| Lecture: | Tuesday 10:00-11:30 Room Gund 505 |
| Section 1: | Thursday 10:00-11:30 Gund 516 |
| Web Site: | http://www.gsd.harvard.edu/pbcote/courses/gsd6322 |
| Instructor: | Paul Cote |
| Office: | Gund Hall 520 |
| Office Hours: | By Appoinment |
A university education should prepare students to handle and share information in a professional/scholarly manner. We handle information in order to better understand problems and opportunities. We mix information with ideas to form logical models. Models let us perform experiments and test hypotheses that we hope will create new understanding that has some validity or utility. If our understanding is to have impact in the world, we must learn to present information and ideas effectively with credibility. Credibility requires that we understand and properly attribute the sources of our information. Professionalism and conventions of scholarship require that we are able to hand off the source materials and procedures used in developing our new information so that they may be thourougly understood and even replicated by our collaborators, critics and successors. To this end, graduate students in university should be familiar with conventions and technologies for compiling, documenting and sharing data, information, and models; should have a critical understanding of the procedures by which information is transformed out of data; should have an outlook on how data, transformations and the resulting information are evaluated.
Linking independent pieces of data to create new information is an ancient craft that has been shifted into overdrive by information systems and the internet. Spatial databases and Geographical Information Systems are of particular interest in this regard owing to the fact that spatial referencing systems, even those in use thousands of years ago, are easily reconciled to create associations between information layers -- even if the original compilers of such information never envisioned this potential. Today vast amounts of information regarding the physical lay of the land, of the built environment, of cirulatory networks, natural systems, the regulatory environment, are being developed and maintained that have this capacity. As independently generated and maintained databases grow in number, the new information that may potentially be gleaned in the combination of databases grows exponentially. Particularly now, as the technical barriers to sharing information recede into history, each of us -- especially leaders in scholarship and decisionmaking -- have a responsibility to understand the issues and technical details involved with integrating and sharing spatially referenced information.
To create new information using data is easy. It is somewhat more difficult and much more important to understand the utility of the new information that we create or that is presented to us. A framework for evaluating data and models and the information derived from them gives us a basis for making the best possible models; for presenting information responsibly and effectively; and ultimately for having an appropriate level of confidence in our assumptions, recommendations and decisions. As with studies of rhetorical composition, new ideas created by transforming information with logical procedures can be evaluated by examining critical aspects of the data inputs and intervening procedures, along with the logical rationale behind their assembly. Developing a formal understanding and technical capacity for composing, sharing, and evaluating complex arguments that use spatial data is the goal of this course.
As in other logical endeavors, the problem of building new information begins with a reduction of a complex system to the elements that are considerd critical -- a coceptual model. The use of electronic information systems for mapping and modeling places is an exercise that makes this necessity very clear. The heart of all of the mapping and modeling work of this course lies in the effective development of conceptual models, instatiating our concepts and their relationships with data and procedures (data models), performing experiments on these, and portraying, evaluating and explaining our results. In many cases our results may not be all that useful for decision making. But even when this is true, the experience of having made the model and experiments should always produce useful information about the process of modeling itself. The development of working conceptual models and data models in the areas of spatial design and planning is a very important new mode of scholarship which has barely been recognized by senior scholars and practitioners and provides a very valuable skill for new graduates.
One of the most important skills to develop as a scholar is to be able to insert your ideas into the minds of others in a streamlined fashion. This involves establishing credibility, being very concise, and being aware of the ways that graphics interact with ideas. Making maps with GIS is actually very easy these days. But the art of making maps that communicate effectively requires skills that software does not handle in an automatic way. To make a map that communicates with credibility, one must know how! This underestanding will be the subject of the first and second projects.
A key aspect of scholarship, as opposed to just fooling around, is that scholarship involves advancing discourse and collaboration by organizing and sharing the resources and porocedures that one has found useful, so that others may start where you left off. An important aspect of this course will be for students to demonstrate their ability to organize data and processes so that they may be easily re-used by others. Projects deemed worthy in this regard will be posted in the The Harvard Dataverse Project
The Schedule of Lectures and Workshops provides an outline of the trajectory of the course. Each week we will meet for an hour and a half to build a conceptual understanding of spatial data and its uses. These lectures will often consider the social and institutional context; and the generative history of the fundamental building blocks of GIS. We will use this opportunity to practice the development of a critical outlook on the development of well founded GIS models. These lectures will provide references for understand many technical capacities and problem solving techniques and critical mentality that will be of use in your projects.
Each week we go through an exercise together in the lab. These exercises are a means of bringing everybody through some useful technique -- which will be directly applicable to your homework assignments. The in-class exercizes are intended as introductions to families of operations that can be carried out with software. The sample datasets and the cookbook-style instructions are intended to be something that can be worked through independently and applied to data that you may collect for your own studies. References provided in the workshop notes point to more complete documentation of the tools covered and other procedures that you may want to learn.
There are four projects that will be turned in for evaluation over the course of the semester. These are described in detail on the Assignments Page. The first required assignment is due the evening before the second meeting of class. Even if you are requesting late admission to the class, you still must do the first assignment immediately after applying.
Examples of Student Work from Previous Years
This is not an introductory computing class. Students should be experienced and comfortable with learning to use software by reading documentation, and with creating and organizing digital documents, and with keeping backups of all of their work. It is sad but true that, from the point of view of your clients and instructor, work lost due to failure to keep frequent backups is equal in value to work that was never done at all. THough it may be some consolation that things are usually easier and faster the second time you do them.
Presentation documents created for each assignment require textual descriptions so students should be comfortable writing in English.
The primary software platform used for the laboratory and independent work will be ArcGIS version 9.2. This software is available for use for GSD Students, or for other Harvard Users. All of the user manuals for ArcGIS 9.2 can be found at the Harvard Center for Geographic Analysis Web Site
There are some lab exercises and assignments that require access to file-system resources available only within the GSD. To access these resources, students will need to have a GSD computing Account and have thier computer on the GSD network, or use a public lab computer in Gund Hall.
All student work will be will handed back using the MyGSD courseware system. Use of this system requires a valid Harvard PIN. For more information, see Harvard PIN Request Site.
GSD6322 has a budget for Assistant Instructors this year! Normally there are none, and students aere forced to depend on my lab notes, software user manuals, on-line help, and their own powers of experimentation to get through problems (much like the real world!) TAs work during predefined office hours and in places specified in the Assitant Instructor Directory. Things with a TA are not much different than this, actually, except that if you are working during TA office hours, you may be able to get a nudge here and there to help you overcome sticking-points and obstacles that do come up when working out problems that come up even when working through fairly well documented procedures.
Nobody should rely too heavily on the TAs. This will tax thier limited resources (a few hours per week) and will render you a less effective user. You must not use the TAs as a substitute for reading the lab notes anf the on-line help for the software and manuals linked to our tutorials. In fact, to use GIS in the real world requires a good deal of patience, persistance and resourcefulness.
Have a problem when TAs aren't on duty? Send a message to the email discussion group set up exlusively for GSD6322 students and staff! Your message will be forwarded to the teaching staff and to the other members of the course list who have opted to recieve these messages. You can also browse the threads already discussed in this list to see if your question has already been answered.
It is especially encouraged that students contact the instructor for questions of research design. The difficulty of making GIS models can be greatly reduced by beginning with intentions and conceptual models that are well suited to available data and procedures. Please take the time to describe your intentions and questions in an email when you are setting up your appointment. This exercise substantially increases the value of face-to-face meetings.