“. . . tracing a stream of metaphors that runs right through language and flows from the concrete to the abstract. In this contstant surge, the simplest and sturdiest of words are swepts along, one after another, and carried toward abstract meanings. As these words drift downstream, they are bleached of their original vitality and turn into pale lifeless terms for abstract concepts—the substance from which the structure of language is formed. And when at last the river sinks into the sea, these spent metaphors are deposited, layer after layer, and so the structure of language grows, as a reef of dead metaphors.” (The Unfolding of Language, by Guy Deutscher)
Models are, among other things, tools that help us better understand the universe. To paraphrase the words of eminent evolutionary biologist Marlene Zuk “models are devices that allow what-if reasoning.” One could legitimately argue that we constantly using models, and in a multiplicity of ways, including, as examples, linguistically through the use of metaphor, physically through representations of scaled-down or -up objects (solar system, cell models, buildings, etc.), symbolically through mathematics, or graphically through the depiction of relationships or flows. From models come understanding.
To gain a first-hand experience of modeling in ecology (or evolution, if you’re so inclined), I am assigning that you find a study that makes use of a model, and recreate it. There is a great deal to be learned from being in the weeds of a study with an ecologist that is trying to understanding an aspect of the natural world, attempting to create a model that will capture its essential elements, and trying to gain deeper understanding that would otherwise been unknown without the model. Don’t get me wrong, there is a lot to learn from the broad kinds of modeling that we are studying from the text (mostly population-level, theoretical models), but there is also a lot to learn from trying to recreate a newer, (perhaps) messier, real scientific model. In doing so I’m hoping that you’ll grow and learn through learning idiosyncratic approaches the authors used and see how models are created, analyzed, and used in practice—it’s very different from the nice, clean models we read about in the textbook that have been studied for many decades! This might include discovering implicit or explicit assumptions of the model, approaches to parameterizing of the model, learning different quantitative techniques for analysis, seeing new types of visualizations, and lots of other possibilities!
But perhaps most importantly, after creating and analyzing a real scientific model, I am hoping that this project will empower you by revealing to you (and me, of course) that you have the ability to model at a professional, scientific level. As you all are preparing to graduate within the next 0.75–1.75 years, these types of activities will certainly benefit you in many expected and unexpected ways.
The Remodeling Project will have 6 parts:
The timeline will follow:
| Date | Task |
|---|---|
| September 18 | Introduction to the project and assigned literature search |
| September 25 | Present Chris with up to 7 potential papers |
| September 25—28 | Meet with Chris to choose a the original research you wish to recreate |
| October 02 | Proposal presentation (essentially the paper with a light version of the methods) |
| October 16 | Methods presentation |
| November 06 | Preliminary results presentation |
| November 20, 27, December 04 | Remodeling project analysis in lab |
| December 05, 06 | Results presentations |
| December XX | Final research paper due (will announce time when final schedule is released) |
“Remember, always, that everything you know, and everything everyone knows, is only a model. Get your model out there where it can be viewed. Invite others to challenge your assumptions and add their own.” (Thinking in Systems: A Primer, by Donella Meadows)
The first part of the Remodeling Project will be to find a model that you wish to recreate. To do so, you will first search broadly, then narrow your choices down on your own and with my help. I highly recommend you use an actual database to search for articles. Perhaps the most useful tool available to you is through the library, called SCOPUS. Opposed to unmanipulatable blackboxes like Google and Google Scholar (don’t get me wrong, even those can have their time and place), SCOPUS allows you to search all sorts of criteria (keywords, author, article type, etc.), and further sort and filter (e.g., filter specific journals, sort by relevance or date).
Before you take to the interwebs, take some moments to reflect on aspects of ecology (or evolution) that you find most interesting. I would suggest searching for what you find interesting in addition to some of the keywords I suggest below.
