Docstoc

GIS Methods for Quantitative and Qualitative Analysis

Document Sample
GIS Methods for Quantitative and Qualitative Analysis Powered By Docstoc
					              Overview
• Research in geography
• Types of geographic research questions
• Quantitative versus qualitative research
  approaches
• Using GIS for research in geography
Why do we do research in
     geography?
      Research for exploration
• Investigation of little-understood
  phenomena
• Identification of important variables
• Generation of questions for further research
      Research for description
• To describe and characterize patterns and
  phenomena
• To document patterns and phenomena of
  interest
   Research for explanation and
          understanding
• To explain what caused an observed pattern
  or phenomena
• To explain why a pattern or phenomena is
  characterized the way it is
• To understand processes and interactions
  between people, places, and phenomena
       Research for prediction
• To predict future patterns or outcomes for
  different phenomena
• To forecast events and behaviors resulting
  from different phenomena
Types of research questions in
          geography
•   Exploratory questions: learn about a new issue
•   Descriptive questions: describe a phenomenon
•   Explanatory questions: explain a phenomenon
•   Predictive questions: predicting future patterns
•   Research studies often include two or more types
    of research questions
  GIS can be used to answer all
   types of research questions:
• Exploratory: Is there a spatial pattern?
• Descriptive: Has the pattern changed over
  time?
• Explanatory: What caused a pattern to
  change?
• Predictive: What do we expect the pattern to
  look like in the future?
    GIS can be applied in both
quantitative and qualitative research
               studies
Quantitative research approaches
• Application of numerical analytical
  techniques to address geographic research
  questions (of all types)
• Defined as the collection, classification,
  presentation, and analysis of numerical data
 Qualitative research approaches
• Use non-numerical information (e.g.
  conversations, artifacts, visual images)
• Entails a wide range of approaches such as
  unstructured interviews, ethnography,
  content analysis
• Shared belief in grounded theory (generate
  theory from information that the researcher
  collects)
  All quantitative analysis is based on
qualitative judgements (e.g. quantitative
   survey of quality of life in Oslo)
• Did the respondent understand the question?
• Did the respondent understand your answer
  scheme (e.g. 1 = agree; 2 = disagree) ?
• Did the respondent realize the questions
  were only about living in Oslo?
• Was the respondent answering honestly or
  just randomly?
All qualitative data can be measured and
 coded using quantitative methods (e.g.
unstructured interviews about quality of
                life in Oslo
• Code responses in an open-ended interview with
  numbers that refer to data specific references
• For example, code the factors that people see as
  reflecting high quality of life (e.g., bars, skiing)
• Quantitative research can therefore be generated
  from qualitative inquiries.
 What’s the real difference between
       research approaches?
• The major difference between qualitative and
  quantitative research stems from the researcher’s
  underlying strategies.
• Quantitative research is viewed as confirmatory
  and deductive in nature (use data to test theories)
• Qualitative research is considered to be
  exploratory and inductive (gather the data and
  learn what’s happening from the data and then
  generate theories)
 Questions to consider for qualitative
    versus quantifative research
• Is your aim the generation of new theories or
  hypothesis?
• Do you need to obtain a deep understanding of an
  issue? Is the issue too complex to quantify? (what
  does it mean to be poor in Oslo today?)
• Are you willing to trade detail for generalizability?
  (e.g., someone’s experience of poverty vs. a
  quantifiable measure of income levels)
      GIS is a tool for all types of
    research questions and research
              approaches
•   GIS and quantitative, descriptive analysis
•   GIS and quantitative, explanatory analysis
•   GIS and qualitative, descriptive analysis
•   GIS and qualitative exploratory analysis
   GIS and Quantitative Analysis:
  Vulnerability to climate change and
economic changes in Indian agriculture
                      Context
• Agriculture in India
  – 27 % GDP
  – 700 million people
  – more than 60 % is rainfed cultivation


• Both climate change and economic globalization
  are ongoing processes with uneven impacts. Indian
  agriculture will be confronted by both processes
  simultaneously, leading to changing patterns of
  vulnerability.
             Main objectives
• Assess vulnerability of agriculture to climate
  change in the context of economic changes
• Use GIS to identify highly vulnerable areas and
  social groups
• Interview farmers in highly vulnerable areas to
  understand how farmers are coping with climatic
  and economic changes
              Methodology
• GIS-based vulnerability profile
• Village-level case studies
• Integration of macro- and micro- scale
  analyses
Globalization vulnerability
Climate Change Vulnerability
   Double Exposure: Areas that are
Vulnerable to both Climate Change and
             Globalization
                     Case study approach
• Questionnaire-based survey
   –   Economic status
   –   Agricultural practices
   –   Coping mechanisms
   –   Access to facilities (electricity,
       irrigation, health, education,
       loans, etc)

