A two-day symposium on Action Analytics was held in St. Paul on September 22-23, 2009. I took lots of notes via Twitter. I’ll copy and paste several of those tweets from day one and call it a blog post. Twitter hashtag #ActionAnalytics
- Symposium begins in 1 hour. First up: “Why action analytics for higher education?” My answer: to improve student success
- Symposium getting started. MnSCU Chancellor states that the people are fed up with the completion problems in higher ed.
- “I don’t think we can assume that our needs will be funded if we can’t demonstrate high levels of achievement.”
- The key piece of #ActionAnalytics is not the analytics, it is the ACTION. Use the data to lead us to actions to span the achievement gap.
- Who’s been invited to #ActionAnalytics ? Trustees, presidents, political leaders, policy makers, national associations (& me?)
- Donald Norris has a blog titled: Linking Analytics to Lifting Out of Recession https://bit.ly/1TI4s
- Higher ed has never said “Every student needs to succeed.” We accept failure – non-completers are expected & “normal”
- Health care & education are similar in the “laying on of hands.” Nationwide discussion about reforming health care, will education be next?
- Had breakfast this morning with Capella’s president. He told me some surprising things about their student demographics. Very impressed. Capella educates huge numbers of first-generation college students. I don’t think that’s widely know. He tells a great story about access.
- Higher ed measures lagging indicators (persistence, graduation, etc), should focus on leading indicators (Wk1 engagement)
- More open education environment will take us from education opportunity (now) to education assurance (future). Bill Graves
- Getting data out of our systems is hard (costly) – we need better standardization of info systems. Michael Feldstein
- The data is leaving the LMS for Web 2.0 apps – how do we build connections to all these tools (can we)? Michael Feldstein
- “Smart Change” is the aggressive application of change management principles to develop institutional capacity (Baer/Duin)
- Routine change happens everyday, transformative change gets you out of your comfortable box. Academics freak out with that.
- My take on transformative change – largest barrier is TRADITION (think Topol singing in Fiddler on the Roof). For example: 300 bright minds may agree that we should extend the school year – but 300 million people disagree because of tradition of summer vacation.
- John Campbell of Purdue. Building capacity for analytics is really about building community – getting buy-in
- Purdue Signals Program – actionable intelligence – real-time predictions of student success in a course. Try to focus your data analysis – one question leads to another – must make choices and keep it simple. Purdue’s Campbell. Purdue’s info system hasn’t reduced course drops, but they drop earlier. More B/C grades, fewer D/F. Purdue students didn’t think it was Big Brother. They appreciated the info avail to them “thanks for kick in the butt.”
- Alex Ushveridze – Predictive Modeling Expert at Capella. “How” questions depend on the “who”and the “why.” Persistence? Predictive Modeling is a continue cyclic process- a way of acting rationally. Online education is ideal arena for P.M. Capella – Early alert – how early? FIRST WEEK determines everything!! High predictability of grades at end of first week.
- Jeff Gran- mgr of assessment at Capella will focus on measurement of learning. How can we collaborate to measure outcomes? Capella has 1,100+ faculty. Outcomes are measured with a fully embedded assessment model (FEAM) in each course.
- Craig Scheonecker of MnSCU showing MnSCU accountability dashboards. Focus on fewer items and make easy to understand. Check out the MnSCU dashboards – available to the public https://bit.ly/Jq0lb You can even “drill down” (somewhat).
- Conundrum: as we move to individualized learning models – how can we aggregate/analyze data with small N sizes?
- Issue: there is a need for info literacy about these analytics – who’s doing something about that? Michael Feldstein
- Will #ActionAnalytics evolve into a set of open tools that can be shared across institutions? It’s not happening yet.
- ERP vendors, LMS vendors are all looking at (selling) these types of tools – but is that the direction Higher Ed should go.
- https://twitpic.com/iq2vj – Lunch keynote speaker at #ActionAnalytics – Undersecretary of Education Martha Kanter.
- During lunch: “When Blackboard is presented a business opportunity – their response is ‘What would the 19th century robber-barons do?’”
- #ActionAnalytics panel discussion – What info, reports, dashboards are needed? 1st up: Dr. George Boggs, American Assoc. Community Colleges. Dr. Boggs: The old saying was “students have a right to fail.” Luckily, that point of view is changing – but not totally. Spellings Commission concerned with consumer info. That’s not the focus for CCs. We need the data to improve our outcomes. Voluntary system of accountability. One concern is improving the effectiveness of our remedial education programs.
- Foreign dignitaries come to the US. to study our system of higher education. “Sorry, we don’t have one of those.”
- 25-30% of the developmental ed is done at universities – but they don’t do remedial education. Name Game: Banana Fanna Fo.
- Revenue per FTE at public institutions: $7,059 appropriations, $4,004 tuition revenues. (64% – 36% nationally)
- If academe makes accountability studies without outside oversight, will they be viewed as being less than truthful?
- “As an educator masquerading as a technologist, I now realize how little I know about what will happen next.” Michael F.
- https://twitpic.com/iqgyd – From #ActionAnalytics – two bright minds: Mark Milliron and John O’Brien.
- Data Quality Campaign website: https://bit.ly/e2bMb Using Longitudinal Data Systems to Improve Student Success.
- We have a greater supply of data than demand for it. People don’t ask for it, don’t trust it, can’t use it effectively.
- Next generation of analytics – interesting conversation since we are still making baby steps on the current generation. Next gen: move from high-cost business intelligence to value analytics for the masses. Next gen: move from users waiting for results to immediate results with dynamic analysis and changing parameters. Next gen: analytics currently driven by power users, move to end users deploying user-friendly tools. Next gen: move from institutional data sets to cross-institutional analytics and P-20 data sets.
- Discussing some of the user-generated analytics – RateYourProfessors, PickaProf, CollegeResults dot org, etc.
- Statistics is THE math that we need to be teaching in high school and college. More important to our future than Calculus.
- One of the secret sauces for analytics is the predictive model – good place for collaboration. Open-source style. Al Essa
- Our technical capacity is developing faster than our ability to make constructive use of the output. What does this mean?
Overall, it was a very good day, very long day, spent with very intelligent people. My brain is full.