/stats_for_soil_survey

Statistics for pedologists

Primary LanguageHTMLGNU General Public License v2.0GPL-2.0

Course Content

Current Table of Contents

Update on "Statistics for Soil Survey" course

9/3/2015 Shawn McVey, Katey Yoast and Tom D'Avello met with NEDC Training Specialist Stacy Rice August 18-20, with Dylan Beaudette and Skye Wills calling in for portions of the meeting. A listing of the highlights include:

  1. Renaming the course from "Statistical Data Analysis for Pedologists", to "Statistics for Soil Survey." This presents a more expansive scope as the content would also be useful for Ecological Site Specialists, in addition to achieving a 2x reduction in the number of syllables.

  2. Fine-tuned and updated the Table of Contents of the web based content and what will be delivered in the course

  3. Developed a Course Description and Lesson Plans for each module (A module generally corresponds to a chapter)

  4. Proposed delivery methods for the course. Offer Chapters 1-6 via distance learning over the course of 2-3 partial days. Chapters 7-10 would be offered the following week in a traditional classroom environment over the course of 3 full days. Would 1-5 work better for distance learning and 6-10 as conventional classroom?

  5. Classroom material would be developed and presented using power point slides. Links to the web page will be used when further explanation is needed

  6. Exercises are present throughout the course and involve use of common datasets and datasets of individual interest (This is something the group needs to figure out to make sure we have a consistent theme)

  7. Require completion of a Case Study so participants have a chance to demonstrate the ability to conduct analyses, interpret results and offer options for proceeding with the findings. Suggest it be data for their region.

  8. Discussed the purpose of the Web-based content and how it relates to the course

The purpose of the web page is to serve as a permanent reference for all of the content, similar to a text book. The primary difference is the web page will limit esoteric mathematical explanations, concentrate on demystifying the material and serve as a quick reference to refresh ones memory or get a self-starter introduced and functional with the tools and techniques.

There may be overlap between slides and the web-based content. That should not be a problem as a web page is a better means of serving as a reference than slides. Using content from the web page to help author slides will expedite course development. In addition, a link to the web page may be better than a tedious, word dense slide.

  1. Date for the pilot class to be March 1-3 and 8-10, 2016

  2. Volunteers are needed to complete web based content for:

  • Logistic regression
  • Tree based models
  • Validation and uncertainty
  • Clustering and ordination

  1. Slides and exercises/interactions need to be developed for all modules. The individual(s) working on the slides will typically be the same ones developing the web content.

  2. Web content will be posted to GitHub soon. Any edits to the current and future content is fair game, whatever it takes to get the lessons conveyed.

  3. Aaron Achen, Editor at the NSSC has agreed to edit the web content to make sure we are communicating well, rather than "real good"

  4. As a pilot session, we can suggest participants. It may be good to get a few folks in the pilot session that would be strong candidates for becoming future instructors. At the minimum, we need a few savvy folks that will be helpfully critical. In addition, people signed up for courses that have this "Statistics" class as a recommend prerequisite (e.g. DSM with ArcSIE) would be candidates.

Datasets

Standard data sets for teaching, student data for exercises.