Anthropology 298 (ANT 298 — You need to fill out a 298 form to register )
Ecology 298 (ECL 298 — CRN: 60046)
Statistical Thinking in Evolutionary Ecology
SPRING 2010

Instructor: Richard McElreath
Location: 302 Young Hall
Time: Tues/Thurs 3:10pm-4:40pm

Topic Schedule:

Week 1 – Likelihood: Estimation and confidence
Recommended reading: Bolker, Chapter 1 and first portion of Chapter 2.
Files: week1.pdf , week1.r , week1x.pdf , week1x.r , classdata.r

Week 2 – What’s wrong with null-hypothesis significance testing
Files: week2.pdf , week2.r , week2x.pdf , FA meta data.csv , week2xsols.pdf
Articles: Møller_Thornhill_1998.pdf , Gigerenzer_Null_2004.pdf , Odds Are, It’s Wrong.pdf

Week 3 – Linear models
Files: week3.pdf , week3.r , week3x.pdf , foxes.csv , primates.csv , week3xsols.pdf

Week 4 – Hypothesis testing and model comparison (AIC)
Files: week4.pdf , week4.r , coeftab.r , week4x.pdf , week4×1.r , week4xsols.pdf
Reading: AndersonBurnham.pdf , statisticalconfusion.pdf

Week 5 – Linear models and interaction effects
Files:  week5.pdf , week5.r , week5x.pdf , tulips.csv , sickle.csv , week5xsols.pdf

Week 6 – Generalized linear models: Poisson and binomial regression
Files:  week6.pdf , week6.r , week6x.pdf , elephants.csv , salamanders.csv , week6xsols.pdf
Reference: mle.pdf , plotting.pdf
Reading: darwin in the fossils 2008.pdf
Bolker, chapters 6,7 and 8 are relevant, but not all parts are relevant right now. Chapters 5 and 6 of Gelman & Hill are more relevant right now. They notate differently than I do, but you’ll see that similarities anyway.

Week 7 – Generalized linear models: Negative binomial and ordered logit/probit
Files: week7.pdf , week7.r , week7x.pdf , ucb.csv , oceanictools.csv , polrplot.r , week7xsols.pdf
Reading: Gelman & Hill, Chapter 6; Bolker, Chapter 4, especially section 4.5 onwards.
Extra: gigerenzer making sense of health stats 2009.pdf , common ancestry AIC test paper.pdf

Week 8 – Statistical thinking
Files: week8.pdf , week8.r , week8x.pdf , galapagos.csv , tadpoles.r , tadpoles.pdf
Reading: sim.pdf , validate.pdf

Week 9 – Bootstrap and resampling methods
Files: week9.pdf , week9.r , week9x.pdf , willowtitn.csv , week9xsols.pdf , mlesong.txt

Week 10 – Modeling variance: Hierarchical and mixed effects models
Files: week10.pdf , week10.r
Reading: Gelman & Hill chapters 11–15.

Equipment: You’ll want to bring your own laptop to work on, after the first session.

Purpose: This course is a bootcamp in modern model-based statistical analysis. Whether scientists are aware of it or not, all statistics are based upon models, and being able to translate one’s unique research hypotheses into a set of unique statistical models is extremely valuable. Otherwise, you are at the mercy of statistical ritual. The substitute is statistical thinking, and that’s what I aim to teach.

We’ll focus on this kind of translation—which allows one to build a model, starting from scientific principles—and then how to estimate effects within a model and finally compare the explanatory and predictive power of different models. Build, estimate, compare, is the mantra.

This is a practical seminar, not focusing on mathematics nor the derivation of theorems. I’ll assert important theoretical results and give you the citations, but we’ll be focusing on building and modifying the code necessary to be a practicing modern applied statistician (with an evolutionary emphasis).

Preparation: If you’ve had a basic statistics course at some point, you can make it through. If you already know how to program, in any language, you’re well prepared.

Topics: We can’t cover everything here, nor can we cover anything in total depth. I want to cover basic statistical philosophy—the arguments that justify some procedures, like likelihood estimation, and show the silliness of others, like null-hypothesis testing. We’ll cover generalized linear models, because these are the focus of most modern statistics, and they provide a language for translating hypotheses into statistical models. We’ll spend a lot of time comparing models and plotting predictions, because these are essential strategies for interpreting statistical models. We’ll finish up with material on hierarchical and random effects models. I’ll say some things about being “Bayesian” along the way.

Labs: Every lecture is woven together with a computer lab, so you get experience actually performing analyses, as well as building models.

You’ll use the R statistical programming language for everything in this course. The advantage of using R is that it is quickly becoming the standard for scientific statistical modeling. The disadvantage of using R is that it has quite a steep learning curve. I’ll put some stairs on this curve for you, by giving you example code for everything. You’ll never have to generate code from thin air, but instead modify a given solution for a similar problem.

I’ll also introduce key programming concepts during the statistical examples. But it wouldn’t hurt if you bought a book on intro R programming and worked through it on your own time.

