[Scipy-organizers] Diversity blog post

[Bennet Fauber]

Andy,

Thanks for including me on this.  I'll apologize ahead of time for the
length.  I hope haven't strayed too far afield in what follows.

I'd like to see a regular track at SciPy 2015 that is really meant to
be more like training than conferring.  I think that would be more
successful than segregating beginning workshops at the beginning.
Here are the things I'd dream about trying to do in it.


-- Make the 'training track' parallel to and on an equal footing with
the 'presentation track'.  That way people who are there to learn can
do so all the days of the conference, not all crammed into a couple at
the beginning.  And they can step aside to go the regular conference
as desired.

-- Limit the amount of training given in one day to something like
three hours.  There's only so much a beginner can process in one day.
More time should be spent consolidating what's been learned than
forging ahead to new topics before the last has had a chance to
percolate.  Perhaps BOFs for the regular people could happen
concurrently, and training is like an extended BOF of people who want
to learn.

-- If possible, configure the rooms to encourage interaction.  Hold
the workshops in a room with cafe seating, three people to a smaller,
round table (think four-top).  Mix up the tables from workshop to
workshop so people don't sit with the same two people the whole time.
(We do our precalculus and calculus classes here like this, and it
seems to work.)

-- There should be a homework project every day.  Projects should be
presented as talks the following day.  Some people hate it, but I
think it's good to do homework in groups of three to four people.
This is about community, and that means sitting with people, talking
to people, working with people.  Let's provide an atmosphere and
activities where we give people a chance to try that.

-- There should be a separate space, available in the
afternoon/evening where the people who are working on the training
homework can have their own 'sprints' and be with people of
like-interest and more-alike ability.  I would much rather have gone
off to work on a beginning- intermediate- level homework problem in a
room with a half-dozen other people who were just learning python than
going to the happy hours.  The happy hours are good for people who
already know each other.  For me, it's just another party at which I
know no one.

-- There should be some presentations of work that isn't 'awesome' but
that shows some real examples of python getting work done by people of
middling ability; something that can provide an 'Ah, Hah!' moment for
a beginning or intermediate python programmer.  Something that looks
within reach of someone still struggling with list comprehensions.
This isn't going to appeal to the main conference attendees, unless
they're interested in teaching python to people.  Ideally, some of the
topics shouldn't require three years of college math.  This, and the
homework projects, are, I think, the hardest part to get, and
especially to get right.

-- Maybe no one else cares, but I'd like to have a bit more Sci at
SciPy.  We have a thing in Ann Arbor called 'Nerd Nite', which are
short talks about some subject of interest to the presenter, typically
given at a bar.  Topics I've seen are the flu pandemic, brain imaging,
how does malaria work.  I think it would be fun to have some 'my work
for a lay audience' talks.  Kind of like lightning talks, but with
some time in between for people to ask questions and talk to each
other.  Again, I think these would be better done in a smaller room
with cafe seating at much smaller tables.  It should feel more like a
poetry slam, or karaoke, or something.


I may well be projecting my own wishful thinking on an unsuspecting
world because I never got to go do Outward Bound for a week in the
Rockies, and I never got to go to the cool science summer camps, and I
never got the military boot camp (though I think I'm actually pretty
glad to have missed that one), and at some point, I wanted to do all
those things.

I'd want to attend a conference that had a track like that.  I think I
could make that fun, and productive, and I think something like that
has a good chance of kickstarting the nucleus of a new generation of
collaborators, given a little care and watering.

I also think that by providing a more constructive and welcoming
environment for the less advanced, in general, perhaps it would change
the overall tenor and feel of the conference to one where people who
are not already part of a project can come to _start_ being involved
in one, or just plain start one.

The person who's been most helpful to me is Julia Evans
(http://jvns.ca/), and this

http://jvns.ca/blog/2013/02/27/graphing-bike-path-data-with-ipython-notebook-and-pandas/

is the sort of thing that I think would be awesome to present in an
afternoon talk to a beginning/intermediate audience after a morning of
workshops.  It's enticing, interesting, and it's the kind of project
that a beginner could look at and say:  Hey!  I could do that.  In
fact, it turns out, I can do that.  I'd probably have given up without
that example.  She was also receptive when I asked her question about
it, and I even got to submit a bug to the Canadian Weather service
about an off-by-one error on their web site.  How cool ist that?!

That's the kind of experience I'd like to help people have.

The main problem I see with a conference like that is, how will people
pay to go to it?  Many of the people who would be best served don't
have funding, or don't have time, etc.  Maybe parts of that can be
made so people can take only a day or two out of the whole program and
still get to the conference part that provides the budget
justification?  Would some sort of certificate help?

I have some time and energy I can contribute.  I can work on bits of
the outline above, or if people have other ideas and I can contribute,
I could do that instead.

Thanks, again, for the opportunity to put my oar in the water.

                          -- bennet

#pragma
    printf("The views expressed here are mine and not those of my employer.\n");

-- 
University of Michigan
Ann Arbor, MI 48019
1-734-764-6226