WK 28 - CANCELLED - How to Avoid Common Pitfalls in Multivariate Analyses

Sunday, August 6, 2017: 12:00 PM-5:00 PM
A106, Oregon Convention Center
Jeri Peck, The Pennsylvania State University
Just because the reviewers don't catch it, doesn't mean it wasn't wrong. The fact is, the most common advanced analysis tools for ecologists (multivariate statistical applications such as ordination, clustering, and group-testing) are often either never taught formally or not taught by an ecologist, rendering many of us functionally self-taught.  Most of the time that works, but some tricks are worth being taught. In this interactive half-day workshop, we'll discuss the implications of inappropriately structuring your data for analysis, inadequately ensuring correspondence between those data and the distance measures you use, and improperly interpreting the output. I'll demonstrate these problems using polar ordination, PCA, NMS, and MRPP and show you some 'best practices' for avoiding these mistakes. You'll use your own laptop and software provided for the day (PC-ORDv.7) to apply these practices on sample datasets or your own data.

Participants should be familiar with the concepts of ordination (i.e., their interpretation) and distance measures and must bring a laptop capable of running Windows software and will be provided with the software and sample datasets at the workshop (via USB).  If you wish to use your own data, you will need to format them correctly for the software in advance.  Don't forget to bring your power cord and a notepad or the recommended textbook (Step-by-Step, second edition).

Registration Fee: $35

See more of: Workshops