The University of Maryland's College Park campus has
proposed a new
strategic plan for information technology in order to spur innovation, and moreover
invites comment from stakeholders. Unfortunately the opportunity for stakeholders to consider
others' thoughts on the matter is limited because all discussion is held behind a privacy wall.
You need to specifically log in using campus credentials just to see what others might have to
say, you need to know where to drill down once you get there (comments are spread out at
the site - you need perserverance to track down all of them) and OIT's prototype web site which
hosts this "discussion" limits users to those running old browsers on specific platforms to which
the stakeholder is willing to add a plugin of unknown origin and function. Those are probably
sufficient reasons to explain the general lack of stakeholder exchanges on a topic which has a
potentially sweeping impact to the campus community. What follows is one stakeholder's commentary,
copied here so as to not be hidden on the other side of OIT's privacy wall. -- Jim Purtilo
The draft University
IT Plan should be sent back for a serious do-over in order that it reflect both
the IT industry’s best practices and the University’s core missions.
Taken as a whole, this draft paints a picture with OIT at
the center of the universe. In Recommendation 3, for example, we learn that it
will be a mission of the University to build information technology for the
scholarly achievement of its students by means of IT. Leaving aside a question
of how scholars other than students may fare in this brave new world, and with knowledge
that this is cast as a university plan,
not narrower OIT plan, we are
surprised at the shift. The plan might more reasonably have said that it is a
University of Maryland mission to enable scholarly achievement, and moreover
that one of the means by which this may occur – one means among many – is via IT.
The inversion of missions is not a stray editing slip, as the mindset is
pervasive throughout this proposal, and indeed the very first action item under
point 3 places an obligation upon faculty, not OIT. Having declared that technology
is now the campus mission, we are told its interests “require that UMD faculty
fully engage in the use of IT tools …”
Our faculty has just adopted a Gen Ed plan which
willfully avoids use of the word “computer” (in stark contrast to trends among
our aspirational peer institutions) so perhaps it is a good thing that at least
our faculty might know something about IT. After all, then our students might
be exposed to it by indirect means. Nevertheless, that is an issue which can be
addressed more reasonably by patient and focused advocacy of those who control
Gen Ed, not an inversion of the campus mission to serve the imperialist
interests of OIT.
The two most practical implications of an OIT-centric
world view are a likely resource grab (especially to gain control of computing
activities not yet under OIT’s purview, something my friends over in the Smith
School might characterize as “empire building”) and acceleration of the
implementation of MOOC and similar on-line content delivery, a top-down
decision already made by the bean counters. My remaining remarks are intended
to give color to the likely and unwanted effects of adopting the plan as given.
1. OIT should offer
a service but otherwise let units decide for themselves.
A clear implication of this draft proposal is that
activities and resources presently within other units will eventually become
part of OIT. This should not be required.
My department’s fastest rise in international stature many
years ago followed our success at winning fundamental independence from the
administrative unit which today is called OIT. We had to be free to operate,
configure, procure and manage our own resources, which we did at lower cost and
with higher quality. Not all units on campus can or should be so independent,
but it would be a fundamental limit on our ability to compete at highest levels
on our missions to work through a
single procrustean OIT model of computing resources.
To illustrate, today I operate a laboratory which resides
within CS department facilities but whose resources are independent of it. They
were creatively acquired and enable me to perform not only exploratory research
but offer advanced instruction which cannot be done in present OIT-provided
educational resources. The essence of what we need is control, which is at odds
with the proposed OIT model, and responsiveness, which OIT has no track record
of providing. Even if we overcame those barriers, no DRIF provides enough resources
to let us procure hosting from OIT for prospective work on frontier areas. I
would like to think we could agree that a scholar exploring prospective research
ideas in computing would have different needs than an educator interested in
ordinary services, yet this is not recognized in the OIT proposal. If adopted,
the full weight of what are today ‘off book’ computing activities in support of
both research and education would fall onto the departments where, given the budget
realities, they are likely to be dropped. CS would not be able to do as much as
we do today.
A far superior model for computing on campus would foster
growth of OIT as an internal service, which would compete with, but not
necessarily assimilate in Borg-like fashion, other operations. Then let each
member of the administrative community decide how it prefers to meet its own
needs. A little healthy competition will help lower costs and improve
responsiveness to mission needs.
instruction should deploy organically, as its effects and value can be
understood, not driven by non-academics to serve their needs.
