Role of Statistics in
Geography
What Is Geography?
1. Attempt to describe, explain and
predict spatial patterns and activities
2. How and why do things differ from
place to place?
3. How do spatial patterns change
through time?
How Do Geographers Approach
Discipline
1. Positivism- objectivity of scientific analysis
and testing hypotheses to build knowledge
and understanding
2. Humanistic- people create subjective worlds
in their minds- behavior understood only by a
methodology that penetrates the subjectivity
3. Structuralists- cannot explain observed
pattern by examining pattern itself. But rather
establish theories to explain development of
societal conditions within which people must
act
Role of Statistics
Room in all the above interpretations for
quantitative analysis.
But increasingly both quantitative and
qualitative analysis are important
Qualitative analysis involves?
Statistics and measurement are used
commonly in our lives
A. Making home purchase decisions
B. Setting up investments
C. Weather variations are expressed as
probabilities
How Do Geographers Use Statistics?
1. Describe and summarize data
2. Make generalizations concerning
complex spatial patterns
3. Estimate likelihoods of outcomes for
events at particular location(s)
4. Use sample data to make inferences
about a larger set of data (a population)
5. Learn whether actual pattern matches
an expected or theoretical
6. Wish to compare or associate (correlate)
patterns of distributions
Formulating the Research Process
1. Problem Identification
2. Develop Questions to Investigate
3. Collect and Prepare Data
4. Process descriptive data (maps,
graphics)>>>>> Reach conclusions
5. Formulate Hypothesis >>>>> Collect
and Prepare Sample Data
6. Test Hypothesis>>Evaluate Hypothesis
7. Develop Model, Law, or Theory
What Are Models?
Abstractions of the real world
Simplified versions of reality
Easier to examine scaled down and simplified
structures in attempt to understand
Iconic models- look like what they represent (
Analogue models- one property used to
represent another
Symbolic models- equations
Basic Terms and Concepts
Data element- basic element of information
which we measure
Data Set- groups of data (commuting sheds
of industries)
Observations-Cases-Individuals- elements
of phenomena under study
Variable- property or characteristics of each
observation that can be measured,
classified or counted
Values may vary among set of
observations: rainfall, per capita income,
years of schooling
Geographic Data
1. What sources of data are
available?
2. Which methods of data collections
should be used?
3. What type of data will be collected
and then analyzed statistically?
Types of Data
Primary Data- acquired directly from
original source
1. Information collected in the field
2. Usually very time consuming
3. Involves decision about a sample
design so representative data may
be obtained
Types of Data
Secondary Data (or Archival Data)
1. Usually collected by some organization
(United Nations, U S Bureau of Census)
2. Often easily accessible- hardcopy or CD rom
3. Less time consuming but also more limiting
4. Often need to inspect historical records and
archives for diaries, oral histories, official
reports in order to develop a picture of problem
Characteristics of Data
1. Some data are explicitly spatiallocations are directly analyzed
2. Other data implicitly spatial- data
represents places but locations
themselves are not analyzed
(population sizes of towns)
Measurement Concepts
1.Precision- level of exactness associated
with measurement (rain gauge to inches or
fractions of inches)
2. Accuracy- extent of system wide bias in
measurement process
3. Validity- if geographical concept is complex
expressing true or appropriate meaning
of the concept through measurement may be
difficult (levels of poverty, economic well
being, environmental quality)
4. Reliability- changes in spatial patterns are
analyzed over time must ask about
consistency and stability of data
Types of Statistical Analysis
Descriptive Statistics- concise numerical or
quantitative summaries of the
characteristics of a variable or data set
(e.g. mean, standard deviation, etc)
Inferential Statistics- here we wish to make
generalizations about a statistical
population (total set of information or data
under investigation) based on the
information from a sample
Sample- typical or representative or
unbiased subset of the broader, larger
more complete statistical population