In terms of your search criteria, there are several journals that I would recommend you search:
Here are also some keywords that will perhaps help you find a recreatable modeling paper. Use these in combination with the aspect of ecology or evolution you’re interested in (e.g., a taxon, interaction type, location):
The types of models that we are largely focused in in this course are dynamic models, meaning that we study how populations change over time. When looking at the math, if the left-hand side of the equations have a \(\mathrm{d}x/\mathrm{d}t =\) or \(x_{t+1} =\), with \(x\) being population size, biomass, density, etc., then you are probably looking at a suitable model for this project. I understand that I am asking you to find papers with models that you have only just begun to learn about, but that’s because we need to start this project early in the semester and I am also here to help you.
In terms of the age of the paper, I think that either new (within the past 10 years) or really old papers (classics from the 1970s or earlier) would be good choices. (I will strongly suggest the former.)
Last bit on criteria. I would like for you to avoid two types of model: (1) individual-based models (AKA agent-based models, IBMs, ABMs) and (2) species distribution models (AKA ecological niche models, ENMs, SDMs). IBMs require a totally different type of programming, which I would like you to steer away from. If you have experience or want to learn these methods I’d of course support you, but I cannot give you the support that you may need to complete a successful project. SDMs require external data, and these data aren’t always accessible. Like with the IBMs, if you have experience or want to learn these methods I’d of course support you, but I cannot give you the support that you may need to complete a successful project.
For this first set of up-to-7 papers, I would like for you to create
a folder in Private/RemodelingProject/ called
PotentialPapers. In here, save .pdf copies of papers that
capture your interests that have a modeling component. Have all of the
papers there by Monday, September 25. I will look over
them and we will arrange to meet to discuss your interests and
capabilities to recreate the model.
An important note is that I would like for you to look for papers without concern for the “difficulty.” Each project will likely have differing degrees of difficulty, and I will do my best to standardize them all by making lighter models more substantive and more difficult models lighter. Really, I want the subject to be important and interesting to you, because that will bring out your drive and best work!
Congratulations! You’ve hopefully decided upon a paper for your Remodeling Project by this point! This paper will be the major project for this course, which will be one of the main foci of the course follwing break.
As mentioned above in Part I, the next step in this project will be to deeply familiarize yourself with the paper and to share it with the rest of the class. The benefits of these presentations are threefold:
I ask that this first presentation, that will take place October 02 in lab, be 7 minutes long. There will be a few minutes for questions following each presentation. The format of the presentation should be that of the scientific paper itself, with a conceptual background, motivation (e.g., hypotheses, goals, aims), methods, results, and discussion. If I had to suggest a ratio of time for each of the sections in for the presentations (excluding motivation), I would suggest a 2:1:2:2. That’s mostly because I want you to present on the ecology of the paper, why a model was warranted, what the model yielded, and what the model output means for the ecological question. Your next presentation on October 16 will be strictly on methodology, so please just summarize it here. (The methods presentation will be around 1:2 of (what they did):(what you are doing).)
Aside from that, I don’t have any strict guidelines. Generally presentations have little text on slide, a goal of around 1 slide per minute works well, and remember that slides are enhancements for talks, not crutches.
Please send me a link to your slides via email before lab on Monday, October 02.
I’m looking forward to learning about these papers in more depth next Monday!
The next milestone in the Remodeling Project is the a presentation on core of the Project, the methodology. For the past few weeks you have been studying this paper, which includes figuring out the authors’ methods and how you are going to recreate the project. For the sake of sharing your Project with your classmates and because people learn by teaching, I ask that you present the methods in lab in 2 weeks, on October 16.
In terms of this presentation, I ask for the sake of time that it is only 7 minutes. For the content, I think that a balance of 2:3:1 on:
would make for an informative and useful presentation.
To present their methods, do your best to be clear, concise, and synthetic. Keep in mind for this presentation that we want your peers to be able to understand what they did. That means defining terms that they may not know and organizing the presentation of their methods in a way that is succinct, which is unfortunately not always the case. You know your audience and what they know, so please aim for that level.
