Successful transition from secondary to higher education using - - PowerPoint PPT Presentation

Successful transition from secondary to higher education using learning analytics

Erasmus+ projects ABLE and STELA

• verview

1. ABLE project 2. STELA project 3. findings & recommendations

1/3ABLE project

Erasmus+ (Strategic Partnership 2015-1-UK01-KA203-013767) ableproject.eu

ABLE: goals

 strategic partnership, launched in September 2015  How to apply learning analytics to support the transition from secondary to higher education?

• 1. Identify strategies for integrating institutional support around data

generated by Learning Analytics.

• 2. Strengthen research into first year experience and transfer domain

knowledge to Learning Analytics interventions.

• 3. Provide recommendations and resources to the sector.

ABLE: double approach

Nottingham Trent University  institution-wide dashboard  external provider  focus on engagement  engagement  progression & attainment KU Leuven & University Leiden  dashboard to support the live interaction between student and advisor  dedicated development  focus on study success  early academic performance  long term study success

ABLE: Nottingham

ABLE: Nottingham

ABLE: Leuven/Leiden

position students in peer group impact: how similar profiles did in the past guided planning of future study pathway

name student

2/3STELA project

Erasmus+ (562167-EPP-1-2015-1-BE-EPPKA3-PI-FORWARD) stela-project.eu

STELA: goals

 Forward looking cooperation project, launched in November 2015  How to apply learning analytics to support the transition from secondary to higher education?

• 1. student-centered
• 2. beyond identifying at-risk students: inclusive approach for every

student

• 3. beyond single courses: focus on the entire program
• 4. actionable feedback, ability to remediate

STELA: project ideas

four focus areas 1. performance 2. engagement 3. skills 4. well-being three main approaches 1. position students with respect to peers 2. show how other students with similar profile did in the past 3. feedback loop

STELA

Privacy by Design Privacy Engineering

3/3 Findings & Recommendations

What we’ve learned so far.

Learning Analytics in a traditional learning context is a challenge.

 Results from online learning, open universities, MOOC’s … cannot be transferred directly to traditional academic contexts.  Data is less straightforward to collect, information is sparser and it is difficult to get a complete picture.  Careful: we should not move away from traditional face-to-face teaching just for the sake of learning analytics.

REC 1: Focus on data that is available.

 academic performance

 strong relation to study success  widely available

 digital traces of behavior

 card swipes, in-class polls, lab attendance  virtual learning environment

 survey data

 strong body of knowledge in pedagogic research   REC 3

REC 2: Take learning analytics into account when redesigning learning.

 Which activities are expected?  Can these activities leave behind learning traces?

example: blended learning Mechanics 101 at KU Leuven

None of the students that accessed less then 1o online modules passed the exam. Most successful students finish at least 15 online modules.

REC 3: Think beyond the obvious.

 Learning analytics tends to have a “big data” bias.  We can learn from pedagogy and psychology

 self-reported (small) data  standardised tests.

example: academic skills feedback at KU Leuven

Shows a student how he or she is similar to peers. Shows how other students did last year.

privacy and ethics

 Privacy and ethics are big issues!  Rules and practices differ from country to country and from institution to institution.

 Nottingham Trent University (UK): full access (to card swipes, …).  KU Leuven (BE): vice-rector directly involved to unlock / fast track debate.  TU Delft (NL): lots of freedom for MOOC’s, restricted for regular students.  TU Graz (AU): very strict regulations.

 Every project has to spend a lot of effort = huge loss of resources and focus!

REC 4: There is a need for clear national and European policies for learning analytics.

 Traditional educational setting should not be disadvantaged compared with MOOC’s etc.  Eliminate differences between educational institutions and commercial entrants.  Quick win: provide model agreements for institutions to use and adapt to specific needs.

REC 5: Focus on actionable feedback.

 actionable feedback: focus on what can be improved  involve all stakeholders

 students, student unions  study advisors  teachers  management  policy makers

 provide guidelines and good examples

data silo’s

 data is stored in operational silo’s:

 academic performance  central IT system  behavioral data  virtual learning environment  survey data  different for every faculty

 difficult to assemble an holistic view on the student

REC 6: Develop common data requirements.

 example from Nottingham Trent University:

 UK government requires retention and attendance data  all UK institutions track, store and report this data  opportunity for Learning Analytics.

 But: institution should remain owner of the data!

REC 7: Stimulate flexible software solutions.

 Open source remains preferred license model for new projects.  However, universities often rely on proprietary software

 campus administration systems (e.g. SAP)  not necessary a bad choice!

 Project should focus on flexible solutions:

 open source components that can be integrated within existing systems.  reproducible blueprints preferred over highly specific solutions.

 Beware of not-invented-here syndrome (NIHS):

 many great puzzle pieces are readily available.

Example: STELA blueprint

Focus is on drawing the arrows

REC 8: Provide a checklist to evaluate tools and resources.

 Many offers available, of varying quality!  Both institutions and providers need guidelines to avoid mistakes.  Example: data ownership (cloud solutions)

REC 9: Keep on funding European collaboration projects.

 European collaboration is not always easy, but very stimulating!  Partners learn from each other and push progress in their own nations. Examples:

 ABLE: Leiden pushes for data access to support study advisors  STELA: Graz pushes for feedback to students

STELA + ABLE recommendations

 REC 1: Focus on data that is available.  REC 2: Take learning analytics into account when redesigning learning.  REC 3: Think beyond the obvious.  REC 4: need for clear national and European policies for learning analytics.  REC 5: Focus on actionable feedback.  REC 6: Develop common data requirements.  REC 7: Stimulate flexible software solutions.  REC 8: Provide a checklist to evaluate tools and resources.  REC 9: Keep on funding European collaboration projects.

conclusion

 useful projects for Learning Analytics in European context  interesting findings and recommendations so far …  … more to come soon