September 24 CMI Harmonization Meeting, Minutes September 17

[Tyde Richards]

The next CMI Harmonization meeting will be:

Date: September 24
Time: 8am US Pacific; 11am US Eastern; 4pm UK
Duration: 1 hour

Proposed Agenda

- Action recommendation: Renew, revise, withdraw IEEE 1484.11.3 (XML binding of data model)
- Core CMI Data Model (primary discussion topic)
- Infrastructure for collaboration

Please send comments on the agenda or draft minutes from last meeting (appended below) to: tyderi...@gmail.com

Dial-in:
Skype phone: +9900827049304412
Conventional phone: local number + access code 9304412
US 201-793-9022
United Kingdom 0870 0990 931

Visuals:
https://my.dimdim.com/letsi

Draft Minutes: September 17 CMI Harmonization Meeting

Attending

Avron Barr
John Blackmon
Ed Cohen
Don Holmes
Bill McDonald
Jyotirmaya Nanda
Aaron Perrin
Frank Polster
Jonathan Poltrack
Tyde Richards (convener)
Chris Sawwa
Guy Tourigny
Crispin Weston

Agenda

- Introduction to IEEE LTSC Joint Study Group
- Status and plan for IEEE LTSC 1484.11.1 Amendment
- Status of IEEE 1484.11.3 XML Binding of CMI Data Model
- Core CMI data model
- Status of communication protocols (IEEE 1484.11.1, LETSI RTWS)
- Discussion of business rationale for standardization issues
- Collaboration Infrastructure

- Introduction to IEEE LTSC Joint Study Group
In March 2010 the IEEE LTSC and LETSI established a joint study group under the aegis of the "intent to collaborate" document that ADL, the AICC, the IEEE LTSC and LETSI drafted approximately one year ago. The original scope was to identify parts of multi-part CMI standard; investigate ways to accelerate the standards development process; evaluate ways to collaborate with ISO-IEC. The study group planned to end in July but was extended through October to support this activity. It may be appropriate to initiate a new study group to reflect the expanded scope and set of collaborators represented in the CMI harmonization discussions.

- Status and plan for IEEE LTSC 1484.11.1 Amendment
The IEEE 1484.11.1 standard recently when through a 5 year reaffirmation ballot which means that it has been a standard for 5 years and has been reaffirmed as a standard for another 5 years. The IEEE LTSC has an approved project to amend the standard to correct deficiencies identified in balloting (which come from ADL) and to support use in a Web service context (which come from the LETSI RTWS project).

There are a total of 8 new data elements proposed by ADL and LETSI that are in scope for the amendment.

1. Both: attempt_number to content_object_communication
2. ADL: learner_role to content_object_communication
3. ADL: completion_threshold to objective_type
4. ADL: scaled_passing_score to objective_type
5. ADL: weight to objective_type
6. ADL: “extensions” to content_object_communication
7. LETSI: attempt_number to interaction_type
8. LETSI: identifier to comment_type

The participants recommended that all data element be addressed by an amendment except number 6 - an extension mechanism. There was no consensus about the features of an extension mechanism, or even if it is in scope for the standard. Suggestions were made to look at the extension mechanism in the SIF specification and some white papers on extension mechanisms that ADL will make available.

The following plan was proceeding with IEEE standardization of a 1484.11.1 amendment

Phase I - prepare a document for ballot by 12/2010
Phase II - IEEE Balloting 3+ months
Phase III - Fast-track through ISO-IEC 6-9 months
Phase IV - Procedure to make free upon ISO-IEC publication

The participants recommended that WG11 in the IEEE LTSC proceed with Phase I using the proposed data elements.

WG11 in the IEEE LTSC will resume weekly meeting the week of September 28, details to follow. Parties who would like to participate who are not currently members of WG11 should contact Tyde Richards, the acting chair, at: tyderi...@gmail.com

- Status of IEEE 1484.11.3 XML Binding of CMI Data Model
IEEE 1484.11.3 is an XML Binding of the IEEE 1484.11.1 cmi data model. It is due for a 5 year re-affirmation ballot and a decision needs to be made within the next couple weeks about re-affirming, revising, or withdrawing (default) the standard. The LETSI RTWS project established that the XML binding, while technically correct, is so complex that it cannot practically be implemented. The LETSI RTWS project developed a simplified XML binding using the IEEE standard as a point of departure. It is premature to revise the standard because the LETSI binding is evolving and the underlying data model itself is going to change. The majority position was to not reaffirm the 1484.11.3 standard. Consensus was not reached and there will be a brief discussion in the next telecon.

