Wednesday, February 4, 2015

Data ownership indeed

Many thanks to MetaWhat! Data! for bringing this article on "Questions of Data Ownership on Campus" to my attention.  This is a very interesting read, and I appreciated the personal perspective that MetaWhat! Data! brought to her write-up.  I think the concerns about privacy and security are well-founded, and universities must address these concerns as well as be up-front about disclosing what data that they collect and how they use it.  However, the article also makes a valid point that universities increasingly need this data to provide competitive services and justify their existence to ever more tightfisted funding agencies.

To underline the intense interest universities now have in collecting and using data, I will list several ideas that I have heard proposed in the last year (caveat:  these are pie-in-the-sky, none are remotely close to implementation):

  • Software that can associate individual student success (i.e. grades, attendance) with instructor teaching or course style (e.g. flipped, lecture, seminar, experiential, etc.) and recommend specific instructors or courses in which the individual student might be more successful.  (For example, at registration, the student selects Physics 201, and the course system says "because you did well with Dr. Jones, Dr. Smith may be a professor you would like," or possibly "other students who did well with Dr. Jones also did well with Dr. Smith."  Actually choosing Dr. Smith's class is left to the student.)
  • Software that counts occupancy of rooms over the course of the school year based on mobile device connections to the building wi-fi.  The occupancy data (anonymized) informs decisions on where to upgrade the facility (e.g. "study rooms always packed, this justifies the cost of adding more" or "seating areas near west-facing windows fill last in summer--improve climate control or install shades").
  • Tracking whether students are more successful when they receive different kinds of library instruction (in-class with peer instructor, with librarian instructor, outside of class with or without extra credit, etc.).
  • Tracking at-risk students to see which additional support services they use most, and which seem to increase success the most.
  • Tracking research database usage from initial search to article download to determine optimal paths to offer users and justify continued subscription expenditures.
Are these great ideas, or really creepy?  Or both, like Google Now?  Is it Big Brother come to campus, or a more benign Big Sister?  Tell me what you think in the comments.

1 comment:

  1. They seem like well-formed ideas to me ... but it's all about preventing the dreaded "data breach"! So the question is: Can this data ever truly be anonymized? One thing to consider in "big data" operations such as this is how long to keep the data ... 10 years seems about right.