OMnI Projects

Year 2 cohort

Site for sharing information about ongoing progress on projects.  Requires login.

 Albert Ortiz         Economics

This project will develop a two-period consumption/savings model to present the basic factors behind intertemporal consumption.

Among the factors to be studied are:
(1) The link between consumption and permanent income as opposed to current income.
(2) The role of preferences.
(3) The role of interest rates and access to credit.
(4) The storage/investment technology.

Among the elements that students will learn from this exercise are:
(1) How to set-up an economic model.
(2) Alternative solution methods.
(3) The definition of equilibrium and the elements of a solution.

 Zeb Page                Geology

At present we have no computational modeling component in our gateway lab sequence. Tectonics and Climate Change are two topics well suited to modelling that would make a improvement in our lab sections.

 Tracie Paine / Michael Loose            Neuroscience

I would like to introduce modeling into the Introductory Neuroscience lab.  Two components of the course readily avail themselves to discussion regarding computational models: 1) the action potential and 2) learning theory.  Currently we use a computer program to simulate (model) how changing ion concentrations on the inside and outside of a neuron effects the likelihood than an action potential occurs.  Although we use a model, nothing is currently said about the model, moreover the utility of models as a tool is not discussed.  Thus, in this laboratory, I would like to develop the course materials to discuss models in general and then apply that knowledge to model of the neuron and an action potential.  In the process I may choose to change the program that we are using to model the action potential. 

Later in the course we conduct an experiment whereby rats learn to lever press to obtain reward.  Numerous empirical studies have shown that varying the frequency, variability or timing of reinforcement can produce distinct patterns of lever-pressing behavior.  Moreover, the different schedules of reinforcement are more or less amenable to extinction (i.e., a decrease in the behavior in the absence of the reward).  I would like to develop a model that could be used to explore some of these principals.

Importantly, the two models that I propose to develop would focus on different levels of analysis (single neuron vs whole animal).  As a result, these models would allow students to see that models are applicable to many areas of neuroscience. 

 Aaron Santos                Physics

In most introductory physics courses, the vast majority of forces considered are constant.  The reason for this largely practical: it allows one to solve problems algebraically by considering (at most) a quadratic equation.  This approach suffers two major drawbacks.  First, most forces in nature are not constant, so students are missing out on some very rich physics phenomena.  Second, solving problems in this way gives the false impression that all problems can be solved algebraically rather than introducing using calculus which is often necessary.  By allowing students to computationally model forces that are not constant, we can introduce them to methods in computational physics, explore deeper problems, and qualitatively familiarize them with the types of systems they'll be solving in upper level classes.  Incorporating this module will also prepare introductory students to begin physics research projects early in their academic careers.

 Amanda Schmidt                 Geology

There are two modeling units I would like to implement in the class. The first is a classic hillslope diffusion model. There are pre-existing versions in Excel that I hope to use. It shows how hillslopes lower in slope over time through diffusion processes such as rain splatter. It is commonly used in geomorphology classes and should be easily adaptable to this intro class.

The second is a landsliding model. I have thought of a few ways to do this. One is to use an ArcGIS model developed at the University of Washington that looks at how much rainfall it takes to cause a landslide. This model deals with spatial distribution of landsliding. They have a package of data where students can play with a partially finished model and change parameters. The other is to do an infinite slope model where you look at how different soil parameters cause landsliding on a specific slope due to rainfall. It would allow for students to adjust rainfall and cohesion to see when slopes fail. I think we could do a nice job showing what earthquakes do to slope cohesion with both causing it to fail right away and with lowering cohesion so future failures are larger.

 Jordan Suter             Economics

I plan to incorporate a section on modeling when I cover the economics of common property resources (CPRs). A resource is considered a CPR if its use by one agent reduces the quantity or quality of the resource available to other agents. In particular, I intend to model the use of a groundwater resource by multiple users. In this case, if private water users follow their private incentives they will extract more water than what would be considered economically efficient. This is because the users ignore the costs that they impose on other users of the groundwater resource, who must now expend more effort extracting the water from the ground.

The model that I would like to introduce will illustrate how specific hydrologic characteristics of the groundwater resource influence the consumption and availability of the resource over time. It will also enable students to compare the effects of different types of economic policies (ie a pumping tax or quantity limit) that could be introduced.

Continuing from Year 1

Daphne John, Robin Salter, Cindy Frantz, Hanna Joy,

Year 1 cohort

Site for sharing information about ongoing progress on projects.  Requires login.