- Deferred topics
Discussion about a core CMI data model, communication protocols, business rationale, and collaboration infrastructure was mentioned but deferred due to lack of time. The priority topic for discussion at the next telecon is a core CMI data model.

[Daniel Rehak]

Tyde

An FYI on SC36. 
ISO has changed various procedures including balloting and fast track (as I understand effective immediately for new work).

As I understand one of the results is that the ballot period is extended, such that if SC36 authorizes something to go to ballot a meeting X, the ballot and associated process will close *after* the X+6 month meeting (or close enough to the meeting that it will not be on the agenda) and thus the work will not be considered until the X+12 month meeting.  So effectively a ballot cycle is 12 months.  To mitigate, before going to a year long ballot cycle outside of SC36, SC36 will iterate internally on their 6 moth cycles until they get 100% consensus on a work item.

Thus my guess on a realistic time line for SC36 to get something (2 months before meeting), ballot, approve, and publish is 18-24 months at best.
    - Dan

--
---
You are subscribed to the "CMI Harmonization" group.  More information can be found on the website: https://sites.google.com/a/adlnet.gov/scorm-harmonization/home

To post to this group, send email to cmi-harm...@adlnet.gov

To unsubscribe from this group, send email to
cmi-harmonizat...@adlnet.gov

For more options, visit this group at
http://groups.google.com/a/adlnet.gov/group/cmi-harmonization?hl=en_US?hl=en

--
Daniel R. Rehak, Ph.D.

ADL Technical Advisor
Skype: drrehak
Email:  daniel...@gmail.com
            daniel.r...@adlnet.gov
Twitter: @danielrehak
Web:   learningregistry.org
           adlnet.gov
           lsal.org

Google Voice: +1 412 301 3040
Tel:      +1 412 931 7317
Work:   +1 412 931 7319
Mob:    +1 412 805 7683

[Tia]

Hi,

I was not able to come to the meeting last week and I have a question
about one of the data elements. What is expected from the
"learner_role to content_object_communication?" Exactly what is this
element suppose to capture?

Thanks in advance.

On Sep 20, 1:54 pm, Tyde Richards <tydericha...@gmail.com> wrote:
> The next CMI Harmonization meeting will be:
>
> Date: September 24
> Time: 8am US Pacific; 11am US Eastern; 4pm UK
> Duration: 1 hour
>
> Proposed Agenda
>
> - Action recommendation: Renew, revise, withdraw IEEE 1484.11.3 (XML binding of data model)
> - Core CMI Data Model (primary discussion topic)
> - Infrastructure for collaboration
>

> Please send comments on the agenda or draft minutes from last meeting (appended below) to: tydericha...@gmail.com

> WG11 in the IEEE LTSC will resume weekly meeting the week of September 28, details to follow.  Parties who would like to participate who are not currently members of WG11 should contact Tyde Richards, the acting chair, at: tydericha...@gmail.com

[Jonathan Poltrack]

Hi Tia -

We received several requests for the learner_role element via the
ADLNet.gov help desk.

This use case is a result of SCORM not specifying reporting
requirements for LMSs. For example, when the content (SCO) sets
values, you know that if the SCO gets them during the same attempt,
the data model values will be the same as what was set earlier. This
is a SCORM requirement. However, there is no requirement for LMSs to
provide the data model values for later use. Of course LMSs do
provide some of this data (individual/class/yearly - scores, pass/
fail, etc) but this is not a requirement in the SCORM. This is used
by LMS vendors as a differentiator in their products.

So - on to the learner_role element... Some of the ADL community
wanted the ability for an instructor, administrator, etc to view
scores/interactions/success_status/etc - but this was done differently
in different LMS systems. Further, in many systems the interactions
(cmi.interactions.X) are not available via LMS reporting screens after
the SCO is complete. So, some in the community are developing their
content with several views.

