STATISTICS AND GEOGRAPHY

DEFINITION: STATISTICS
Statistics is the methodology used in studies that
collect, organize, and summarize data through
graphical and numerical methods, analyze the data,
and ultimately draw conclusions.
1. To help summarize the findings of some inquiry, for example, a study
of the travel behavior of elderly or handicapped citizens or the
estimation of timber reforestation requirements.
2. To obtain a better understanding of the phenomenon under study,
primarily as an aid in generalization or theory validation, for
example, to validate a theory of urban land rent.
3. To make a forecast of some variable, for example, short-term
interest rates, voter behavior, or house prices.
4. To evaluate the performance of some program, for example, a
particular form of diet, or an innovative medical or educational
program or reform.
5. To help select a course of action among a set of possible
alternatives, or to plan some system, for example, school locations.
 Descriptive statistics and Inferential statistics.
Descriptive statistics deals with the organization and
summary of data. The purpose of descriptive
statistics is to replace what may be an extremely
large set of numbers in some dataset with a smaller
number of summarymeasures.
• Inferential statistics, Descriptive
statistics is linked with probability theory so that an
investigator can generalize the results of a study of
a few individuals to some larger group. To clarify
this process, it is necessary to introduce a few simple
definitions. The set of persons, regions, areas, or
objects in which a researcher has an interest is
known as the population for the study
Statistical Analysis and Geography
The Role of Statistics in Contemporary
Geography :

1. statistics
is an important component of the research methodology of
virtually all systematic branches of geography. A substantial
portion of the research in physical, urban, and economic
geography employs increasingly sophisticated statistical
analysis. Being able to properly evaluate the contributions of
this research requires us to have a reasonable
understanding of statistical methodologies.

2. Maps are inherently

mathematical and statistical objects, and as such they
represent one area of geography where dramatic change
will continue to take place for some time to come. This
trend has forced many geographers to acquire better
technical backgrounds in mathematics and computer
science, and has opened the door to the increased use of
statistical and quantitative methods in cartography.
Geographic information systems (GIS). Large sets of data
are now stored, accessed, compiled, and subjected to
various cartographic display techniques using video display
terminals and hard-copy devices.
3. Many of the fundamental problems of
displaying data on maps have clear and
unquestionable parallels to the problems of
summarizing data through conventional
descriptive statistics.

4. statistical methods find numerous

applications in applied geography.
Retail location problems, transportation
forecasting, and environmental impact
assessment are three examples of applied
fields where statistical and quantitative
techniques play a prominent role.
 DEFINITION: STATISTICAL POPULATION
 A statistical population is the total set of elements
(objects, persons, regions, neighborhoods, rivers,
etc.) under examination in a particular study.
 For instance, if a geographer is studying farm
practices in a particular region, the relevant
population consists of all farms in the region on a
certain date or within a certain time period
• DEFINITION: POPULATION CHARACTERISTIC
• A population characteristic is any measurable attribute
of an element in the population.

• A fluvial geomorphologist studying stream flow in a

watershed may be interested in a number of different
measurable properties of these streams. Stream
velocity, discharge, sediment load, and many other
characteristic channel data may be collected during a
field study. Since a population characteristic usually
takes on different values for different elements of the
population, it is usually called a variable.
 DEFINITION: VARIABLE
 A variable is a population characteristic that takes
on different values for the elements comprising the
population
 DEFINITION: POPULATION CENSUS
 A population census is a complete tabulation of the
relevant population characteristic for all elements in
the population
 DEFINITION: INTERNAL DATA

Data Data available from existing records or

files of an institution undertaking a study
are data from an internal source.

primary and
secondary
external data.

DEFINITION: DEFINITION:
SECONDARY PRIMARY DATA
DATA Primary data are
Secondary data obtained from the
are data obtained organization or
from a source institution that
other than the originally
primary data collected the
source. information.

If you must use external data, always use the primary source. The
difficulty with secondary sources is that they may contain data altered by
recording or editing errors, selective data omission, rounding,
aggregation, questionable merging of datasets from different sources, or
various ad hoc corrections. For example, never use an encyclopedia
to get a list of the 10 largest cities in the United States
 DEFINITION:
EXPERIMENTAL METHOD
OF DATA COLLECTION
An experimental method of
data acquisition is one in
DEFINITION: which some of the factors
NONEXPERIMENTAL under consideration are
METHOD OF DATA controlled in order to isolate
COLLECTION their effects on the variable
A nonexperimental or variables of interest
method of data
collection or statistical Only in physical geography
survey is one in which is this method of data
no control is exercised collection prominent. Fluvial
over the factors that may geomorphologists,
affect the population for example, may use a
characteristic flume to control such
of interest. variables as stream
velocity, discharge, bed
characteristics, and
gradient. Among the social
sciences, the
largest proportion of
experimental data is
collected in psychology.
 Metadata

