2 3 4 5 6 7 8 9 10

-3 -2 -1 0 1 2 3 4 5 6 7

INTERVAL SCALE

DW200705 Introduction 11

TYPE OF GEOLOGICAL DATA (6)TYPE OF GEOLOGICAL DATA (6)

0% 10 20 30 40 50 60 70 80 90 100%

CLOSED

ORDINAL

1 2 3 4 5 6 7 8 9 10 11 12

DW200705 Introduction 12

TYPE OF GEOLOGICAL DATA (7)TYPE OF GEOLOGICAL DATA (7)

N

S

EW

360 o 0o

DIRECTIONAL

360 = 0 .. ?

DW200705 Introduction 13

TYPE OF ANALYSIS (1)

UNIVARIATE METHODS

Each variable analyzed in isolation, the data can be portrayed as a series of points along an appropriately scaled line

0 5 10 15 20 25

xxx x x x x xx

DW200705 Introduction 14

TYPE OF ANALYSIS (2)

BIVARIATE

METHODS

Two variables analyzed together. The two measurements made on one object give coordinates of a point in a two dimensional (2D) space, and a data set can be portrayed as a 2D scatter, for the purpose of investigating the relationship between the data points and/ or the relationship between the variables.

Y

X0 5 10 15

10

20

30

40

x

xx

x

x

x x

DW200705 Introduction 15

TYPE OF ANALYSIS (3)

TIME SERIES METHODS

Sequences of data in time (or space) can be in various forms. Some can be treated as simple bivariate data, with one variable happening to be time. Often, the situation is conceptualized as a continuously varying curve

1

2

3

4

X

0

1 5 10 15 20 25

Time

DW200705 Introduction 16

TYPE OF ANALYSIS (4)

10

40

30

20

30

40

20

40

30

30

SPATIAL ANALYSIS

Three (or four) variables analysed together, two (or three) of which are spatial coordinates; grid references or latitude/ longitude, with or without altitude or depth. The other variable is geological measurement of interest, and as regarded as varying continuously over the area.

DW200705 Introduction 17

TYPE OF ANALYSIS (5)TYPE OF ANALYSIS (5)

MULTIVARIATE ANALYSIS

General methods applicable to any number of variables analyzed simultaneously, and usually applied to more (often many more) then three variables. If these are m variables, the data may be imagined as points in an m-dimensional space. The prime objective is to reduce the dimensionality so that the shape of the data scatter can be viewed. Relationship between variable can also be investigated.

DW200705 Introduction 18

Sample and PopulationSample and Population

SAMPLE

Population

DW200705 Introduction 19

Sampling Strategies (1)Sampling Strategies (1)

RANDOM REGULAR

DW200705 Introduction 20

Sampling Strategies (2)Sampling Strategies (2)

CLUSTERED UNIFORM

DW200705 Introduction 21

Sampling Strategies (3)Sampling Strategies (3)

TRAVERSE

DW200705 Introduction 22

THANK YOUTHANK YOU

DW200705 Introduction 23