mod 3 tester comments

[kcudding]

Mod 3

• One topic that could be included will be Discriminant Function Analysis.
• include stats like permanova
• I am missing, however, real examples in which these analyses were used by research to solve ecological questions. Showing the graphs researchers used in their papers, with an explanation of the interpretation will allow students to see how these analyses have been used, and in which cases.
• a list with the type of ecological questions that had been addressed
• For some explanatory figures the code is missing, and it could be helpful for the students (e.g., figures 4.16 and 4.17). I do not think this is bad or wrong, but is more a question from me to you: why are you not using ggplot to create the graphs?
• kclust=kmeans(niso[,c(“C,” “N”)], 4)), and this is how it should be: kclust=kmeans(niso[,c("C", "N")], 4)). NOTE the quote marks!!

· Examine the significance test. Does this automatically selected model differ from the full model and your manually selected version?” in section 4.3.10 Forward Selection of Explanatory variables.

The answer is yes, but there are multiple reasons that make this model different. What are you expecting students to notice? discuss: guidance for students to think as a hint

Should it be possible to have a box to answer and then get the accurate response, to see if what they are interpreting, and thinking is right or not?

-look at figure in this paper ?

put reference to answer later in the chapter, so students not tempted to look up initiallly

• I am missing more exercises with different sets of data, and examples of how people interpret these analyses in real ecological contexts.

examine kmeans function.... not acting random?? need to look at this??

Modify figure 4.3 to emphasis difference betten simple and complete, use an oval

---

• took 6-7 hours

• Could the data in the examples (multivariate resemblance) also be provided? I think it might be helpful

• Also, not sure if all the R functions have associated packages next to them, if they don’t, I’m assuming they are base R functions. Also, would it be possible to have hyperlinks for the R functions that lead to further information?

• Also, maybe the references could also be hyperlinked.

• Also, some functions are bolded, but some aren’t. It might be helpful to bold all of them.

• 4.1 could have been longer. That said, I understand that the analyses covered in ordination and cluster analysis might be more applicable to current research.

• An exercise in 4.1. Offer an additional introduction in R before introducing more involved exercises later in this section.

referencing solutions in the literature: what are the copyright issues for reproducing graphs??

[kcudding]

lda provided in classification module should provide examples in lit as in classification module. Obviously cannot supply figs etc due to copyright no list of ecological questions since 1. this is biology not ecology, and 2. above would more or less do that