Gayblack Canadian Man

Foreign Policy Analysis
Creating an Analytics Culture at OSU

Creating an Analytics Culture at OSU


One of the challenges
we had we had just completed a system conversion
of our student information, our human resources information
and our financial information systems. We did not have a good
way to the reporting at that point or to provide
information to people throughout our campus. And so SAS really was the tool
that we chose, in addition to the great visuals,
just to simply be able to pull data
from our system, combine it, create our files,
create all the background snapshots that we needed
to be able to do reporting and to provide visuals
for our campus. You can do things
that’s public-facing. You can do things
that are internal. You can do things that
are security-based. So there’s all different levels
of reporting that you can make available through the SAS
Visual Analytics tool. And then again the ability
to connect to various systems across campus. I mentioned that a lot but
that’s a big deal for us because it has given us so much
more data than we have ever had before. And some data that we had
never had access before. And so I think the complete
suite of tools and again within the Visual Analytics there’s now
Visual Statistics which allows you to do the analytics part
and the predictive modeling part which again will be kind
our next step down the road. In addition to the fabulous
visuals that we can now create again, Enterprise Guide
allows us to pull data, to organize data, there’s
data management tools and then the ability to connect with
various other systems across campus. Higher Ed is notorious
for different offices having different side systems. And the ability for us to
connect to all of those systems and bring their data in to
combine with ours has given us more information more data than
we have ever had access to. And it’s allowing us to do
different types of reporting and visuals for our end users. We’re able to get things
out to people quicker. We have a student
profiler a factbook. Every higher Ed
institution has that type of document or publication. It used to be that we would
get our data clean after drop and add ended. We would have programs we’d run
then we’d take the data from the programs and dump it into
spreadsheets and then have to make PDFs and then finally
about a month and a half later we had “a student profile”. Now drop and add-ins,
we finalize our dataset, we drop it into Visual
Analytics and we have our student profile. A week, you know, a week
after drop and add-in’s. So three weeks into the
semester and it’s done. And people have that
information available. It’s a hard thing to track
how every room is utilized throughout campus and we
used to provide some reports. They were very tedious. It took somebody numerous
months because they were hand entering information
about the time slots that rooms were scheduled. And so we actually collaborated. We have a wonderful Master’s
in Business Analytics Program. We have since we’ve implemented
SAS been hiring their students as graduate research assistants. And the first young lady
that we hired came in and she put together a
classroom utilization dashboard. So we can now look across every
building on campus and look and see what the
percentage of times rooms are utilized from 8:30 to 4:30
which are peak class times. And then within that
the percentage of seats that are used within
those classes. So we know that
only if a room is being used a large
percentage of the time but if the number of students in
the class is actually optimal. And so that’s been a great
way to look at our classrooms across the campus and know
whether we need more space, are there certain
size rooms we need? Are there rooms that
we’re not utilizing? And so are there perhaps
buildings where we have rooms that aren’t well utilized but we
could make them one larger room that would be utilized? So looking at renovating
instead of building a whole new building. So it’s been a really good
problem-solving tool as we have been looking at the classroom
utilization across campus. Of course, the
classroom utilization that was not something
we could have done other than using SAS Viya. It gives you the
ability to drill down. It gives you the ability for us
look at different time periods during the day. It gives you the ability to
look at different room sizes. And so it gives us
that specificity to be able to drill
down and get to very granular levels of the data. The visuals are just
fantastic because there are things you can show
people in a visual that resonates so much more
than me saying something. So, for instance,
we typically say, of the students that don’t
graduate we lose half of them in their first year. Well, we have a path analysis
of retention, enrollment and retention and
graduation rates that now shows we lose 800
students their first year, and in years 2, 3 4, 5 6, 600. So, seeing those numbers is a
much more striking comparison to people than me saying
here’s what happens. The other thing we uncovered
from that in looking at it is, of the students that come
back for that second year they have an 80% graduation rate. So, the emphasis we put
on first-year retention is well placed because
getting those students back their second year increases the
probability of them graduating significantly. And we have retention dashboards
that you can really slice and dice in so
many ways: gender, ethnicity, first-generation,
pell grant vs. non-pell grant, by ACT score, by college. And so, as to different people
across campus look at this, they identify different
things that they go, hey we saw this and we think…
Here’s something we wanna look at in terms of predictive
modeling for what might identify students who
might be a risk coming in. So that’s been a great thing. And one of the things that
we looked at that we were- I was having a meeting with our
faculty console and they said in advance they said, you know
we’re interested in looking at how students who take classes
online their first semester do. And with very little
effort we were able to add, we had course
information, we were able to add that has an option
in our retention dashboard. So that I could go
to a meeting and say, here’s how they’re retention
changes and here’s how their GPAs differ based on
whether they’re taking online classes their first
semester or not. And so it makes it very easy to
go from one thing to the next question that follows up and
be able to adjust to answer numerous questions that we
haven’t really ever been able to answer before. First of all, when we initially
rolled out our dashboards I actually met with
our President, Provost and our Senior Vice President
for administration finance to give them their own little
tutorial so they immediately knew what we had available. I always for the Vice President,
the Senior Vice President I report to, I constantly
keep him in the loop, let him know about feedback
we’re receiving from people. Let him know how we’re
utilizing the dashboards. And it was a great thing
that just happened recently. We had at Oklahoma
State University, we had a Retention
Conference which was an internal conference. It’s the first one
we’ve ever had. It had about 250 to 300 people
across our campus that signed up to attend the conference. And I actually got
to give the keynote and talk about our
retention dashboards because we had just
rolled them out. And so it was a
great way to share what we were doing with not just
the executive administration but other Deans
and Vice Presidents and department
heads across campus. Our campus was really hungry
for this type of reporting tool and this type of report
development and visuals and interactivity and dynamic
reporting that we’ve been able to provide. I think that we are
