September 17, 2015
9 MDT/11 EDT/16 BST
Attending: Tom Creighton, Erick Emde, Lucas Huang, Ellen Meiselman, Valerie Smothers and David Topps..
David provided an update from the AMEE meeting last week. One presentation was a project from Karolinska that included an Open Labyrinth virtual patient as part of a MOOC. They had 19,000 registrants. 700 individuals completed the course, and there were 600 concurrent users of one VP case. They had previously done statistics by hand, but heard about xAPI and tapped into it. Valerie will contact Nabil to discuss further.
1 Review minutes
The minutes were accepted as submitted.
2 Review draft proposal for Executive Committee
David provided comments to the proposal which Valerie circulated prior to the meeting. Valerie gave an overview of the formal ANSI standards development process whereby standards are proposed and reviewed by the Executive Committee. It was determined to launch an official standards development effort for a MedBiquitous xAPI profile. The profile would be made available under our open license.
David questioned including patient health assessment and education under Scope. He suggested possibly changing it to patient outcomes research, limiting it to education. That does expand the scope of this work which would have more impact, but also means it could become a several year project. Erick thought that piece could be added at a later point. Valerie suggested including a caveat that it “may” include patient health assessment. David thought it was worthwhile being aware of what is happening in that area. He would hesitate to include it in the scope. Valerie explained the reason it was included was from a related use case on patient health assessment. She suggested saying it was out of scope for version 1. Ellen described the vision was to use xAPI track and do assessments with patients. The ideas were pre-research kit; using technologies to drive decision based on activities of patient. David was concerned about tracking all patient research. Valerie will make the change to “may include.”
David was concerned about using the term “social media” to describe the Activity Stream format. Valerie noted the proposal will be going to the Executive Committee who are a fairly sophisticated audience. It will be posted on the website and viewed by a larger audience, but could be reframed from. She suggested adding “modern Internet technology” to replace “social media.” Ellen mentioned explaining the importance of systems integration, and how it is not just e-learning.
David continued with a conversation he and Jason had. David was working under the assumption it is crucial to use existing verbs from the activity stream set. ADL is deprecating use of Activity Stream verbs; however, Jason is open to groups creating new verbs, potentially based on the activity stream verbs, if there isn’t a good fit with an existing verb. Valerie forwarded a note from Jason on Aug 26th regarding his conversation with David and the process for adding new verbs. Valerie agreed to talk with Jason to see how to proceed. She encouraged the group to comment on the discussion by email. David suggested she recirculate the Google documents. Valerie noted they are on the wiki under important links.
3 Continue review of xAPI gap analysis.
Valerie relayed that the recommendations for xAPI implementation are influencing the broader RESTful design guidelines developed by the MedBiquitous Technical Steering Committee (TSC). The TSC volunteered to have a joint meeting with this group to iron out any issues that need to be discussed. Ellen would love to hear what they have to say about what others are doing. Valerie commended this group on being pioneers in this area. She suggested including a section in the guidelines for creating PURLs. The group agreed. Ellen wanted input on tracking learning activities to competencies across all modalities. She noted it is hard to picture how you would track things to competency without a breakdown of where things fit. She asked if there was anyone who has done work to track data based on the performance framework and competency framework. She has seen a proof of concept system. Valerie noted Tom put that together. Tom mentioned a video that explains what he did. He looked at different ways to attach competency objects with xAPI statements.
David commented that this relates to the gap analysis section Competency and Performance Levels to Statement Linking. Under Possible Solutions: Represent the competency as an Activity object, using the competency URI as the Activity’s URI. The positives are the competency activity object can be used in different ways; as the object of a statement, as a Context Activity, and as a related different statement. The negatives are, only one Activity ID may be used when querying an LRS.
Tom suggested grabbing the larger scope and then filtering what is received. Ellen asked about using graph databases for these relationships. Tom commented that would require storing data outside the LRS, which is not fully fleshed out. Ellen commented that she liked the idea of using an activity object to represent a competency. If you did so, could you use it in both places, as context of an activity and an activity object? Tom confirmed. He wants to be able to attach a competency object or competency framework to multiple statements, as context activity and evaluate based on other roles. Lucas asked about representing a score on a likert scale. Valerie had the same question; how does the performance framework fit it in to that context? Tom offered to think about that.
Valerie noted we need to decide if we are going to say that Tom achieved a performance level of 4 for competency X using the performance framework developed by XYZ and here is the ID. Tom mentioned you could link to a performance framework as an activity and define the performance level scales (ex. Tetras).
4 Open discussion
- Valerie will edit the proposal based on the group’s feedback.
- The group will revist how performance frameworks and competencies are represented in future meetings.