Notes on Spatial Data and Analysis in Research

Check your Data!

When looking for and collecting data, keep asking the following questions:

• Can you obtain the data you need to perform your research? If not, change the research.
• Is the data available in your price range?
  • Is it "free". If so, nothing is free, what will it take for you to use it?
  • If not, is the cost worth it vs. doing it yourself or changing your questions?
• Does the data have the required uncertainty?
  • Check the metadata
  • Talk to the folks that created the data
  • Don't assume it does, check it out for yourself!
  • Talk to others that have used it
• Can you use it for your intended purpose?
  • What are the copyright and publication limitations?
  • Are there any ethical concerns? (i.e. can publishing this data get anyone hurt or sued?
• Get "intimate" with the data
  • Map it, graph it, histogram it, run stats on it
  • Look through the attributes

Be Mindful with Analysis

Your analysis is key to using spatial data.

• Document the analysis steps with flow charts and/or step-by-step procedures
• Repeat the steps over and over again to make sure it is correct
• Have someone else execute the same steps, did they get the same result?
• Does the analysis make logical sense and does it support your conclusions?

Document and Archive Everything

• Uncertainty maps are very cool. If you can't create one, at least add a "Caveats" section to your discussion.
  • A caveats section explains the limitations of your results based on the uncertainty of the data, processing, and analysis.
  • This may feel like it makes your work weaker but it actually makes it stronger. In other words, would you rather be caught using data that had errors you did not document or be up front about the errors and their implications?
• Your final products are not the end!
  • Archive your data, processing steps, and other documentation with an organization that does long-term archival (e.g. the library, NOAA).
  • Save a copy for yourself as you'll at least refer to the materials in the future and you may be asked to create an updated product, sometimes 10 years in the future! Also, you may end up teaching a class and can use them as materials!

Don't Just Execute, Think about Process and Long-Term Success

• Read "7 Habits of Successful People" (at least checkout the Wikipedia page periodically)
• Setup an organized data store for your spatial data
• Create a group around you that has the diversity of skills to be successful and enough process to be efficient and happy.