Search this site
Hanowell.info‎ > ‎Academic Home‎ > ‎

Evolutionary Modeling Survival Kit

Stop Crying and Train!

    In addition to (and more important than) lots of coffee, a pad on your desk to soften the blow when you fall asleep over reams of Mathematica code, and lots more coffee, you're going to need to start training in analytical and computational modeling techniques early, and especially if you an uber-newb like I was.

     Listed below is a series of books that I recommend to anyone embarking on a career in evolutionary and ecological modeling, especially one that integrates cultural, genetic, and epigenetic processes.  I do not purport that this list is complete.  Nevertheless, if you read the books I recommend, they will point you toward yet more books that will be of use.  

    Take it from a guy who began by barely remembering his high school calculus courses to being able to develop, understand and analyze stochastic models, all through diligent study and much banging of his head against the desk through tears of frustration and occasional bouts of enlightened euphoria, all over the course of one year (it's been rough; thanks for understanding, Malyse [she's my girlfriend; swoon]).  

    We'll start with the uber-basics and move to the sweet, useful stuff that will help guide and inform your models-based research.  Notice that this reading list is dominated by math primers and mathematical texts.  That's because it is my opinion that one benefits most by starting with the mathematical techniques, both in research, and in training.  See below for links to the various publishers.

   Anyway, as Nabakov once wrote (although in a much raunchier context), "Look at this tangle of thorns."

(1)  Need to brush up on your calculus before jumping into the theoretical application (yes, you do)?  I recommend (and this is no joke):


Calmly put your ego and your sense of pride into a safe deposit box for a week and remind yourself what a derivative and an integral are and why they are important in the real world.  If you read through the book and realize you've forgotten high school algebra, as well, go back to step 1 and replace Calculus for Dummies with Algebra for Dummies.  Rinse and repeat with Calculus for Dummies.  Thankfully I didn't have to go that far, and I suspect that if you are remotely interested in becoming a theoretical modeler, you won't either.  But if so, that's okay.  Baby steps.    

(2) Okay, so you've brushed up on your calc skillz (and your whining like a baby to your annoyed but patient girlfriend skillz, too; those will continue to be useful).  Next you've got to get a basic introduction to the whole concept of mathematical modeling.  There are a few options for this, but I recommend the following, while recognizing that ANY book you are likely to read has some errors in it, is going to require work, and that getting through all of these texts requires discipline and patience:

    (a) 

Otto and Day's awesome introduction to mathematical modeling in evolution and ecology is the primer you need to get started.  It is, in my opinion, the best introduction out there for this gig.  It includes an introduction to modeling philosophy, recipes for developing and solving analytical models, primers for function formation and approximation, linear/matrix algebra, and probability theory, and step by step examples of mathematical models in action for a wide range of theoretical problems in biology. 

If you are just getting into the modeling gig and you are for real, you need to have this bad boy on your shelf.  Oh, and work through the Mathematica supplements on their website, as well!  Because if you want to be a modeler, you need to know how to use some modeling software, and Mathematica is one of many options.

    (b) 

Kokko's text is a funny and light-hearted introduction to a broad reach of modeling concepts in biology, and it has the added bonus of providing examples of MATLAB (another piece of mathematical software) code for each chapter.  This is a great companion piece to Otto and Day.








    (c) 
McElreath and Boyd is a good text, particularly if you want to get a clear introduction to social learning models, multi-level selection, and especially if you want a great primer for covariance genetics (the Price equation) as a tool that is widely applicable in the theory of evolutionary ecology.  The book mainly focuses on social evolution (kin selection, reciprocity, some indirect reciprocity).










    (d) 
This book is a testament to Martin Nowak's ability to clearly, succinctly deliver ideas in a step by step fashion, and the reason why I have a man-crush on him.  I wouldn't recommend reading this unless you have gone through step 1 (or don't need to go through step 1) and gone through most of Otto and Day and McElreath and Boyd, mainly because once you have done that, you can kick your feet up and just enjoy the clarity (and formulate constructive critiques, as well!, particularly of the cooperation and language evolution sections).  Prepare to be impressed.






3) Okay, so you've been training in analytical modeling.  What about statistical modeling?  Yeah, you're going to have to confront those models with some data at some point, aren't you?  Well, maybe...Epstein would say that prediction isn't the only purpose of modeling.  Nevertheless, if like me you want to interface your theoretical with your empirical work, you've got to confront the new statistical revolution.  Read any papers that use the terms "MCMC" or "burn-in" or "Bayesian" or "prior distribution" or even "posterior distribution" lately?  Feel like shutting the garage door and starting the car, and just leaving it running, or perhaps giving up on scientific research every time you do?  

Well, then Benjamin Bolker's new book is just the thing for you!  It is an awesome, step by step introduction to ecological modeling and data analysis and it will teach you how to do it in R!  For those of you who aren't in the cool group, R is an open-source software "environment" with very flexible graphics, statistical analysis, and even analytical modeling capabilities (although I'd recommend Mathematica, my fav, or MATLAB for the final purpose listed), plus publicly available extensions for things ranging from social network analysis to...well, you name it, I suspect.  Anyway, Bolker covers data management, maximum likelihood methods, Bayesian analysis, and much more.  Get this book or something remotely like it and nerd out despite your girlfriend's protests (besides, you know she thinks deep down in her heart that Bayesian inference is sexy!).



Need an introduction to basic statistics?  I'm sorry to say I can't recommend any basic statistics texts because I've never read any for self-teaching.  I've heard from an anonymous source that they are all awful, but I imagine this is because learning statistics for the first time is usually a horrible process for most human beings (and if it's not, I hate you).  I learned basic frequentist statistics, elementary regression modeling, logistic regression, and introductory survival analysis by living through three quarters of biostatistics courses at the University of Washington.  My advice: take a course in statistics in an anthropology, economics, psychology or sociology department for a basic frequentist primer (usually the econ, psych, and soc courses are the better ones).  Biostat departments are good with this, too.  Stat departments are a maybe, because they might not give any examples that seem relevant to the questions you would eventually use the method to try and answer.  Chances are, this is already a requirement in your department.  Who knows, maybe there is a program at your school that is revolutionary in that it doesn't start with basic frequentism, but instead starts with likelihood methods (which are philosophically frequentist in many respects).  I hope that soon, this option becomes available, because in the opinion of another anonymous individual I respect greatly, learning statistics in the traditional way may be sub-optimal nowadays given the cheap, fast computing power that is available.

All right, hop to it.  I expect you to have developed a cogent, concise, and elegant theory of everything evolutionary by Christmas.

And lastly, I give you the old adage: "No pain, no gain."

Publisher sites: