I've just finished teaching a one-quarter undergrad intro stats class
for the second time, using a combination of Verzani's "Using R for
Introductory Statistics" and Gonick and Smith's "Cartoon Guide to
Statistics." The two complement one another usefully: Gonick and
Smith is a nonthreatening introduction to a given topic, and Verzani
covers the same ground in more depth with programming examples. (And
on a purely strategic note, Gonick and Smith also makes it difficult
for students to claim that the course material was too challenging....)
I'm committed to using R because I want them to be able to use their
skills once they leave school with minimal difficulty and expense.
But I am realizing (and Jake seems to be experiencing this as well)
that the course is rapidly dominated by R programming tutorials,
usually at the expense of a focus on the nuts and bolts of statistics
-- throwing lots of R and lots of equations at them is a recipe for
panic. I've settled into a pattern of showing them the equations,
explaining how they work and why they make sense, but emphasizing more
applied aspects in exams, and that seems to make everything manageable.
Bear
On Mar 23, 2009, at 11:20 PM, Jake Bowers wrote:
> Dear All,
>
> I'm teaching undergrad stats for the first time this term and have
> been quite happy with the current draft of Danniel Kaplan's book:
>
> http://www.macalester.edu/~kaplan/ism/
>
> The book is worth a look. It is unconventional and leaves out many
> topics (no contingency tables, no simple differences of means) and
> adds others (hypothesis testing and confidence intervals are covered
> mostly using the bootstrap and permutation techniques). So far, I
> am happy with the mix of what he includes and excludes because I
> think the book offers a way for undergraduates to analyze data in
> pursuit of answers to their own questions about politics and
> society. And some basic skill in using data to pursue questions is,
> as I see it, the most important outcome of this class (the second
> most important outcome might be to get them interested enough to
> take a "real" statistics class which bases inference on the central
> limit theorem rather than resampling, and which covers the analysis
> of tables and other forms of testing not depending on the linear
> model and which uses linear algebra and calculus to unpack the
> linear model).
>
> My class has some freshmen as well as some seniors. We are using R
> and are following the book rather closely this first time around. I
> think this course (currently with about 40 enrolled) with the kinds
> of assignments I've designed (all of which emphasize writing,
> computing, and interpretation) requires a creative and energetic TA
> (a role played extremely well by Aya Kachi this term).
>
> Our biggest challenge so far is making sure that people know (1)
> what a zip file is, (2) how to find said file on their computers
> once it arrives from the web (i.e. how to find it using their
> Windows GUI and also how to tell R where the file is), and (3) how
> to make sure that R interprets the spss/sas/stata/csv files
> reasonably. Otherwise, they are doing a good job choosing
> interesting datasets, making interesting plots, and talking
> intelligently about coefficients from linear models. [Confidence
> intervals and the bootstrap arrive next week.]
>
> As I understand the role of this class in our department, PS majors
> must take one of a set of five courses before they can graduate.
> This course is one of those five. Thus, it is not directly required,
> but can be used to fulfill a requirement.
>
> We don't currently have a separate research design class for
> undergraduates in general. The few students who write senior honors
> theses in our department have some special classes which cover
> research design, as I understand it.
>
> I'll be interested to hear what other folks are doing.
>
> Jake
>
>
> Jake Bowers
> Assistant Professor Dept of Political Science
> Research Scientist, National Center for Supercomputer Applications
> University of Illinois @ Urbana-Champaign
> http://www.jakebowers.org
>
>
>
> On Mar 23, 2009, at 3:02 PM, Paul Gronke wrote:
>
>> Colleagues
>>
>> It sounds like there is widespread interest in having this
>> conversation. The old thread is out of date according to Jeff Gill
>> because some important new texts have been released. Just to toss
>> out
>> a few questions. Add yours and any thoughts or comments.
>>
>> 1) What text do you use for introductory undergraduate stats?
>>
>> 2) What computer program / web program do you use for training the
>> students (if any)?
>>
>> 3) What text do you use for introductory research design (if a
>> separate course)?
>>
>> 4) Are either or both of these courses required of your majors?
>>
>> 5) What kind of school (R1, R2, LAC, other, private, public)?
>>
>> And any other comments. Thanks!
>>
>> --
>> Paul Gronke
>> [log in to unmask]
>> Professor of Political Science
>> Reed College
>>
>> **********************************************************
>> Political Methodology E-Mail List
>> Editors: Melanie Goodrich, <[log in to unmask]>
>> Xun Pang, <[log in to unmask]>
>> **********************************************************
>> Send messages to [log in to unmask]
>> To join the list, cancel your subscription, or modify
>> your subscription settings visit:
>>
>> http://polmeth.wustl.edu/polmeth.php
>>
>> **********************************************************
>>
>>
>
> **********************************************************
> Political Methodology E-Mail List
> Editors: Melanie Goodrich, <[log in to unmask]>
> Xun Pang, <[log in to unmask]>
> **********************************************************
> Send messages to [log in to unmask]
> To join the list, cancel your subscription, or modify
> your subscription settings visit:
>
> http://polmeth.wustl.edu/polmeth.php
>
> **********************************************************
Bear F. Braumoeller
Assistant Professor
Department of Political Science
The Ohio State University
2168 Derby Hall, 154 North Oval Mall
Columbus, Ohio 43210-1373
**********************************************************
Political Methodology E-Mail List
Editors: Melanie Goodrich, <[log in to unmask]>
Xun Pang, <[log in to unmask]>
**********************************************************
Send messages to [log in to unmask]
To join the list, cancel your subscription, or modify
your subscription settings visit:
http://polmeth.wustl.edu/polmeth.php
**********************************************************
|