Ultimately, I want this to be a fun and engaging experience, so please feel free to be as creative as you desire.
As somewhat of a check point, I ask that you present the progress you’ve made on your Project. At the point of this presentation, November 06, you should have a minimum model running and have recreated at least one of your paper’s figures. These are just goals and given that every Project will be different, you will all be at different stages at this point. The presentation should include an introductory slide where you remind us of what the paper is about, a general slide summarizing the methods (to give us the big picture of the forest), followed by however much of the model you are able to present (and any accompanying methods). I am imaging these being 5—7 minutes.
The culminating event of your Remodeling Project will be the presentation you deliver the final week of class, December 05 and 07. Everybody’s presentations will be due to me before class on December 06 (11 AM, emailed to me), although the presentations will be over a period of two days.
This presentation should be strongly weighted on the presentation of your results. That is, you will have been working for roughly 8 weeks on understanding and recreating your modeling paper, and the presentation is your time to share how you were able to reproduce the model from the paper you chose. This will be different for every individual Project, as every Project has its own unique set of hurdles. Some unique hurdles may include putting the model together, some are understanding the mathematics, some are translating what they did into R, and some are mimicking the model results.
The class periods are 75 minutes each, meaning that we must divvy up the time between the 9 presentations to ~17 minutes. Therefore, I ask that your presentations be a maximum of 10 minutes, which will allow a few minutes for questions and some time to switch presenters.
In terms of presentation guidelines, since you have already given presentations on the overall paper, the methods, and a minimum model, that material doesn’t need covered. If you do feel compelled to present your methods, then please do so; just remember that this is a presentation that should maximally focus on how you recreated the paper’s model and results.
I think an effective way of presenting your results would be to show, on one slide, the figures from the paper, and your recreated figure, side-by-side. Take your time explaining what you did to make the figures, what their figures are showing, and how precicely and accurately you were able to recreate them.
I hope that you all feel good about the hard work that you’ve put into this project and proudly showcase what you’ve done.
Pat yourself on the back for successfully recreating an ecological model! (Seriously, you should be very proud of yourselves.)
I hope that you learned a lot about population modeling, whether that is understanding and interpreting a published model, learning methods for creating or analyzing a model, how to translate a written model into R, or the multitude of other things that could be learned from this assignment. I hope that this assignment has empowered and emboldened many of you through the practice of working with theoretical and mathematical literature, but also opened your eyes to this literature that’s probably new to most of you.
I digress . .. you’ve been intimately working with your modeling paper now for over two months and I want to read about your experience! Also, I presume that your future selves will also be interested in reading it, which is why the format is part summary and part description of your experience (more below). This will be the final assignment for this course, due 2022 December 16 at 6 PM via email. I would like for you to sent a zipped file of your .Rmd, .html, and all other files I would need to knit your .Rmd (i.e., everything in your working directoy).
Please write your paper using R Markdown. A couple of notes:
knitr::opts_chunk$set(echo = FALSE) in the setup chunk in
the beginning of the document. Every chunk of code you wish to be in the
document, simply write echo = TRUE in the chunk
options.<img src="img.jpg" style="width:500px;height:500px;">
You can remove the style options and it will appear in your
knitted html document without resizing. Like \(\LaTeX\) code, this html code works outside
of the R chunk.But most importantly, I ask that your paper is structured in the following manner:
The page descriptions are only guidelines to help describe the amount of detail that I’d like, not including figures, single spaced (so, maybe 3-to-5 paragraphs per page, depending on your writing style). Use your discretion in terms of including the most important details. Like I mentioned above, this assignment is the culmination of what you’ve learned from this project, and really, this course. I would like you to demonstrate command and understanding of the modeling and ecological processes in documenting what the paper was about, their modeling process, your modeling process, your ability to reproduce their results, and some specific questions that I will ask you.
I’m happy to read over a draft or outline, and please be sure to edit and proofread before submitting it to me, via email.