Tip:
Highlight text to annotate it
X
Well, hello
my name is Jeff Grabill and I'm a faculty member
in a department called Writing, Rhetoric and American Cultures, in fact in August
I'll be Chair of that department so you will never see me again.
My creative activity will cease instantly.
Hopefully not. I also direct the research center
with Bill Hart-Davidson called Writing in Digital Environments and some
of the work that I'm going to talk to you about today
comes out of some research and development work that we've been doing
in that center over the last five or six years.
And in September I helped found an educational technology company called Drawbridge
that owns the technology we'll talk a little bit today. It's a company
that's also owned by the university and so starting that venture something
that Kari can talk a little bit more
intelligently about, but I'm happy to talk with you about if you guys have questions
about how we did that. I'll try and give you honest answers and not make things
up 'cuz they sound good, which is my
general response to answering questions is to lie to you through my
teeth constantly. So Burt you didn't know this when you invited me is that
I was gonna make up stories the entire time.
So let me tell you a little bit about what Eli is
and our path towards getting Eli out into the world which is exactly the
language that we've always unused and always had in our head
as we've thought about Eli and talk to you as best I can about
the decisions we made and the lessons that we've learned to this point in time.
So Eli, we invented Eli by mistake.
So what we were trying to do with Eli is, so let me back up.
I've been in conversations with a number of people
outside the university about commercializing ideas that have been
generated by the research center that I helped erect.
Almost from the first six months that we were in operation
there were people that I knew and people who came to know us who thought
that there was a tremendous amount of innovation possibility
in the domain of writing instruction and writing education.
And this is surprising to a lot of people, but actually
not too terribly surprising when you think about the amount of
writing intensive knowledge work that takes place, not just in school, but
also takes place in organizations. So
keeping an eye on commercialization possibilities is something that
we've had, we've done since almost since the very beginning of WIDE.
But in fact Eli was a mistake. So what we were doing when we invented Eli was
we were working at the College of Business on trying to develop
some innovative ways to teach organizational communication
and ways that didn't, and ways that scaled beyond the university. So
we were trying to do some stuff that was useful and that would be adopted by the
College of Business, certainly, but we were trying to do it in such a way that
it would be useful for audiences outside the university. So
what were we trying to do?
Well we're trying to reduce the amount of instructional time to the fewest
number of student-teacher-instructor contact hours as possible.
So we were trying to compress the number of hours,
we were trying to compress the curriculum,
and we were trying to leverage what we understood, what we understood via the research
to work in writing instruction. And so the work that we were doing was the most
conceptually, theoretically, and pedagogically tight course design and course delivery that
that we've ever done in our careers as academics.
And it turns out that it worked. And but it works; but it didn't work at first.
So what didn't work was we had to find a way to teach writing that
didn't rely on the teacher to provide all the review feedback.
Well it turns out that
there are some very standard, commonplace
pedagogical practices in writing instruction that we call peer review
but are really just a species of peer learning.
And I'm going to come back to peer learning in relation to Eli cuz I think that's
actually really the big idea.
So we were trying to figure out a way to leverage the power of pure learning
'cause we didn't have the time and space in that instructional
environment for teachers to review the work of students.
We had to have students do
all of the review feedback given the constraints that we placed on ourselves,
and the students had to get good at it.
So not only did we have to have a way to do high quality review work that
fed high-quality revision because
the learning theory we were working with here is that all learning and writing
happens in revision.
That's a little bit hyperbolic but not too terribly much, the research is pretty
clear about what gets, what helps people get better at writing, they have to revise,
they have to do high quality revisions,
they have to learn from those revisions, they have to be able to execute those
revisions again and again and again.
Well one of the things that is required for high quality revisions is high quality review feedback.
And so one of things that doesn't happen in many instructional contexts is the teaching of review.
So we realized that in order to facilitate the high-quality revisions that we needed to
facilitate, we had to be able to teach review.
And you can't teach what you can't see.
And so one of the problems that
teachers, writing teachers everywhere have is the ability to see review feedback. And
this is true in lots of instructional contexts, if you can't see
the ways in which students are responding to each other then as a teacher
you have no way of intervening and teaching those review and revision processes.
These problems of efficiency and problems of visibility
were the problems that we couldn't solve with existing technologies.