Text: I’ve tried a few different textbooks, but none of them really work perfectly. We’ll refer to a new book by Ben Bolker, “Ecological Models and Data in R” for further information and references. Another book I’ll refer to sections of is Gelman & Hill, “Data Analysis using Regression and Multilevel/Hierarchical Models”. You can buy both books from Amazon.com, if you want them.

Grading: I will score your lab assignments. There will be one lab assignment each week. Grading will be fairly gentle, however, as students will approach this course with different levels of relevant training and experience. I’m more interested in improvement than perfection.

EPHE Core Schedule of Lectures and Assignments

ECL 298-008 (40087)

Time: TR 12:10-1:30
Room: Storer 2342

Introduction

Jan 5 – The unification of the social sciences (Richerson) – Read: Gintis Unity BBS

Economic Approaches to Human Behavior (Sanchirico/Springborn/Lin)

Jan 7 – Theory of consumer/firm, overview of economics and fields (Sanchirico)

Jan 12 – Public Goods and Externalities (Sanchirico/Lin)

Jan 14 – Dynamic Resource Models (Lin)

Jan 19 – Decision-making under uncertainty (Springborn)

Cultural Evolution (Richerson/McElreath)

Jan 21 – The nature of culture (Richerson)

Jan 26 – Modeling cultural evolution (Richerson/McElreath)

Jan 28 – Analyzing cultural evolution (Richerson/McElreath)

People and Conservation (Darwent/Borgerhoff Mulder)

Feb 2 – Many roads to conservation (Borgerhoff Mulder)

Feb 4 – (Darwent)

Feb 9 – TBD

Feb 11 – TBD

Cultural Ecology/Political Ecology (Orlove/Winterhalder)

Feb 16

Feb 18

Feb 23

Feb 25

Environmental Policy and Politics (Lubell)

Mar 2 – Policy Tools

Mar 4 – Overview of US Environmental Laws

Mar 9 – Theories of the Policy Process

Mar 11 – Frontiers in Policy Research

Adrian Bell’s page

I am a 4th year student in the Graduate Group in Ecology at UC Davis. My primary advisor is Peter Richerson.

Please take this link to my webpage

Interests

I am intensely fascinated by biological and cultural evolution, the fundamental difference between the two, and the implications for the interaction of the two systems.  I study how ecological and biophysical factors exert selective forces on human institutions, and how the study of cultural evolution might aid in the search for sustainable (and durable) institutions.

PhD Research

I am currently analyzing and writing my dissertation in Human Ecology at U.C. Davis.  My research  tests the influence of ethnic/caste diversity and hierarchy on a traditional cooperative irrigation system in southern India.  A sizable literature demonstrates that ethnic diversity negatively effects public goods provision in realms such as public schooling, policing, and environmental management.  This literature might overlook the importance of inter-ethnic relationships in determining cooperative regimes within societies.  I use a range of methods, from  traditional ethnography, to surveys, to quantitative experimental games to measure human behavior.

Recent Research

I study the interaction between cultural processes and environmental conditions in Southern India. Because cooperation is a key component of sustainable resource management, I measured social norms of fairness and sharing in a small group of day laborers in Tamil Nadu to measure the strength of three social preferences: self-interest, equality, and social efficiency. The results suggest that people have preferences for how to distribute wealth that go beyond classical economic utility maximization, and include concern for how others are treated. Future research of this type will be used to measure cooperation between social groups as a component of successful sustainable environmental management. This work also will contribute empirical evidence to ongoing discussions about the nature of human social preferences.

Web Pages

Family Site
Personal Site

Resumé

Resume in PDF

Previous Research

Water use and water quality in Bardiya National Park – Nepal

Spatially explicit ecological simulation of the Everglades – SFWMD

Social learning and environmental fluctuation – UC Davis

Social norms of sharing in day laborer population – Kodaikanal, India

This post is password protected. To view it please enter your password below:

Password:

Francisco J. Gil-White. 2001. Are Ethnic Groups Biological “Species” to the Human Brain? Current Anthropology 42:515.
Download PDF

First, a review article (and letters responding to it):

Étienne Danchin et al. 2004. Public Information: From Nosy Neighbors to Cultural Evolution. Science 305:487.
Download PDF of review
Download PDF of letters and response

Second, a specific case study of putative socially-transmitted tool use in dolphins.

Krützen et al. 2005. Cultural transmission of tool use in bottlenose dolphins. PNAS 102:8939.
Download PDF 

Just one reading for this week. It’s long enough itself, and it has some commentary you should read, too.

Scott Atran, Douglas Medin, Norbert Ross, Elizabeth Lynch, Valentina Vapnarsky, Edilberto Ucan Ek’, John Coley, Christopher Timura, and Michael Baran. 2002. Folkecology, Cultural Epidemiology, and the Spirit of the Commons (with Commentary). Current Anthropology 43:421-450.
Download PDF 

Hahn and Bentley. 2003. Drift as a mechanism for cultural change: an example from baby names.
Download PDF

Fryer and Levitt. 2004. THE CAUSES AND CONSEQUENCES OF DISTINCTIVELY BLACK NAMES.
Download PDF