March 2, 2017
11 EST/10 CST/9 MST/8 PST
Attending: Ellen Meiselman, David Topps, Co-Chairs; Susan Albright, Scott Kroyer, Rosalyn Scott, Valerie Smothers, and Luke Woodham
1. Review minutes
The minutes were accepted as submitted.
2. New Members: Introductions
- Scott Kroyer - Medhub /eValue
Adam Dodson - Johns Hopkins, Lead Simulation specialist
Scott Kroyer - Medhub /eValue – Scott described MedHub’s current focus on redesigning their entire platform using XAPI as the backbone. Different systems will use xAPI to feed in and report out to one another. Ellen thought that was a great integration point between pieces of the system. She asked about their use of xAPI for multiple purposes. Scott provided an example of making evaluations standalone related to learning event. There may be many things that people want to evaluate within a single event or course. David commented it is a challenge to associate an activity stream with a particular context when a learning activity has multiple pieces. Scott agreed especially in terms of external interoperability.
3. Blended Simulation Profiles
Preliminary discussion of what pieces that already exist could be pulled in, and what needs to be tracked, and why.
See abstract for description of blending virtual patients and mannequins:
- What makes sense to track
- What questions need to be answered?
- Possible functional reasons for tracking: integration between mannequin and branching VP, adaptive features, etc.
Existing profiles and vocabularies that might be useful
On the xAPI CoP website:
And from TinCan Registry:
Rosalyn shared her experience at the VA related to integration of Virtual Patient (VP) cases with other simulation modalities. They used two different blended learning approaches. In the first approach, they removed mentor nodes (feedback) from an existing VP case, added a node to prompt interaction with the mannekin (ie “now it’s time to interview the patient” or “now it’s time to examine the patient”) and created a facilitator’s guide from the mentor nodes. The learner would interact with the mannekin, and based on the responses or findings form the mannekin, make decisions in the VP case. Rosalyn noted one of the advantages of this method was learners had consistent presentation with an equivalent experience for everyone. The simulation experience extended the time line for the case.
In the second approach, they created a VP case modelled around methodology for communication. She provided the example of using a standardized patient with different clinical problems where end of life discussions were appropriate. The diagnosis was different but the communication model remained the same. They also use task training to provide consistent experience across a group. David noted that there are tools that allow for Natural Language processing of student entered narrative that offers an efficient option for reviewing student communication. This can be used with xAPI to track the text entered. Valerie asked if the tools used xAPI. David confirmed it tracks what is discussed in the narrative, NLP text based. Rosalyn expressed concern with this generation’s ability to communicate with patients verbally. David commented they tracked text based conversations that transmit to a learner record store. Ellen questioned Rosalyn whether a facilitator was in the room in her modality. Rosalyn confirmed it is videotaped and the debriefing includes reviewing video tape.
Valerie summarized the potential streams of data as: video, data entered by facilitator, data from the VP activity, and data from the mannequin or task trainer. Rosalyn added data from an observer.
Luke commented that WAVES’ focus is primarily on VP in group sessions, so tracking would be difficult; however, ways to capture other activity around VP sessions would be of interest.
Ellen noted there are different profiles suggested breaking these into different use cases. Valerie added the need for guidance on how to use multiple profiles together and what is needed to do that. David suggested making sure profiles are layered and interconnected, and when using a particular verb, the meaning of it is sufficiently similar and scalable up and down. He noted cross linking is important. He recommended standardizing a way to link and describe profiles.
4. Update on Open Labyrinth 4 project
- Plans for incorporation of xAPI in the re-architected OLab4.
David noted that he plans for incorporation of xAPI in the re-architected OLab4. He is interested in input on how to do this. Pulling information from multiple sources will be a core piece and will solve data reporting problems for them. They are interested in having other contributors for XAPI.
5. Highlights of DISC (Data Interoperability Standards Consortium) call
xAPI profile interoperability effort (A kickoff webinar is scheduled for Monday, 27 March 2017, 11:00am Eastern - details to come)
Machine readable profiles
Disseminate and curate profiles by credible organizations that could be subscribed to
Generate profile sets that would power authoring tools
These would automatically put in statements you need
Help with validation of statements by LRS
Inform general reporting and visualization tools
Ellen mentioned DISC looked into providing certification of xAPI. They are now working on profiles to make profiles interoperable and machine readable. A kickoff webinar is scheduled for March 27 with the hope of having communities of practice generate profile sets and be able to improve validation and context.
Susan questioned whether the Curriculum Inventory and xAPI standard are related. Valerie clarified xAPI mainly tracks learner activity. The Curriculum Inventory provides a standard for curriculum data. They are beginning to develop an API for competency data. Ellen noted that xAPI has an Activity Profile. The system producing xAPI statements may consume curriculum data to produce an Activity profile.
David suggested fleshing out using multiple profiles together and volunteered to work with Valerie on that. Valerie added, she, Rosalyn, Luke and Deb had a call tomorrow to discuss further. She would send an invitation to David as well.
- Discussion with xAPIApps
- ADL-submitted abstracts
- Using xAPI to connect high-fidelity simulators to an LMS
Chad Jackson, American College of Chest Physicians
Andy Hicken, Web Courseworks
Rosalyn mentioned that xAPI will have value in the development of a VA virtual medical center. They are creating face to face courses for staff and patients using flipped classroom strategy and plan to put didactic pieces in the Virtual Medical Center. There will be different tracks with different requirements to be tracked. xAPI could be used for tracking. They plan to issue badges for patients.