Assignment: Statistics for Geographers PROJECTS
Introduction
The project is an opportunity to synthesize what we have learned across the entire semester. You are asked to complete a statistical analysis of primary* data from start to finish, beginning with articulating a hypothesis and identifying appropriate variables and ending with a conclusion based on inferential statistical testing. Your grade on the project will be based on the extent to which you can demonstrate mastery of the concepts we have learned in class. The exact manner in which you design your project is up to you, but the ideal project will demonstrate mastery of:
• Organizing and categorizing data
• Descriptive statistics
• Statistical distributions
• Means comparisons
• Correlation and regression
*By primary data, we mean data that you "collect" on your own - although these data could be collected many different ways, including your own observations, analysis of imagery or text, surveys, etc. Downloading a data set from the internet and then generating new metrics from those data for your analysis may be acceptable in some cases.
Logistics
You may work individually or in groups of no more than four students for your project. Each "team" must collect its own primary data, but teams can work on the same type of problem and even collect the same type of data (although, ideally, not at the exact same location/time). Groups of three or four students should designate a coordinator/manager who will be responsible for ensuring that the team works cohesively to achieve its goals.
Expectations regarding the amount of work that teams complete vary by group size. As a rough guideline, teams should collect and analyze three variables per group member. Possible configurations for variable collection include:
One person 2 independent variables, 1 dependent variable
Two people 3-4 independent variables, 1-2 dependent variables
3-4 people 5-9 independent variables, 1-3 dependent variables
To reasonably conduct most of the statistical tests we have been using, groups should aim to collect at least 20-30 observations, and at least 10-15 observations for each "category" of any independent variable. More observations are certainly welcome and will increase the quality of your analysis, but please do not spend dozens and dozens of hours collecting data. Division of labor and effort should be equal among all team members. Each team member should be fully capable of answering any question about the data collection, analysis, or interpretation thereof, and thus it is strongly recommended that all team members be involved in all parts of the project.
The project represents 35% of your semester grade. Your grade on the project will be based on your data collection plan (5%), sound analysis of your data (10%), a written report summarizing your findings (10%), a short in-class presentation (5%), and a brief reflective statement on your own experience this semester (5%).
One report per group uploaded to Blackboard, plus R code
One statement per student (250-400 words) uploaded to Blackboard
Reports
The exact length of project reports will vary by group size. You should use as much space as you need to communicate your findings. Individual reports will probably not need to exceed 7 pages, including figures. Teams of two will probably not need to exceed 10 pages, and teams of three or four will probably not need to exceed 15 pages.
Your reports should describe what your group completed at each stage of project design, analysis, and interpretation. A recommended table of contents for the report, which may also serve as a useful guide for completing your project, is shown below.
1. Project Title
2. Team Members
3. Research question(s)
4. Description of independent and dependent variables
Type of variables
Expected distribution of variables
Potential unusual cases/difficulties
Specific hypotheses about the relationships between the independent and dependent variables
5. Data collection and analysis plan
Estimated number of observations
Location(s) and schedule for sampling, sampling interval
Measurement type/unit/mode (for each variable)
Anticipated data collection challenges
Descriptive statistics to report
Statistical tests to employ
6. Data collected (mirror section #5, after data collection is complete)
Problems encountered
Metadata, missing data codes, other necessary coding
7. Data analysis
Descriptive statistics for each variable
Graphical summaries for each variable and combinations of variables
Comparison of distribution for each variable to expected distributions
One sample tests for certain variables against expected values
Two sample tests for certain variables based on groups
Analysis of variance tests for certain variables based on 3+ groups
Correlation analysis for certain variables
Regression analysis for certain variables
8. Interpretation
Decision to accept or reject original hypotheses
Potential causal explanations for observed patterns
Variables excluded from the analysis that may have been important
Extent to which data collected meet necessary assumptions for certain analyses
Strengths and weaknesses of study design
Recommendations for future projects on related topics
Broader impacts/societal implications of your findings
9. Conclusions
10. External data sources/references (if necessary).