R help - Oct 2007 - Input appreciated: R teaching idea + a way to improve R-wiki

Hi all,

I will be teaching a graduate-level course on R at CU Boulder next
semester. I have a teaching idea that might also help improve the R
wiki page... I wanted to know what you all thought of it and wanted to
solicit some advice about doing it.

During the latter part of the course, students will choose a topic of
interest (e.g., hierarchical linear modeling), and show how to achieve
it in R. They would present their findings to the class, and would
also be responsible for writing a concise but well-written "How To"
manual on the topic. These would be ~ 5-10 pages and would include
basic background of the statistical procedure and a commented example
with code in R. The goal would be for these to read like Baron & Li's
"Notes on the use of R for psychology experiments and questionnaires."

Originally I was going to post these as PDFs on my own web-page and
let them grow into a compendium of how-to manuals as I teach this
course over the years. However, perhaps a better idea, and one that
probably benefits more people, is to have my students post their short
manuals (not as PDFs but rather typed in) on the R-wiki page.

Does this seem like a good idea to folks?

Another question has to do with how barren the current R wiki page
is... is it still being actively developed or has the community given
up on it?

Finally, any thoughts on where on the R-wiki site we should post our
"How To" manuals? The "tips and tricks" section seems to
barely be
more than snippets of conversations from this list-serve (often sans
the context). My guess is that the "Guides" section is where these
should go.

Your input would be most appreciated. Best,

Matt



-- 
Matthew C Keller
Asst. Professor of Psychology
University of Colorado at Boulder
www.matthewckeller.com

I think you need to see how things work before making any decision on
this.  While the principle seems OK, in a optimistic sort of way, you
may be a little disappointed by the outcome.  Some will likely be
superb, useful, well written and accessible.  Others, I suspect, will
fall short of this ideal, with some falling a fair way short.  That's
the way students learn, after all.  They should use these exercises to
straighten things out in their own minds, and some of them seem to have
rather twisted ideas, at least initially, even at "graduate-level".

Some people argue it's useful to see the learning process in action, and
some books I could mention seem to be written this way - but they don't
get very good reviews.  I just think there is a real danger here of
giving misleading and inefficient teaching materials a spurious cloak of
legitimacy, even if there are disclaimers all over it.  I see a need to
be very cautious about this, in other words.


Bill Venables
CSIRO Laboratories
PO Box 120, Cleveland, 4163
AUSTRALIA
Office Phone (email preferred): +61 7 3826 7251
Fax (if absolutely necessary):  +61 7 3826 7304
Mobile:                         +61 4 8819 4402
Home Phone:                     +61 7 3286 7700
mailto:Bill.Venables at csiro.au
http://www.cmis.csiro.au/bill.venables/ 

-----Original Message-----
From: r-help-bounces at r-project.org [mailto:r-help-bounces at r-project.org]
On Behalf Of Matthew Keller
Sent: Monday, 22 October 2007 9:45 AM
To: R list
Subject: [R] Input appreciated: R teaching idea + a way to improve
R-wiki

Hi all,

I will be teaching a graduate-level course on R at CU Boulder next
semester. I have a teaching idea that might also help improve the R
wiki page... I wanted to know what you all thought of it and wanted to
solicit some advice about doing it.

During the latter part of the course, students will choose a topic of
interest (e.g., hierarchical linear modeling), and show how to achieve
it in R. They would present their findings to the class, and would
also be responsible for writing a concise but well-written "How To"
manual on the topic. These would be ~ 5-10 pages and would include
basic background of the statistical procedure and a commented example
with code in R. The goal would be for these to read like Baron & Li's
"Notes on the use of R for psychology experiments and questionnaires."

Originally I was going to post these as PDFs on my own web-page and
let them grow into a compendium of how-to manuals as I teach this
course over the years. However, perhaps a better idea, and one that
probably benefits more people, is to have my students post their short
manuals (not as PDFs but rather typed in) on the R-wiki page.

Does this seem like a good idea to folks?

Another question has to do with how barren the current R wiki page
is... is it still being actively developed or has the community given
up on it?

Finally, any thoughts on where on the R-wiki site we should post our
"How To" manuals? The "tips and tricks" section seems to
barely be
more than snippets of conversations from this list-serve (often sans
the context). My guess is that the "Guides" section is where these
should go.

Your input would be most appreciated. Best,

Matt



-- 
Matthew C Keller
Asst. Professor of Psychology
University of Colorado at Boulder
www.matthewckeller.com

______________________________________________
R-help at r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide
http://www.R-project.org/posting-guide.html
and provide commented, minimal, self-contained, reproducible code.

Possibly.  I find it difficult to argue the case either way in the
abstract, though.  I think once you see some of the outcomes, it will
become clear which are good enough for posting and which are not.
 
Bill Venables.
 
 
  _____  

From: pietr007@gmail.com [mailto:pietr007@gmail.com] On Behalf Of
Ricardo Pietrobon
Sent: Monday, 22 October 2007 2:56 PM
To: Venables, Bill (CMIS, Cleveland)
Cc: mckellercran@gmail.com; r-help@stat.math.ethz.ch
Subject: Re: [R] Input appreciated: R teaching idea + a way to improve
R-wiki


Bill, very interesting comment.  However, do you believe that by posting
these tutorials on a wiki they could, even if initially faulty, be
improved by the community over time?  

Ricardo




On 10/22/07, Bill.Venables@csiro.au <Bill.Venables@csiro.au> wrote: 

	I think you need to see how things work before making any
decision on
	this.  While the principle seems OK, in a optimistic sort of
way, you
	may be a little disappointed by the outcome.  Some will likely
be
	superb, useful, well written and accessible.  Others, I suspect,
will 
	fall short of this ideal, with some falling a fair way short.
That's
	the way students learn, after all.  They should use these
exercises to
	straighten things out in their own minds, and some of them seem
to have 
	rather twisted ideas, at least initially, even at
"graduate-level".
	