We tried, we tried everything we could leveraging the most common review and
revision pedagogies in the world is Microsoft Word plus email.
So we tried all of those existing technologies and we simply couldn't do it, we
were failing. And so we realized
pretty quickly that we needed to invent a technology that allowed us to do that.
And so that was the happy accident is we invented Eli
which is a, and Eli per se
is a service focused on writing instruction. It's commercial implementation right now
is in the market, is in the writing instruction market in higher
education and in K-12 education.
What Eli really is, is a generalized review service.
And the way in which we're framing it and reframing it over the next year is
to support peer to peer learning and education. This is the really hard
problem: teaching computer
programming or whatever to ten million people or ten gajillion people.
And the way in which people are trying to do that is the easiest thing
in the world, that's the easiest, that's the lowest hanging fruit and
not particularly interesting fruit in my view. In other words
if the problem of education that people are trying to solve with massive online
education is the transfer of information from point A to point B, let 'em have it.
That's low quality, that's a low value proposition.
I don't think that that scales in particularly interesting ways to
increasing access to education. The way
to teach complex things in complex ways that have high value and high value propositions,
different kinds of modalities and technologies are going to have to be brought to
beer and one of those is peer to peer learning.
People are going to have to help each other get better at doing complex things and
they're gonna have to do more complex things than filling in bubble sheets
in order to do massively online education and in a way which is
high value education. So
this is actually an area that we're deeply interested in for intellectual,
and now for commercial reasons. And we see Eli as
our first foray into that world.
And so Eli was an accident and so the question became;
so we invented it, we developed it in our own classrooms
at the College of Business
and then the question became what we do now?
And it was pretty clear to us that we had something
and every teacher who used Eli; and the Eli that people were using at this
point in time it was duct tape and spit and twine and it was MacGyver code, it was awful.
But we managed to make it work and with
a lot of patience and love and tender care and good wishes.
But even in that state the value proposition to teachers was pretty clear.
And what we could anticipate--and this wasn't built into Eli at the time--
are the analytic possibilities. So
this notion of visibility, what you can do when you can see things
and what you can do when you can count what you can see
and the kinds of analytics that are possible in relationship to that,
this is where also a lot of magic, I think, will happen around Eli in the future. So
Keri mentioned in passing in her presentation that there are a couple
of patents associated with Eli and there are,
but Eli per se is not patentable, but we have an algorithm
that's associated with Eli that is the source of one patent. And I honestly can't
remember what the second one is.
The patent process is intellectually fascinating but otherwise is mind numbing.
So then the question is we knew we had something, what are we gonna do with it?
And this is where I want to come back to
this notion of getting ideas out in the world to make a difference. So
you know as academics and certainly as partners in this research center,
what Bill and I were most interested in doing is precisely that,
getting ideas out in the world so that they can make a difference. And
software for us from the very beginning of WIDE
was really just a materialization of ideas. It's theory made material.
So what Eli is and what all the software is that we've made,
they're just simply ideas.
One of the ways that our ideas live in the world of academics is
they live in the world as books and articles and book chapters and the kinds
of things academics typically produce.
I'm bored with that, that's not, that's not that interesting to me anymore. And
certainly we could have written books and we've could have written a number of
articles to our colleagues, six people would have read them.
My mother would have read them, she would have written about them on Facebook,
you know life would have gone on.
But we were interested in something else with Eli and some of the other things
that we're doing. And we were interested in making a difference, honestly,
in the way in which writing is taught in the United States.
So I am unashamedly a romantic and an idealist in this regard and I have no
problem standing up in front of people saying that
what we're really trying to do with Eli is change the way people teach writing
and we knew that the only way we could change the way people teach writing--and
we don't think Eli is the only arrow in our quiver in that regard--
was to get Eli into the hands of as many people as quickly as possible at
the lowest cost possible and in a way which was sustainable.
And this is a really hard problem and Dean's going to talk a little bit more
hopefully about open source.
So one of the things that we instantly thought about was is this something we could do?
And doing the math we just didn't think that open source was a viable path
for us. There wasn't, there's not a built-in group of
developers out there interested in making writing software.