The OIT-centric mindset in this draft similarly casts
technology-intensive education as an inevitability. It should be softened to
recognize that perhaps the professoriate, not administrators (many of them
non-academics) should make the calls on when and how to involve OIT. The jury
is simply out on what is the effect on quality due to use of trendy tech-based
content delivery. Undoubtedly technology will help us discover effective ways
to prepare our students for the future, but it is too early to adopt a policy
which locks us in, and potentially will detract from our mission of excellence.
Yes, the visibility of MOOCs and other trends make them exciting, and we
understand why leaders want to make one-time investments into technology which purports
to shave on-going expenses of delivering content via expensive faculty. But
please – let’s exercise some patience and good judgment. Change the draft to
sound less like a case for major equipment acquisition today and more like we
want to be ready to support experiments in content delivery tomorrow.
3. Campus should
show it can follow its existing strategic plans before making new strategic
The 2008 strategic plan – described as a ten year document
and thus still in force – establishes:
“The administration will establish review
processes to evaluate the performance of individual units in all
divisions in meeting standards for improvement in their quality and effectiveness.
We will use state-of-the-art information technology to ensure that
institutional data are accurate and available for effective decision making. We will review all academic and administrative
approval processes and change them as necessary to ensure that decisions can be
made expeditiously and responsibly. Institutional
units will provide responsive, customer-oriented services to all
constituencies. Annual surveys will monitor the progress made in improving
academic and administrative operations.”
Obviously UM’s strategic plans are observed selectively,
with administrators free to cherry pick dicta as suits a need of the day, but
had the existing language concerning reviews or state-of-the-art IT been
treated as fully in force, then not only might we have not needed yet another
plan, but we’d have taken note of OIT’s track record on some fronts:
- The centralized campus effort to refactor
academic data systems under Kuali is years behind, has cost an immense amount
of money and has yet to positively impact its first stakeholder outside of a
few very specialized internal users. During the same period of time, whole
generations of organically grown, non-OIT based – and genuinely free – systems were
built, deployed, used successfully with stakeholder acclaim and yet ultimately killed
for want of campus adoption of software which OIT would not accept.
Administrative unwillingness to lead, follow or get out of the way is the mark
of 1970’s era Soviet centralized planning, not a tenet of a successful
university in the new millennium.
- Years of professorial advocacy that we should do
real data mining in order to improve student outcomes have been in vain.
Scholars today know serious techniques for learning from our data. Corporations,
government agencies and scientists apply these techniques to expose a wealth of
insights about what works and what doesn’t in their respective operations. Not so
at the University of Maryland, where our antique data services continue to age,
guarded by an organization that seems to spend more energy protecting
stakeholders from accessing the data than it does upgrading services so
stakeholders could make better-informed decisions. As noted, data-driven
decision making is part of the strategic plan, yet those few who are allowed to
try it find Byzantine data models, having ad hoc updates from inconsistent and
unreliable sources. This is not a state of affairs which will be fixed by granting
OIT with greater centralized authority or by creating even more plans which
will be selectively followed.
- Over the last decade, by my direct observation,
OIT has done an outstanding job of stonewalling campus map and geolocation
services. If not invented or controlled by them, then organically-grown software
or data is not accepted. OIT has held out instead for tremendously expensive
commercial packages which only they would manage. In the meantime, while OIT
has held out for funding, nearly half a dozen administrative units on campus,
all with serious mapping needs, have over-paid for conflicting resources that needlessly
replicated one another. Taxpayers and stakeholders alike are the losers in this
classic empire building game, which would be rewarded under the proposed
- For an example of what life would be like under
a fully centralized campus computing system, one only needs look at the very
TerpSharePilot site they make available for collecting comments about the IT
draft. The organization and structure of this site was so confusing that UMIACS
ultimately had to post a detailed guide for its computer science research
faculty to help them figure out how to navigate the site and add comments.
These examples suggest that caution should be exercised
before OIT is placed even more squarely on the critical path of campus
stakeholders – and before making up more plans.
A centralized service model like seems to be proposed in
this plan flies in the face of corporate trends toward getting control into the
hands of stakeholders to affect their own lives. Companies which are successful
in today's economy respect a basic reality of economics, which is the
debilitating impact on quality when you concentrate authorities and
responsibilities in a monopoly. Institutionally, lean and agile methods which freed
my unit from dependency on OIT have worked. We should not change this.
Any unit on this campus should fear a bureaucratic model
which places its success in the hands of staff whose incentives and reward structures
are independent of that unit. This is true for departments that have only ordinary
tech needs, just as it is for departments that educate and research out at the
cutting edge. Each department should be free to incentivize performance from
OIT by also being free to satisfy their own needs as they see fit.