 Robin Salter        


 Spread of Disease

Description:  I will introduce modeling in Human Biology, a general audience course (BIOL 090) that enrolls ~40 students.  An important component of this course has been to research and discuss the mandatory nature of childhood vaccination in this country.  I plan to use the model to demonstrate the importance of vaccination in preventing an outbreak/epidemic of childhood diseases.   Human Bio students will explore the relationship between infection rate (i.e., how contagious a particular pathogen is) and vaccination rate on the spread of disease within a population.   

Immunity to some pathogens is lost over time.  Students in my upper-level Immunology course (BIOL 327) will use an expanded version of the model to explore how the duration of immunity affects disease transmission through a population.   

Since many of my students are interested in public health careers, my ultimate goal is to show how modeling can be used to inform important policy questions. 

 Ellis Tallman


 Economic Growth

I plan to revise the treatment of models of economic growth. This topic is essential to modern macroeconomics. I have found that it often is hard for students to grasp onto growth issues as we shift from analysis of levels of economic activity to the analysis of the dynamics of economic growth. In this treatment, I want to emphasize the theory of endogenous growth from the investment in human capital as well as from investment in research and development.

The topic should be more accessible in a setting that exploits computational modeling for demonstrating the comparative dynamics of alternative policies.

 Daphne John


Labor Markets

Two potential topics that I could revise to incorporate computational modeling would be 1) group formation and 2) participation in the labor market and associated outcomes.
The first part of this introductory course focuses primarily on micro-level social behaviors. One text used on this topic is Julie Bettie's _Women Without Class: Girls, Race and Identity_. This ethnography provides insight into how and why particular individuals identify with each other and form race/class/gender specific groups in a high school. Computational models could be constructed to illustrate the salient factors involved in group formation and persistence. For example, how does family socioeconomic status impact a student's perception of the value of formal education and thus affect the peers with which she associates?
The second part of the course focuses on macrostructural behaviors, such as labor market composition, the functions of unemployment and access to economic resources. In particular, models of social stratification are examined in order to illustrate how inequality persists. One text read in the class is Sudhir Venkatesh's _Off the Books: The Underground Economy of the Poor_ which illustrates how and why an underclass responds to economic barriers and formulates and informal economy. Topics of economy and poverty seem well suited for computational modeling exercises given the availability and quality of empirical data. For example, longitudinal data (like the PSID) might be used to illustrate the variables involved in movement from employment to participation in public program subsidies once one's employment status changes.
Additional Information
I am trained as a quantitative sociologist and could efficiently integrate computational modeling into my introductory course. I also teach the research methods courses in the department. I am very interested in this project and look forward to participating.

 John Petersen    

 Environmental Studies

Intro to simulation for biologists

I am a good part of the way through developing a simulation modeling lab for Bio102. Indeed various components have already been user tested with students and faculty in that class over a two year period. Basically the concept is to introduce students in this class to the power of simulation modeling. So far it is STELLA based, but I might consider switching it to a freeware program which would make lab logistics a bit easier.

 Rumi Shammin

 Environmental Studies

 Environmental Analysis

 This course is designed to be a survey of various types analytical approaches available and used in environmental analysis. While it is designed for students with limited mathematical preparation (often social science students) in order to offer them a more accessible approach to quantitative skills. However, in previous semesters, natural science students and others with strong mathematical training also seemed to benefit from the course - as it introduced them to new ways of framing questions and approaching solutions. The course also introduces students to a wide range of skills (statistical analysis, modeling, spatial analysis, etc.) that may be further studied in upper level courses.

I maintain a hands-on approach in this course. In the modeling component of the course, I currently go over introductions to modeling and have students develop and solve models by hand (and using the spreadsheet as a secondary tool) - directly working with the underlying equations (with fewer iterations, of course).

Under this project, I would like to develop a two-week module (4 classes) on modeling. The topics covered will be:

1. Introduction to modeling
2. Stock and flow analysis
3. Population analysis
4. Benefit-cost analysis

Currently, I use Excel spreadsheets partly for #2 and more extensively for #4 mentioned above. As part of this project, I will develop a short introduction to Stella as part of #1, Stella models for #2 and #3, and refine the existing Excel spreadsheet based model for #4. I will also update related assignments for items #2, #3 and #4 based the new teaching modules.

I have more experience of using Stella and hence have chosen it as the desirable platform for #2 and #3. However, I have also used Vensim as a graduate student and would be open to the possibility of using it instead.

 Dennis Hubbard


 Reef Ecology

I want to create a model that will illustrate the relationships between grazing fish, nutrients and algae/coral balance on a modern reef and one that has been impacted by either overfishing or nutrification/pollution. This is a relationship that is difficult to teach abstractly, and by having students change variables and examine the changing outcomes, it should help students who learn differently to come up with their own way of reaching conclusions and understanding concepts. The model would be used in conjunction with existing lectures that will be modified to better compliment the new hands-on component.
Additional Information
I hope to also investigate ways to develop models for my upper-level courses at the same time, so I hope there will be opportunities to gain some experience with several different tools and discussing the possibilities of working with someone to start to put together rudimentary versions of the models that will replace older ones that do not run on newer platforms.

