We were struck today by an article in technopedia.com on business intelligence and its relationship to the challenges of data and higher ed. Near the article’s conclusion, appears the following observation:
“Studies by Merrill Lynch indicate that 85 percent of all business information is made up of unstructured or semi-structured data, including emails, news, reports, Web pages, presentations, phone conversation notes, image files, video files and marketing information. In the IT industry, management of such data is considered a major unsolved problem.”
We would argue that this “unstructured or semi-structured data” is also a “major unsolved problem” in higher ed data analytics as well. It is not just the challenge of placing all of this data in a location and form for schools to analyze for the benefit of the student in terms of support and program structure Privacy and access issues arise when this data is to be shared with success stakeholders across different departments within the institution.
First, let’s set the technological stage for schools wanting to use this data. How can emails exchanged between admissions officer(s) and the student be conveniently collected and placed within a particular student’s data warehouse? What about exchanges on SKYPE or via SMS (text)? When a student mentions to an admissions counselor that their father is ill and requires the daily attention of the prospective student in terms of both time and financial support, how is this information captured from a text, email or SKYPE chat? The short answer at this technological juncture is yes if that school admissions officer notes it in a data base. It is then incumbent upon the school to train admission officers to capture such information. It is also necessary that this data not be on an Excel spreadsheet or WORD doc safely ensconced on the officer’s laptop. Assuming that this critical data is captured, and is placed in an accessible data location to which the school has access, who should be allowed to “see” this information. We place “see” in quotes as there are different levels of sight when it comes to data.
To be sure, various departments within a school will alternately argue that access to this data should be controlled, shared, or private. A Solomon-like approach to giving “sight” to this data might be to assign risk factor(s) to the data. Such risk factor(s) could, sua sponte, or in concert with other factors, place a seemingly on path student into an at risk status. In this approach, other success stakeholders would see a heightened risk for the student (and operate accordingly) but not know the exact reason(s) for the heightened risk. As an aside, we are reminded of a recent presentation at the Higher Learning Commission Conference in which a university staff member, commenting on predictive analytics, said: “It’s a challenge to intervene with a student who has a B+ GPA.”
This example of communication between an admissions officer and student can just as easily be imagined as between the student and an orientation staff member, adviser, professor, financial aid officer, tutor, supplemental instructor and other staff member. The medium in which all of these success stakeholders communicate with the student can be just as varied (e.g. SMS, SKYPE, FACEBOOK, email, telephone).
Discussions about “unstructured and semi-structured” data must take place at all institutional levels where decisions about methods of recordation, training, and the sharing of data must not only be made, but just as critically, be instituted and adopted by all success stakeholders. We welcome engagement in this dialogue with your school and conferences.