Feedback on 2-intro-ds

[jjrob]

Great job on this presentation. Here are some specific suggestions. I do not know if I will provide this many for the other presentations but I went deep into this one as a way of getting my mind on track for what I will present afterwards.

• Slides 6, 7: In the last bullet, you should explicitly say N_hat = D_hat * A = 590. To be perfectly clear to the reader, do not leave out the D_hat * A part.
• Slide 9:
  • Say "in real life" rather than IRL. Some readers may not be familiar with acronyms like this.
  • Third bullet, suggest you say "object" rather than "line". I know you are trying to get into how we model with perpendicular distance but in this very first slide on detectability, I suggest keeping it as simple as possible.
  • Suggest you add a photo showing this. How about http://www.nordhavn.com/egret/images/7962.jpg (credit someone on this page http://www.nordhavn.com/egret/captains_log_sept11.php). Shows near and far dolphins in same page. Which are easier to see?
• Slide 11: need to talk here or someplace nearby about perpendicular distance vs. distance to the observer
• Slide 16: I know we expect people to know math(s) but it could be helpful to show the detection function here (and possibly integral from slide 17) so when you talk about integration you can directly show how the integral is computing the area under the curve, which is < 1, and how if there was perfect detectability there would be a line at g(x)=1, and the area under that line would be 1. Walking people through this graphically makes the integral on slide 17 less scary.
• Slide 17: Fourth (round) bullet: show the result of n / p_hat = 100.32. This may seem obvious, but (I argue) it really drives the point home to see that, once you correct for detectability, "look at how many more animals there were that you missed!"
• Slide 21 or 24: Could show bowriding dolphins, e.g. http://www.corkwhalewatch.com/whale-watching-in-west-cork/ (scroll down)
• Slide 22: I know we're not getting into availability and perception biases in detail in this workshop, but they can be a real problem for marine species. I suggest we at least put in a figure that at least reminds people to consider it later. For availability, you could take Fig 1 or part of it from http://rspb.royalsocietypublishing.org/content/279/1731/1041. You could easily fit this below your bullet points. For perception bias, you could use something like https://pixabay.com/en/seal-head-water-surface-eye-591292/ and point out that it could be difficult to see compared to, say, animals that exhibit behaviors like this: https://commons.wikimedia.org/wiki/File:Orca_porpoising.jpg
• Slide 30: It would be nice to explain in passing what "hazard rate" actually means. Where did it come from?
• Slide 31: Show an example of an unadjusted function and then one with the adjustment applied. I can provide a real-world example if desired.

[dill]

Thanks for these comments Jason! I've made modifications in 66fab6a

I didn't include the depth figures from Hooker et al (2012), as I don't think it adds that much (if you don't know how to read such a graphic already), also non-colourblind-friendly plots :'(.

I'm a little nervous about using other people's photos.

I'll talk about hazard-rate, rather than include more text on the slide.

I've also added in the animation I made of line transect sampling, which I'd not included so far for some unknown reason...