geographical statistics ge 2110 - The State University of Zanzibar

advertisement
GEOGRAPHICAL STATISTICS
GE 2110
Zakaria A. Khamis
Introduction to Geographical Statistics
• The study of geographical phenomena often requires the
application of statistical methods to produce new insight
• Methods of statistical analysis pay a central role in the
study of geographic problems
• What are the effects of global warming on the geographic
distribution of species?
• Such question make use of statistical analysis to arrive at
the conclusion
3/15/2016
Zakaria Khamis
2
• The role of statistical analysis in geography may be placed
within a broader context through its connection to the
scientific method
• GEOGRAPHY !!!! Is it art or science?
• Scientists (social and physical) make use of scientific
methods in their attempts to learn about the world
CONCEPTS
THEORY
3/15/2016
DESCRIPTION
LAWS
Zakaria Khamis
HYPOTHESIS
MODEL
3
• Laws are universal statements of unrestricted range
• Theories refer to collection of generalization (laws) 
Einstein called theories ‘free creation of the human mind’
• Statistical methods allow us to suggest and test hypothesis
using models
• Geographers use spatial analysis within the context of the
scientific method in at least two distinct ways
• Exploratory  methods of analysis are used to suggest
hypothesis
3/15/2016
Zakaria Khamis
4
• Confirmatory  used to help confirm hypothesis; however,
it does not always confirm or refute hypothesis
• STATISTICS  historically, it was a branch of political
science dealing with the collection, classification and
discussion of facts bearing on the condition of state or
community (Hammond and McCllagh 1978); e.g. in todays
situation is vital statistics
• McGrew and Monroe (2000) define statistics as the
collection, classification, presentation and analysis of
numerical data
3/15/2016
Zakaria Khamis
5
• Is a science of organizing, assembling and anlysing data as
well as drawing conclusion about what the data means.
• Is a science of observing, collecting, recording, summarizing,
analysing and presenting data in organized way.
• Modern defintions have in common the objective of inferring
from a sample of data the nature of a larger population from
which the sample was drawn
3/15/2016
Zakaria Khamis
6
Branches of Statistics
Descriptive: Aim to summarize and describe the data .It is a
straightforward presentation of facts in which a desicion have
been made.
Numerical (mean, mode, median etc) and graphic approch
(charts, garphs and etc) are the basic methods used in this
type.
Inferential: This type of statistic enables us to make
generalization or prediction about population from a sample.
The name implies, allow inference about a large population
from a sampe It uses confirmatory methods
3/15/2016
Zakaria Khamis
7
On the last 3 Sunday , henry sold 2, 1 and 0 new car
respectively
Descriptive: Henry everage 1 new car sold for the last 3
sundays
Inferential: Henry never sells more than 2 cars on Sunday
Although the inferential statement is true for the last 3
Sunday we do not know if it is true for all Sunday
3/15/2016
Zakaria Khamis
8
Data
• Is the collection of facts, they can be numbers, words,
measurements or description of things
• Data can be qualitatives or quantitatives
• Qualitative data is a descriptive information, they based on
observable quality but not measurable; example, of qualitative
data are marital status, gender, job title and etc.
• Quantitative data is a numerical information, they are expressed
by means of numbers. This include all data which we can take
their measurement; example, the height of objects, land
elevation, temperature, rainfall, number of residence and etc .
• Quantitative data can be discrete or continous
Disrete data
• This data can only take certain value (like the whole numbers)
• No possiblity of values between the adjacent data values;
example number of birth, number of death, number of
students .
• This include the data which we can count rather than take
their measurement, therefore they cannot be expressed in
decimal form.
Continuous data
• These are the data which can take any values (fractions, whole
number or decimal).
• There is a great possibility of values between adjacent values
number. For instance, if we record the time taken and distance
covered by a motor cycle travel from one station to another;
we can have 122km/122.6km covered and 2hrs and 30min or
4hrs.
• Example of Continous data: height, temperature,presure,
distance, areas, time and etc.
• Grouped v/s individual data Assignment
• Attribute data can be categorized into 4
– Nominal
– Ordinal
– Interval
– Ratio
Nominal
 These are variables that provide descriptive information
about an object.
 These are the observations that have been placed into a set
of mutually exclusive and collectively exhaustive categories
 They do not imply order, size, or quantitative information
E.g. city name, vegetation type code etc
Ordinal
 The values of these variable imply rank.
 An ordinal attribute may be descriptive such as small,
medium or large.
 They may be numeric. However, the number represent only
order and not magnitude, for we assume no linearity
between the values.
 It is possible to say that one observation is greater than
another; however, we cannot say by how much the
observation is greater than another
E.g. Soil erosion can be ranked 1 to 10 classes. Each class
show the order.
3/15/2016
Zakaria Khamis
13
Interval
 Are spread along a regularly spaced scale of measurement
units.
 Equal changes in interval involve equal changes in the
object being measured.
 In this scale zero is not meaningful  No absolute zero,
and consequently ratio interpretations are not possible
E.g. Temperature data which is not in Kelvin  440C is 12
degree warmer than 320C
But 440C is not twice as warm as 220C
Ratio
 These are also interval; however they contain an additional
property that the measurements are related to a meaningful
zero point.
 They always have meaningful zero  absolute zero.
E.g. Rainfall data – zero rain means no rain and 2inchs rain is
twice as 1inch rain.
Temperature measured in Kelvin  1000K is twice as warm
as 500K
Population data etc.
3/15/2016
Zakaria Khamis
15
Sources of data
•
•
•
•
Government and non government agencies
Religious agencies
Population
Physical environment
Primary souces
Is the one in which the person or organization is responsible
for data collection also uses and/ analysis.
Interview, questionnaire, focus group discussion, surveying,
photographing are some of the methods used (collection).
Secondary sources
It implies that the person or organization that is currently
using or analyzing the data was not involved in their
collection.
That is to say it comprises second hand reports that
contained data which have been condensed or summarizes
or statistically manipulated or modified from primary or
secondary sources.
Examples of secondary sources
• Books, such as biographies (not an autobiography), textbooks,
Encyclopedias, dictionaries, handbooks
• Articles, such as literature reviews, commentaries, research
articles in all subject disciplines
• Criticism of works of literature, art and music
Spatial data
 It is also known as coordinate/geometric data – define the location and
shape of the feature.
 The coordinate most often consists of a pair of numbers that specify location
in relation to an origin.
 Spatial data most often use a Cartesian Coordinate System. However, it
can also use Spherical Coordinate System.
 Spatial data can be discrete feature or Continuous feature
 Discrete feature may include – lake, building, well etc.
 Continuous feature may include – elevation, temperature, precipitation etc.
Download