We have no capacity
to sustain a developer community anyway, nor did we have any capacity to sustain
a user community. So one of the worst things that could have happened
actually as a commercial enterprise right now, one of the worst
things that could happen to us but it be a problem we'd be happy to solve is to
have twenty thousand users next week. As a commercial venture we would
have a hard time supporting those twenty thousand
users next week. We would like to be able to solve that problem,
that's a nice problem to have.
But these issues of sustainability in scale are very hard problems.
And so in our running of the math the only way that we could think to make
to make this work,
that is to sustain the operation and to make a difference in the world,
was to commercialize this and that was a very complicated process that
that I won't go into in any detail but I'm happy to answer as many questions as I can.
But fundamentally what we've done was
in creating, in creating a company and spinning a company out of the
university to which we've licensed this technology and a few other technologies
with MSU as a partner what we've tried to do is maintain some control and some integrity
over our inventions to allow us to achieve the goals we wanted to achieve as
intellectuals and as human beings and now as business people.
But also to do that in a way which is sustainable and
in a way in which we can keep some eye on integrity and I'll hopefully
try to do this the right way and to do a little good at the same time.
So that's, I don't know whether we've learned any lessons other than
starting a business and trying to help run a business as a
founder with you know hands off as I am is
unquestionably the most difficult thing I've ever been involved in. This is
mind-numbingly difficult work. So maybe one lesson is don't,
but I actually don't want to communicate that.
But do this, get some help, use the university resources
to help you through this process.
Talk to as many people as possible and find some people who have
expertise in areas where you don't have expertise because
the work of making software that lives as a commercial product and
and attending to the commercialization process and the sustainability of that
enterprise is exceptionally difficult work.
There's lots of time, right? Nine minutes? Excellent.
Questions, thoughts about Eli, about the process?
Yeah, go ahead. How long
and how much time did you wind up spending talking back and forth to MSU's Intellectual Property?
Kari, I don't know, how long did the process take for us?
A year or so? It depends on when you consider it started. I mean we've been talking about Eli
probably for a couple years but the actual negotiations were probably six months, maybe.
yeah, they went really fast once we got serious about it.
And the time lag, it took us longer than I
wanted it to take but it was not, MSU was not the reason we were slow,
I was the reason we were slow. So trying to figure out,
trying to line our ducks up on
our side of the equation to make sure we did this the right way
was very hard to do 'cause this was new to me.
Does that answer your question?
That was the same question? How crazy is that? You guys are on the same wavelength.
How is the business going? Well, how is the business going?
I own, I own a large percentage of zero at this point in time.
It is, start ups are really hard.
It's rocky, it's exceptionally difficult but
despite that it's going pretty well. So
we've had, so last year we had
two thousand somewhere in the neighborhood of two thousand
paid users of Eli at Michigan State University
and we did that with absolutely no sales and marketing whatsoever.
Okemos Public Schools used Eli last year, no sales or marketing whatsoever.
We had a number of highly successful class, what are called class tests
or research based uses,
the most successful, all of them successful, the most successful one at West Point.
So West Point is going to be a user and that faculty member went to
Washington State so we have a seven hundred license
use at the University of Rhode Island happening next year. So we have
we'll have over the next couple of years on our own with no sales and
marketing; so getting a sales and marketing operation up and
going when you're not really a sales and marketing company, I mean we're not gonna have
two hundred sales people running around the country selling this thing. So how do you
have to sell this? This is crazy, right?
How do you sell this without sales and marketing? Finding partners in the
appropriate markets to help us with this has been where most of our effort has
existed. So I'm hopeful that within the next
two months we'll be able to announce a very, very significant partnership in higher education
which will instantly make Eli available to
every college and university in the country
with the sales and marketing enterprise to back it.
So we've been doing a lot of really hard and quiet work to this point in time.
But what's been beneficial is that where we've had an opportunity to show Eli,
people have bought it.
Do you have time to demonstrate of few features of it or something? I was kind of
hoping to see a little bit of how it works. I don't actually to keep the time but
how much time do I have? Five minutes.
So let me try and show you something.
And try to do this in an intelligent way.
So we're right in the middle of pushing--
aww and I'm going to have to remember my password now. I don't know if I know it.
That's what it should be.
Oh my gosh, now you've just witnessed something that's truly divine [laughter].
I don't know how I did that, this is magic. This is magic.
So what you're looking at a teacher view.
So let me go to an old course. So Eli is,
Eli is, let me explain something really fast about this.