working with our campus. We’ve provided training for
people that want to come and learn to use the dashboards. Because what we really want
is to make sure people are comfortable and are utilizing
and understand how to utilize the dashboards, understand
how to obtain benefit from the information we’re providing. Executives are, they have
been using it as well. And so some of the things
we have tried to do for them is to make
higher-level dashboards that they have
access to that they know where to look for
these because it gives them the big-picture data. We are really fortunate, our
president, our senior vice presidents they are
enormously in tune with data and wanting information and
wanting to make decisions that are informed by data. We’ve been really fortunate
having the support from our executive administration. I think the culture
of analytics, in higher Ed everything again
is happening so much faster. We’ve got to be able to react
more quickly when we see changes. We can’t take months to do some
sort of study to come up with what should be happening. We’ve got to be able to have
a dataset that we can start looking at things and say
hey here’s what we’ve seen. Much more quickly
and respond to it and be able to proactively
respond instead of reacting several months down the road. We have someone
in our office who has built probably
85% of our visuals and he was not someone who
had programming expertise. We implemented the SAS tools
and he just jumped on it and learned it and
he’s fantastic at it. Because he- even without
being a programmer- he’s got this level of
creativity that really serves us when thinking about the
best visual for the best report and what works the best,
what’s the most benefit. We were able to have our
student profile probably 80% of our profile
out on our website within a week of going live. That was a lot of time spent
prior to going live so that we could have those things ready. But I consider that
to be a great success. I consider just the sheer volume
of the dashboards that we’ve been able to develop this
year to be a success for us. And that was really
our first-year goal was to make data,
make information more available to our end-users
and to give them information was much more flexible
and dynamic in how that it can be used. And so that’s that stage
one of our success. Another stage of that
again is helping the campus become accustomed to that. And I think we have
been successful in that. I think we certainly have in
addition to our executive team we have the advocates
across campus that are very
data-informed individuals. And so they advocate
for this as well. And then I think our
next stage of success again is looking at OK we
have all of this information. We have the data. What’s the information? There’s that kind of cliché
of we’re drowning in data and starving for information. We’re now at the point where we
want to start looking at what all- what is all this telling
us and how can we benefit? What benefit can this
provide to our campus once we start looking at
some of these relationships? What I think is happening
for us is it’s changing the conversations. It’s no longer a, can you
give us these numbers? It’s we have these numbers and
here’s what we see and here’s what we’d like to look at. How does this number affecting
this other number over here? Or how are these combinations
working together to help or not help the student? And so I think it’s changing
the way we have conversations and we’re looking more towards
analytics and how we do modeling now based on the
numbers we see instead of just looking at all the different
types of numbers and categories of numbers. The information we have
been able to provide again has been game-changers
across our campus. We hear time and time
again, we never knew this. That data- we now
see this in the data or we always thought this
and now we know the data tells us something different. And so being able to
give all that our campus and our end-users
and know that we have these partnerships across
campus where our goal is now to take this and move our campus
forward and see how we can make things better is just a great
thing and a wonderful time to be working at OSU for me. Academic ledgers are something
that we’ve had for years and years and years. And they were a
stagnant PDF that had five years of data that
had student data, faculty information, research
expenditures, and financial information. And we had them from the
department up to the college up to the university level. It’s how our colleges when they
come in and have budget talks with our Senior Vice Presidents
that they come in and they have this information with them. We’ve now completely rebuilt
those as a dashboard. And so now a Dean of a
college or even a department head or even the President
can get into any of these. They’re on our public dashboard. So actually anybody
can get into them. And you can drill
down and you can see how has the student
enrollment changed over five years? How has the faculty
counts for tenure track vs. non-tenure track
changed over five years? What about the
research expenditures? You know how much
external funding are colleges, different colleges
and departments bringing in? And so people have been thrilled
to see those now developed as a dashboard that’s
much more interactive. One of the things that we talked
about was the fact that we’ve been able to collect information
and connect with so many different sources across campus
to get additional information that we have never had before. One of those is our Student
Affairs have a database where they’ve tracked card swipes. So as students go to different
activities across campus they can track what
students are involved in. So being able to identify
some of those students that might be at risk. So again with the example
using swipe card information from students and
looking at the engagement if you see a student who their
activity has- their patterns have shifted. That can be a red flag
if they have never- if they’ve come in and they
haven’t really been involved in all. That could be a red flag because
it’s you know the research will show that students feel-
either feel connected or don’t in their first 6 weeks
of being on a campus. And so if you see a student
that comes in and they have no activities at all that should
be an immediate red flag that someone needs to reach out to
them whether it’s Residential Life or the First Year Success
office or their advisor and see if we can get them involved. So I think there is
some of that, that helps with student success. But I think the more we get
into the analytic side of it and looking at now that what
the retention data is telling us and how we can better
identify a student who may come in and need some
of the additional resources that we offer will be
a great thing for us. We really want to expand this
to include our Human Resource reporting, our faculty,
and staff reporting, as well as our
financial reporting. We have started
working with our- what we call our HRIM, our
Human Resources Information Management, and our financial
information management groups to start working
towards that goal. And then in addition to that
Oklahoma State University we are under what we call
the ANM Board of Regents. So there are other institutions;
Langston, Panhandle, NEO, and Connors who are umbrellaed
under that board but don’t have the same types of
resources that OSU has. What we want to do is have-
thru a shared service model be able to provide
those campuses some of the standard
reporting templates and then standard
visual templates so that their administration
can also benefit from this tool.

Leave a Reply

Your email address will not be published. Required fields are marked *