Learning Technology: Balancing Privacy and Functionality
University of California-Berkeley
There are a number of interesting privacy vs. functionality issues arising
as we consider ways to improve the ability of Learning Management Systems (LMS)
to provide information to students and faculty.
To put this issue in some perspective, consider the approach the library has
taken with respect to privacy. Historically, librarians have gone to great lengths
to protect the privacy of patrons as they access library resources. In particular,
many librarians have expressed deep concerns about provisions of the Patriot
Act as it may impact the privacy of their users.
But, with these concerns of the library in mind, now consider the case of an
instructor who wants to be able to track student use of assigned course materials.
For example, an instructor might be interested in determining if students are
accessing all the material posted for an assignment, and whether they are doing
so in a timely fashion. There are logical needs for such tracking since faculty
want to know if students are making progress toward completing their assigned
course work.
Interestingly, there is no parallel to this type of automatic tracking in the
library world. If students are given the assignment of finding materials in
the library, instructors are not able to directly track how their students are
using the library's resources. And, if an instructor attempted to enlist the
library in the monitoring task, I think it would quickly raise all sorts of
red flags.
Searching for learning resources is another area where issues of privacy versus
functionality might arise. For example, we know that when people are presented
with a common search interface such as Google they tend to do really simple
searches, which don't utilize the metadata associated with the resources that
are out there.
This fact has been noted by those concerned by the amount of work going into
marking up learning objects and other resources with metadata. Of course, if
users would utilize existing metadata appropriately, then searches would be
more targeted and effective. But, as things stand now, most students and faculty
simply don't use the metadata that professional indexers think they should be
using. However, by tracking LMS user behavior and by gathering preference data
we may be able to make searching more effective.
Consider the case of a student who provides preference information, such as
a particular learning style, to a LMS so it can provide adaptive learning experiences.
Given this information, the LMS may also be able to improve search results (i.e.
by automatically incorporating the preference information into the search).
In addition to storing information on learning style differences, next generation
LMS will also be able to store learner accessibility data. For example, if you
are visually impaired, the system would know your preference for large type.
It can be argued that the blending of accessibility data and learning style
information could improve the services schools provide. In fact, schools may
be obligated to maintain accessibility information to better serve disabled
students.
I think you can see, however, that there are privacy implications to storing
and using this information. Where d'es one draw the line on the usage and access
to this type of preference and accessibility information? Should this information
only be available to the student, or should instructors have this information
as well? And, to what extent should instructors be able to update this information?
I am sure you can construct scenarios where benefits and abuses might arise.
Clearly, there are trade-offs.
Campus student information systems also have a wealth of information that some
instructors might want, but there are also good reasons for not giving them
access. For example, some instructors may want to access student SAT scores
to tailor their presentation of course content and improve their interactions
with students. But, others might argue that instructors should not be able to
access student SAT scores since it might bias their interactions with students
who have relatively low scores.
There is another set of issues involving student evaluations of faculty. Generally
speaking, instructors are only given summarized student evaluation information
once the semester has ended. And, of course, students are supposed to be able
to provide their feedback anonymously. However, some faculty would like to have
student feedback during the semester (e.g. on the effectiveness of course-related
activities). So, there is the whole question of being able to do anonymous and/or
non-anonymous surveys during the semester to provide feedback to instructors.
In summary, as we develop technology to improve the effectiveness of instruction,
we need to consider a variety of privacy versus functionality issues. In some
cases, students may be unwilling to provide instructors with their personal
data for fear that their instructor's judgments may be biased as a result (e.g.
the SAT score case). But, some students may be willing to make this information
available to various types of learning software, assuming that the software
would not be biased. Also, how should campus policy makers balance the need
for student privacy with the needs of instructors who want to improve their
performance by knowing more about their students? Are there lessons we might
learn from the library community to help us address these issues? And, what
might the library world learn from learning technologists?