 Michael Parkin


 Predicting Voter Choice

Both Polt 100 (Introduction to American Politics) and FYS 169 (Campaigns and Elections) offer an opportunity to model various aspects of campaign behavior. My plan is to show students how regression can be used with national survey data to design a predictive model of vote choice. I would then multiply the coefficients by certain parameters to show how events in a campaign can affect candidate support and vote choices. I would like to, if possible, create a computer simulation that allows users to easily manipulate various parameters that would then produce differing election outcomes. One could, for example, model how an economic downturn near the end of an election affects the election outcome under certain circumstances.
Additional Information
I would imagine that I would be the only member of the Politics department interested in this. As such, I could help to introduce computational thinking to a relatively popular major.

 Cindy Frantz      



For PSYC 100: a social-cognitive phenomenon that represents overlap in my and Joy Hanna's expertise.
For PSYC 300: sampling distribution of the mean; the effect of violated assumptions on ANOVA.
Additional Information
Intro Psych is moving from a 3 person course to a 2 person course Spring 2011. Joy Hanna and I will be the first 2 person team, and as we work on restructuring the course we're interested in together implementing a several-class modeling component to Intro. In the past I've done some simple exposure to interactive-activation connectionist style models of word recognition, but I imagine that Joy and I would like to have a modeling component that could bridge sub-areas of psychology, for instance we plan to model the genetic and environmental contributions to intelligence. We think it is important for both of us to participate in the workshop in order to effectively implement the changes.

 Joy Hanna



There are a number of topics I could imagine revising to fit a modelling approach in Intro Psych. In Cognitive I work with a simple perceptron dealing with conceptual learning to expose students to the structure and processes of connectionist models, and especially the mechanics of learning. I'd like to link my approach in Intro to whatever I do in Cognitive at some level, but the area of application is wide open. I imagine working with a connectionist style model since that is dominant in my field, but could apply it to any of the sub-areas of psychology.

In Cognitive there is a stand-along set of classes on modeling that are somewhat removed from the content topics, but I'd like the students to be more actively involved in exploring model parameters and learning. It would be great if I could have something that the students could "damage" in various ways to try to replicate behavioral changes that result from aging, brain damage, etc.
Additional Information
My priority for this workshop is modifying Intro Psych to include a computational component.

Intro Psych is moving from a 3 person course to a 2 person course Spring 2011. Cindy Frantz and I will be the first 2 person team, and as we work on restructuring the course we're interested in together implementing a several-class modeling component to Intro. In the past I've done some simple exposure to interactive-activation connectionist style models of word recognition, but I imagine that Cindy and I would like to have a modeling component that could bridge sub-areas of psychology, for instance with both cognitive and social implications. We think it is important for both of us to participate in the workshop in order to effectively implement the changes.

 Nancy Darling


 Understanding Sampling

I would like to enhance student understanding of measurement error and sampling in Psych 200/201 and improve student understanding of intervention, sampling, and error in my lab and seminar in adolescent development.

 Viplav Saini


 Demand and Supply Analysis

A central question in economic science is understanding how prices are determined. Demand and supply analysis is the standard tool used to answer this question. The idea behind demand and supply analysis is that we aggregate the decisions of a large number of consumers and producers of a good into a market demand and a market supply curve. The price is then determined through the interaction of the demand and supply curves.

Since the treatment of most topics at the Econ 101 level is deliberately non-rigorous, students sometimes tend to treat the aggregate analysis as a black box, failing to see how the incentives and actions of individual economic agents determine an aggregate price.

I plan to revise the Demand and Supply analysis portion of the course using a computational model of the demand and supply sides of a market. We will first consider the maximization problem facing an individual consumer. We will then aggregate over the choices of a large number of consumers to derive a market demand curve. Similarly, we will begin with the maximization problem of an individual producer and then derive a market supply curve. In order to add a dash of realism we will allow both consumers and suppliers to be heterogeneous in terms of their willingness to pay and marginal cost respectively.

We will use the setup to show (a) how a market price is determined, (b) comparative statics of taxes, (c) comparative statics of shifts in the demand and supply curves, (d) welfare effects of price ceilings and price floors, and (d) the effect of externalities on consumer welfare.

The key learning goal of this project will be for Econ 101 students to unpack the demand and supply model in terms of understanding the computations that make it work. This tool can also serve as part of a larger toolbox of more intricate general equilibrium models in economics. Excel, Mathematica, Matlab