GEOG 5023/5033
Quantitative Methods in Geography
Recent
Additions
(non)stationary random
field movies
simple kriging lecture notes
(note: this lecture was first given by Phaedon Kyriakidis at UC Santa
Barbara).
Nicholas Nagle
Office: Guggenheim 207
email:
nicholas.nagle@colorado.edu
Phone: 2-4794
Spring 2007
Lecture: M/W/F 9:00-9:50
Lab: T: 2:00-4:50 Click here for the lab
webpage.
Office Hours: Wednesday - 10-12:00 or
by appointment
Course Webpage:
http://www.colorado.edu/geography/class_homepages/geog_4023_s07
Teaching Assistant:
Bryan Jones:
bryan.jones@colorado.edu
Prerequisites: An introductory statistics course. It is assumed that you are familiar with the fundamentals of hypothesis testing, particularly t-tests. If you have taken an introductory statistics course but can not recall this material, please see the instructor.
Description:
This
course satisfies the requirement in quantitative methods for MA and
PhD students in Geography. This
class and its associated lab are grounded in both theory and
practical experience for the description, analysis and description of
spatial and other data frequently encontered in geographical
research. The purpose of developing theoretical knowledge in these
subjects is to
enable students to competently appraise the the scientific claims
made by other researchers (as well as yourself) when quantitive
methods are used. An underlying theme in this class is that the
choice of methods is not automatic, and that once a method has been
chosen, constraints have been placed on the nature of the scientific
process, many of which may seem odd or unnatural for geographic
phenomena.
Required
Textbook: Bailey, Trevor C.,
and Anthony C. Gatrell. (1995). Interactive Spatial Data
Analysis. Prentice-Hall.
Other required readings
will be periodically provided on the course webpage.
Grading:
Mid-Term Exam: 40%
Final Exam: 40%
Paper Review: 20%
Paper Review: Review a academic paper of your choosing. The article must containt a sufficient element of statistical analysis. It does not need to present of novel method of analysis. A good source of articles is Professional Geographer, but please feel free to choose from any journal you like.
Your review write-up should be part report, summarizing what the author's scientrific problem is, how they set up their experimental method, what they did, and the conclusion they draw. Most importantly, however, you must critically evaluate their method and the results, and explain whether their arguments are convincing and why. If applicable (i.e. Unless the analysis is perfect), comment on how the research or analysis might have been improved upon. You may get get help from the TA or myself to help understand the method used.
Course Content and Topics
Topic 1: Calculus Introduction/Review
Topic
2: Review of Basic Statistics: Sampling distributions, point and
interval estimates, hypothesis testing vs. significance testing.
Readings:
McCloskey, Donald N. 1987. "The
loss function has been mislaid." American Economic Review.
75(2): 201-205.
Lenhard, Johannes. 2006. "Models and
statistical inference: the controversy between Fisher and Neyman-Pearson."
British Journal of the Philosophy of
Science." 57: 69-91.
Gould, Peter. 1970. "Is Statistix Inferens the
geographical name for a wild goose?" Economic Geography. 46
(Supplement): 439-448.
Topic 3: Review of Linear Regression: Bivariate, multivariate and ANOVA. Understanding the classical assumptions as empirical commitments.
Topic 4: Analysis of Time Series Data: Autoregressive (AR), Moving Average (MA) and ARIMA representations. Generalized Least Squares.
Topic 5: Geostatistical Models: Variography, Kriging and stochastic simulation
Topic 6: Multivariate Methods: Principal Components and Factor Analysis.
Topic 7: Point Pattern Analysis: 1st order (density estimation) and 2nd order (clustering/regularity) analysis.
Topic 8: Geographic Sampling
Topic 9: Statistical Analysis of Networks: Moran's I and Geary's C statistics. Simultaneous and Conditional Autoregression.
Topic 10: Analysis of limited or censored dependent variables: Logit/Probit regression. Models of hazard and survival.