 Metadata or simply “data about data” provide information about the content, quality, type,
dates of creation, and usage of the data. Metadata are useful for any information source
including pictures or videos, web pages, artifacts in a museum, and of course statistical data.
For a picture, for example, we might wish to know details about the exact location where it was
taken, the date it was taken, who took the picture, deta

A metadata document includes several common

In general, the
components:
goal of roviding
1. Data definitions
metadata is to
2. Method of sampling.
facilitate the
3. Data quality
use and
4. Data dissemination and legal issues
understanding
5. Lists of studies based on the data
of data
6. Geographic data
7. Training.
 Characteristics of Datasets
1. First type, sometimes known
as a structural matrix, is shown
in Table 1-1. In this example,
the observational units are
DEFINITION: DATASET
climatic stations. Five variables
A dataset is a collection of Geographers collect or analyze are contained within
statistical information or two typical forms of datasets. the dataset. The information on
values of the variables of
every variable from one
interest in a study.
observational unit is often
termed an observation; it is also
common to speak of the data
value for a single variable as an
observation since it is the
observed value. In this case,
the rows of the
dataset represent different
locations, and the columns
represent the different variables
available for analysis. These
places might represent areas or
simply fixed locations such as
cities and towns. Such a matrix
allows us to examine the spatial
structural matrix variation or spatial structure of
these individual variables.
2. an interaction matrix,
in which the variable of interest is
expressed as the flow or interaction
between various locations (A through G),
Qualitative variables are sometimes termed categorical
variables since the observational units can be placed into
categories. Male/female, land-use type, occupation, and plant
species are all examples of qualitative variables
Measurement Evaluation

The utility of any statistical analysis

ultimately rests on the quality of the
data used, regardless of how
sophisticated the analysis itself is.
For example, suppose our data suggest
that men travel 10 minutes longer than
females on the journey to work. On the
face of it, this appears to be important.

It is therefore essential to rigorously

evaluate the quality of our data with
respect to several key principles—before
we undertake any statistical analysis.
These principles include measurement
validity, accuracy, and precision.
 Validity
DEFINITION: MEASUREMENT VALIDITY
Measurement validity refers to the degree
of correspondence between the concept
being addressed and the variable being
used to measure that concept.

There are, however, instances in both

physical and social sciences where the
situation is less clear, and measurement
validity is questionable

.In the first case, it may be that the concept

being studied is imperfectly defined. To
pick just a few examples, we could mention
intelligence, social status, drought, and
ecosystem diversity.
Accuracy
DEFINITION: MEASUREMENT ACCURACY
Measurement accuracy refers to the absence of error, or the degree of
agreement between a measured and true value.

the total error of measurement divide

into two components: systematic and random error.
Total error = Systematic error + Random error

The first component, systematic The second component is error that is not
error, arises if the instrument attributable to poor calibration of the
consistently gives high or low measurement instrument leading to
values. systematic error, but appears to be random
or unpredictable in nature.
Precision

We can see that thermometer A has no

systematic error and leads to readings that
may be above or below the true
temperature within a range of random error
of about one-half a degree.

Thermometer B has a systematic error of

about +1° since the average reading
seems to be around 21°, but the
measurements also vary within about one-
half a degree. We would say that both of
these thermometers are equally precise,
but thermometer A is more accurate.
Consider now thermometers C and
D. Like thermometer A, thermometer
C has no bias, but it is much less
accurate than A.

The problem with C is its low

precision, since readings from the
thermometer vary between roughly
18° and 20°. This is almost four
times less precise than A.

Thermometer D is biased but has a

small random error, leading to an
accuracy higher than the unbiased
thermometer C.

In the search for high overall

accuracy, one must consider both
sources of error.
 1.4. Data and Information

Data: specific observations of

measured numbers

DEFINITION: INFORMATION
Data that have been processed
into a meaningful form, one that
has value to the user, have been
transformed into information.

Statistical analysis can thus be

viewed as the link between data and
information.
1.5. Summary
Statistical analysis includes methods used
to collect, organize, present, and analyze
data.

Descriptive statistics refers to techniques

used to describe data, either numerically or
graphically.

Inferential statistics includes methods used

to make statements about a population
characteristic on the basis of information
from a sample. Statistical inference
includes both hypothesis testing and
estimation methods.