Thanks for the details Uwe.

It kind of makes sense but I feel that this should have to be quite fine
tuned for it to work properly. i.e. you are mixing up basic command
line usage/concepts with a whole program (git) running *inside* the
command line. Except if they already have some command line basics
before, in which case I can see everything fitting more clearly.

Have you developed your own lessons/story for this in order to keep the
flow or you just interleave the SC ones?

Iñigo




On Mon, 5 Mar 2018 23:25:52 +0000
"Hilgert, Uwe K K - (hilgert)" <hilg...@email.arizona.edu> wrote:

> Thank you for your response, Inigo. Actually, in our experience
> teaching git/GitHub as a separate, isolated item had come with an
> unnecessary cognitive burden that we reduced by integating it with
> Shell. The audiences we worked with were researchers (grad students,
> postdocs, staff, faculty) and were driven to acquire additional
> skills to help them with their research and data analysis objectives.
> Not by the intent to go out and "learn Unix," or "learn programming,"
> or "learn git/GitHub." Their primary objective was and is to solve
> problems and to learn whatever tools are needed to address a
> challenge. If then the challenge becomes to access a file shared by
> the instructor and alter it as part of the Shell session, then
> accessing this file on GitHub, messing with it and pushing altered
> versions back to their own GitHub accounts turns actually into rather
> effortless adoptions of the practises required to use git to achieve
> the objectives of accessing, updating and sharing a data file.
> 
> So, from my experiecne having observed this at work during a few runs
> of the workshop my anser to your question about "too much cognitive
> load" is "No."
> 
> Does this make any sense?
> 
> Uwe
> 
> 
> ------------------------------------------------------------------------
> Uwe Hilgert, Ph.D.
> 
> Associate Research Professor
> Director of Industry Relations, Workforce Development & STEM Training
> BIO5 Institute (http://www.bio5.org), The University of Arizona
> (http://www.arizona.edu) O 520-626-1367  |  F 520-626-4824
> 
> 
> -----Original Message-----
> From: Inigo Aldazabal Mensa [mailto:inigo_aldaza...@ehu.eus] 
> Sent: Friday, March 02, 2018 1:51 AM
> To: Hilgert, Uwe K K - (hilgert) <hilg...@email.arizona.edu>
> Cc: Byron Smith <bsmit...@gmail.com>; Software Carpentry Discussion
> <discuss@lists.software-carpentry.org> Subject: Re: [Discuss] Fwd:
> Default git/python lesson order
> 
> Hi all,
> 
> Sorry for the late reply, it seems we are having email problems
> here...
> 
> About integrating shell and git... isn't this too much cognitive load?
> 
> And about doing Git as the last lesson, I think Byron raises some
> really good points there. We'll have to think more about this.
> 
> 
> Iñigo
> 
> On Thu, 1 Mar 2018 22:49:40 +0000 "Hilgert, Uwe K K - (hilgert)"
> <hilg...@email.arizona.edu> wrote:
> 
> > Please don’t forget the option we have adopted at the UA, that is
> > to integrate shell and git.
> > 
> > ----------------------------------------------------------------------
> > --
> > Uwe Hilgert, Ph.D.
> > 
> > Associate Research Professor
> > Director of Industry Relations, Workforce Development & STEM
> > Training BIO5 Institute (http://www.bio5.org), The University of
> > Arizona (http://www.arizona.edu) O 520-626-1367  |  F 520-626-4824
> > 
> > From: Discuss [mailto:discuss-boun...@lists.software-carpentry.org]
> > On Behalf Of Byron Smith Sent: Wednesday, February 28, 2018 9:17 AM
> > To: Software Carpentry Discussion
> > <discuss@lists.software-carpentry.org> Subject: [Discuss] Fwd:
> > Default git/python lesson order
> > 
> > On Tue, Feb 27, 2018 at 11:26 AM, Maneesha Sane 
> > <manee...@carpentries.org<mailto:manee...@carpentries.org>> wrote:
> > 
> > I agree, I don't see a big difference in the two schedules Iñigo 
> > proposed.
> > 
> > My intuition (supported by some anecdotal evidence) is that
> > learners will be worn out on Python/R if they're taught in one
> > contiguous block.  I much prefer splitting the programming language
> > across two days not only to give the instructor a break, but also
> > so that learners have time to stew on it.  Teaching shell and git
> > on separate days has a similar effect, since the latter is a good
> > review of the former.
> > 
> >  On Tue, Feb 27, 2018 at 11:26 AM, Maneesha Sane 
> > <manee...@carpentries.org<mailto:manee...@carpentries.org>> wrote:
> > [...] by the end of Day 2, instructors and learners alike are quite 
> > tired making it harder to introduce an entirely new concept like
> > git that afternoon. [...] So I'd recommend against doing git the
> > afternoon of day 2.
> > 
> > On Tue, Feb 27, 2018 at 11:55 AM, April Wright 
> > <wright.apr...@gmail.com<mailto:wright.apr...@gmail.com>> wrote:
> > 
> > I agree with Maneesha about not introducing git in the second 
> > afternoon. Git can be challenging, and hard to understand in terms
> > of the motivation for use.
> > 
> > I'll argue for git as the last session; I'm a little bit surprised 
> > that no one else has.  I like putting git last because:
> > 
> >   *   it reinforces and complements the "reproducibility" thread
> > that's implicit in all of the lessons;
> >   *   you can refer back to the other lessons (writing code) to
> > motivate version control;
> >   *   you can calibrate the cognitive load based on how engaged
> > learners are;
> >   *   some of the alternative storylines have opportunities for
> > humor, and collaboration is always exciting, both of which do a lot
> > to keep learners' attention. We've used this schedule (shell,
> > python, python, git) in almost all of the workshops I have been
> > involved with. It's kinda interesting that I haven't overlapped
> > with the other camps a whole lot.
> > 
> > Not that we need to enforce one on instructors, but it's
> > challenging to take an evidence-based approach to recommending a
> > schedule.  Few instructors have a large enough sample size to
> > overcome the variability between workshops and confidently endorse
> > one of the two. Can we do any assessment based on exit surveys?
> > Which schedule gets the best ratings?
> > 
> > -Byron
> > 
_______________________________________________
Discuss mailing list
Discuss@lists.software-carpentry.org
http://lists.software-carpentry.org/listinfo/discuss

Reply via email to