                                            • Participatory rural appraisals
                                            • Focus group discussions with
                                            small and marginal farmers
                                            • One-to-one meetings with
                                            village heads and district
                                            administrative officers
  GIS and Quantitative Methods
• GIS can also be combined with statistical
  techniques such as correlation and regression
• Correlation: are observations correlated across
  space (e.g. do high income counties cluster
  together)
• Regression analysis: incorporate correlation across
  space into a spatial regression model
     GIS and Quantitative
    Explanatory Analysis:
Accounting for Income Variation
on American Indian Tribal Areas
  Rural Poverty and Tribal Areas
• Persistent poverty is an enduring problem for rural
  areas
• In the US, persistent rural poverty is especially
  evident on tribal lands
• Tribal lands also tend to be located in very
  “marginal places”
• Cross-sectional economic and geographic
  literature has paid relatively little attention to tribal
  areas (despite many case studies)
          Research Questions
1. Are there significant differences in per
  capita income levels between tribal and
  non-tribal areas, after controlling for
  locational and other characteristics?

2. Across tribal spatial areas, what accounts
  for income level variation?
          Tribal Counties in the
              United States




       Reservation and Trust Area
                                                 N
       No tribal area
       OTSA-TDSA area                       W        E
1000            0            1000   2000 Miles
                                                 S
        Per Capita Income (1999)




       Per Capita Income (1999)
            4896 - 19382
            19383 - 24423
            24424 - 33398                                 N
            33399 - 81665
                                                      W       E
1000                0             1000   2000 Miles
                                                          S
         Per Capita Income (1999)
        and high AI share counties




            High AI-share
       Per Capita Income (1999)
            4896 - 19382
            19383 - 24423                                 N
            24424 - 33398
            33399 - 81665                             W       E
1000                 0            1000   2000 Miles
                                                          S
What might explain variation in
   spatial income levels?
• Locational factors: proximity to urban areas, cost
  of living, natural amenities, transport
  infrastructure
• Structural factors: industry structure (shares in
  manufacturing, agriculture/resources, federal gov)
  unemployment rate
• Individual factors: educational attainment, age-
  structure of the population
• Tribal and social capital: AI population share,
  presence collective economic activity (gaming),
  type of tribal area (presence of tribal land base)
              Key Findings
• Locational, Structural and Individual factors
  all matter in accounting for income
  variation across all county groupings,
  including all tribal types of tribal counties
• Consistent factors include market size,
  unemployment, educational attainment, and
  shares of retirement-age population
              Key Findings
• Tribal counties do not have significantly
  different income levels than other counties
  (once locational and other factors are
  controlled)
• But, tribal and nontribal counties with high
  shares of American Indians do have
  significantly lower incomes than other
  counties (even controlling for other factors)
  GIS and Qualitiative Methods
• GIS has historically been applied primarily
  to quantitative questions
• Newest frontier in GIS research entails use
  of GIS in qualitative research
      Marianna Pavlovskaya,
     Professor of Geography
         (CUNY-Hunter)
2002 "Mapping urban change and
  changing GIS: Other views of
     economic restructuring,"
forthcoming in Gender, Place and
            Culture.
         Research Questions
• How did the end of communism affect
  people’s everyday lives?
• How did their participation in the economy
  change (changes in the nature of work)?
• How did their access to consumer goods
  changes?
   Mei-Po Kwan, Professor of
          Geography
    (Ohio State University)

Evaluating Gender Differences in
Individual Accessibility: A Study
Using Trip Data Collected by the
   Global Positioning System
         Research Questions
• How do typical daily travel patterns vary
  between men and women?
• What do these variations imply about
  employment opportunities and leisure time
  activities?
                  Summary
• GIS can be used to answer all types of research
  questions, including exploratory, descriptive and
  an explanatory questions
• GIS applies to both quantitative and qualitative
  work
• The choice of GIS techniques depends on the type
  of research questions that you are asking
• The research questions should always come
  first

				
DOCUMENT INFO
Shared By:
Categories:
Tags:
Stats:
views:32
posted:10/29/2011
language:English
pages:42