Eli is a review service so
and it’s intended to be a simple service, it's not intended to be a
learning management system.
We don't want it to grow in scale in that regard, it's actually intended to bolt
into LMS's. So some of the very quiet conversations we've been having with
people is bolting Eli into
many of the LMS's that MSU is considering and adopting and/or adopting
over the next thing you know however many months.
What teachers do with Eli are really two very simple things. One
of them is really boring and one of them is more interesting.
The really boring thing is that teachers create assignments, big surprise.
And students respond to those assignments by
writing into Eli or copying pasting text into Eli or uploading text into Eli.
On our development arc by the end of the year we'll have the ability to deal with
media, multimedia a little bit more effectively than we deal with it now.
That's all the boring stuff.
The more interesting stuff is that, is when teachers create
reviews and structure reviews and students participate and conduct reviews.
It's in the creation and structuring of reviews that teachers enact
all of the important pedagogical practices, right? So how a teacher sets up
a review, the criteria, how those criteria are taught,
understood by students, how those criteria are used by students in the
conduct of a review,
that's where all the teaching and learning magic happens.
And by the way, in the conduct of a review by both teachers and students
this is what our analytics pay attention to on the back end.
And so the analytics that are the subject of the
patent, we compute a helpfulness score for a student. So we can track
if a student gets more or less helpful over time
by using seven, I think seven different variables and weighting those
variables differently. Seven different activities in the system.
So what you're seeing here--let me give you a real quick--
so what you're seeing here are archived assignments and archived
reviews of an old class of mine.
And let me, let me pull up an old review.
And so what you're seeing here, let me show you something,
let me just cut to the chase and show you something that's
sort of really important. What you're seeing here
is something that no writing teacher in the country is able to see.
And that is a dashboard view of not just all of the writing activity in her classroom, but all
of the review activity in her classroom.
No writing teacher in the world can see this right now. Now whether or not seeing
that activity is transformative or not, that's a different question. But right
now in order to see what I'm showing you actually in this display
I would have to somehow collect all of those papers.
It's the milk crate, what we call the mill crate problem. And whether
it's a paper milk crate
or digital milk crate, it's still a milk crate problem.
Somehow as a teacher in order to see what I'm showing you in this interface I
have to be able to collect twenty to twenty-five papers.
Oh and by the way, collect
the reviews on those twenty to twenty-five papers whether they're
written on separate sheets of paper or whether they're written in the margins of paper
and whether that's analog or digital doesn't matter,
nothing right now aggregates and collects this information and makes it available to
a teacher. So what I can see right now is Laurel's performance
in relationship to; one of the review types that we have
is essentially a criteria matching, it's a binary yes or
no, does this thing exist or not exist in the text?
I see what the class averages are on this and I see
what Laurel did on this.
This is actually a bad choice because Laurel wasn't there that day. But here's Rebecca.
What I can also see
is all of Rebecca's feedback from various students and I can endorse this
feedback or I cannot endorse this feedback as a teacher.
But I can see it, I can see it, I can teach it, I can help,
I can help Lauren get better at responding to Rebecca
and I can interact with this review feedback and say to Rebecca, "You need to
pay attention to everything that Lauren has said to you because this is just
rocket science good reviewing activity".
So what I'm showing you here real fast and then I'll stop now,
are the two primary value propositions that teachers see immediately.
One is I can do in fifty minutes what it used to take me two to three class periods
using Microsoft word and email or paper. And two I can see
my class activity in one computer interface,
I can see what I've never been able to see before as a teacher.
Now I can teach it. I can see it, I can value it, I can teach it, I can grade it, I can do all
sorts of things with it as a writing teacher that I couldn't do before because I
simply couldn't see it.
So those are the two immediate value propositions. Teachers, the more
experienced the teacher, the more easily they associate with it.
And then there's downstream there's more analytic possibilities
and certainly more programmatic data
that schools are much more interested in
that's associated with use of Eli.
So quick, inadequate I realize, I apologize I'm probably a little bit over time
but this is really what we're trying to get at
with respect to Eli. And this is by the way straight down the middle of the road,
research based writing pedagogy. There's no, we're really
not on the bleeding edge of what works in writing instruction. And I
actually think that that's important in terms of the value proposition of what Eli's about.
Alright, thanks.