	Some people argue it's useful to see the learning process in
action, and
	some books I could mention seem to be written this way - but
they don't 
	get very good reviews.  I just think there is a real danger here
of
	giving misleading and inefficient teaching materials a spurious
cloak of
	legitimacy, even if there are disclaimers all over it.  I see a
need to 
	be very cautious about this, in other words.
	
	
	Bill Venables
	CSIRO Laboratories
	PO Box 120, Cleveland, 4163
	AUSTRALIA
	Office Phone (email preferred): +61 7 3826 7251
	Fax (if absolutely necessary):  +61 7 3826 7304 
	Mobile:                         +61 4 8819 4402
	Home Phone:                     +61 7 3286 7700
	mailto:Bill.Venables@csiro.au
	http://www.cmis.csiro.au/bill.venables/
	
	-----Original Message-----
	From: r-help-bounces@r-project.org
[mailto:r-help-bounces@r-project.org ]
	On Behalf Of Matthew Keller
	Sent: Monday, 22 October 2007 9:45 AM
	To: R list
	Subject: [R] Input appreciated: R teaching idea + a way to
improve
	R-wiki
	
	Hi all,
	
	I will be teaching a graduate-level course on R at CU Boulder
next 
	semester. I have a teaching idea that might also help improve
the R
	wiki page... I wanted to know what you all thought of it and
wanted to
	solicit some advice about doing it.
	
	During the latter part of the course, students will choose a
topic of 
	interest (e.g., hierarchical linear modeling), and show how to
achieve
	it in R. They would present their findings to the class, and
would
	also be responsible for writing a concise but well-written "How
To" 
	manual on the topic. These would be ~ 5-10 pages and would
include
	basic background of the statistical procedure and a commented
example
	with code in R. The goal would be for these to read like Baron &
Li's 
	"Notes on the use of R for psychology experiments and
questionnaires."
	
	Originally I was going to post these as PDFs on my own web-page
and
	let them grow into a compendium of how-to manuals as I teach
this 
	course over the years. However, perhaps a better idea, and one
that
	probably benefits more people, is to have my students post their
short
	manuals (not as PDFs but rather typed in) on the R-wiki page.
	
	Does this seem like a good idea to folks? 
	
	Another question has to do with how barren the current R wiki
page
	is... is it still being actively developed or has the community
given
	up on it?
	
	Finally, any thoughts on where on the R-wiki site we should post
our 
	"How To" manuals? The "tips and tricks" section seems to
barely
be
	more than snippets of conversations from this list-serve (often
sans
	the context). My guess is that the "Guides" section is where
these 
	should go.
	
	Your input would be most appreciated. Best,
	
	Matt
	
	
	
	--
	Matthew C Keller
	Asst. Professor of Psychology
	University of Colorado at Boulder
	www.matthewckeller.com
	
	______________________________________________
	R-help@r-project.org mailing list
	https://stat.ethz.ch/mailman/listinfo/r-help 
	PLEASE do read the posting guide
	http://www.R-project.org/posting-guide.html
	and provide commented, minimal, self-contained, reproducible
code.
	
	______________________________________________
	R-help@r-project.org mailing list
	https://stat.ethz.ch/mailman/listinfo/r-help 
	PLEASE do read the posting guide
http://www.R-project.org/posting-guide.html
	and provide commented, minimal, self-contained, reproducible
code.
	





	[[alternative HTML version deleted]]

Just want to chime in and say that I think it's a great idea.  It is, 
after all, a wiki, and even the bad entries will serve to be something 
like stubs that can be expanded upon by others and they play with them.  
When using e.g. the Gentoo wiki, I have run across some well-organized 
entries (as, it seems, all of these would be) that are nevertheless 
incorrect and/or missing steps.  They have nonetheless been quite 
valuable, as they got me pointed in the right direction, and I could 
later come back and patch them up with what I'd learned.

Ari

.

I would love more quality online documentation around the same level  
as at UCLA (http://www.ats.ucla.edu/stat/; the Stata pages are  
fantastic), but I think I would like the first draft responsibility  
to fall to the well qualified instructor (Hi Matt!), and those with  
at least a Masters in statistics.

I chose to learn Stata in 2002, because the documentation was far  
superior. R documentation has improved immensely, and now there is a  
lot more material to get average scientists (like me) comfortable  
with using R packages and writing R functions. However, R  
documentation is still too terse with regard to well explained  
examples, compared to other software. You can find examples, sure,  
but they may not be so well explained in the function help file.

-Dave


> I appreciate the input. Off-list, someone suggested that I set up a
> class wiki, and have this be the first sieve. I could do some quality
> control there first (perhaps sending the link to this list serve at
> the end of the semester for others to check over), and then post the
> final manuals on the R wiki. I think its a good idea and am mulling
> it, but part of me asks: why not just post the (perhaps imperfect)
> manuals on the wiki and allow the wiki to do what wikis are supposed
> to do?
>
> I guess I resonated with Ricardo Pietrobon's point: the essence of a
> wiki is that it is evolving and self-correcting. Even to get something
> started over there would be an improvement. If people wait until they
> are 100% certain that everything is 100% accurate, a much diminished
> pool of people would post... The accuracy of wikis improves as more
> people post. In other words, I think that it is the number of posters,
> and not necessarily the signal:noise ratio, that drives wiki
> accuracy...
>
> Matt

-- 
Matthew C Keller
Asst. Professor of Psychology
University of Colorado at Boulder
www.matthewckeller.com

--
David C. Airey, Ph.D.
Pharmacology Research Assistant Professor
Center for Human Genetics Research Member