For example the cmi.learner_role element could be used by the content
developer to enable the following scenario: An instructor views the
SCO. The SCO gets the learner_role and its value is "instructor". The
SCO skips or hides the instructional content and instead displays a
summary of the interactions. If this approach was taken, the
interaction reporting would be consistent regardless of the LMS system
being used.

I hope this helps.

[Estes Ethan]

In the scenario given how would the Instructors current session in the sco pull the scores/interactions/success_status/etc from the students cmi data stores in the db? Is there a matching set of additions that allow a sco with a instructor logged in to poll the LMS db for that info? The db structure will be different vendor to vendor so there would need to be a required api that a sco could utilize to pull in that 3rd party data, correct?
-Ethan

*****************************************
Ethan Estes
Bullzi Security, Inc.
CourseWare Developer
email: ees...@bullzisecurity.com
mobile: (248) 770-0197
Suite 1073
1052 Montgomery Road
Altamonte Springs FL 32714
USA

Hi Tia -

I hope this helps.

--

[Mike Rustici]

Jono,

Thanks for the explanation of the proposed role element. I actually had quite a different understanding of its proposed purpose.

To me, "role" indicates more of a job role. For instance in a course about a medical device, one might want to tailor the instruction differently for doctors, nurses and medical devise salespeople. I've been a bit skeptical of including such an element because, while it would be quite useful and powerful, it would be rather complex to define in an interoperable manner without just leaving it as a free text string. 

The functionality you describe seems to me to be a more natural fit for an extension to the "mode" element rather than a role. What you are really wanting to specify is "how should the content behave during this launch?". A person with the role of instructor could sometimes want to launch the content as a learner would in "normal" mode, and other times want to go see what the learner did in "grading" mode. 

I'm all for the additional capabilities you mention, but I don't think a new data model element is the correct path forward for this particular case.

Tyde,

I submitted a bunch of additional comments at the IEEE meeting back in January and also at the recent meeting in Alexandria. They are available on the CMI Harmonization Google Group and also (more leigbly) in a shared Google Doc. I believe John Campbell also submitted a list of comments in January. Will these be considered as well?

Most important among these is "Interactions need to allow for more description of the question". That is currently a huge shortcoming of the data model that we hear complaints about all the time.

Thanks,

Mike

--
Mike Rustici
www.scorm.com
615.550.9523

[Thropp, Schawn]

To me this use case sounds like a community of interest extension to the CMI data model.  We have to be careful when reviewing these issue to determine if they belong in the core standard or whether then need to be supported by some extension model.  Especially since we have been talking about a look at a newer "core model" supporting extensions.  I would hate to add this support into an update to the CMI Standard and then remove it during the core model work.

 

The IEEE CMI Working group needs a set of criteria for including additional data model elements, especially in light of the core model concept.  I have always been a fan of smaller more focused and categorically defined data models, i don't think the time is right for patchwork inclusion of new elements.  These might allow some other use cases from communities of interest to be solved, but makes for a more brittle data model standard.

 

We need to think small, focused data models that support extensions for community of interests.

 

Schawn-

[Tyde Richards]

On Sep 22, 2010, at 6:58 AM, Mike Rustici wrote:

> I submitted a bunch of additional comments at the IEEE meeting back in January and also at the recent meeting in Alexandria. They are available on the CMI Harmonization Google Group and also (more leigbly) in a shared Google Doc. I believe John Campbell also submitted a list of comments in January. Will these be considered as well?
>
> Most important among these is "Interactions need to allow for more description of the question". That is currently a huge shortcoming of the data model that we hear complaints about all the time.

We can discuss on Friday because a significant decision needs to be made. Apologies for a bit of background arcana ...

Currently WG11 has an approved "project authorization request" for an amendment to 1484.11.1 (cmi data model) with a very narrow scope:

"This Amendment corrects technical errors identified in the five year reaffirmation ballot of the 1484.11.1-2004 standard and defines updates for use in a Web service context. "

Given this scope, many of Mike's comments are out of scope. What is in scope are the 8 data elements that we discussed in last Friday's meeting, of which there was consensus to address all except an extension mechanism in the amendment.

An amendment usually contains 2-3 new features appended to the base standard. I have spoken several times this week with our IEEE standards liaison about what we are trying to do. He believes that given the anticipated number of changes and the fact that the standard has been out for 5 years that we should be doing a "revision", not an amendment. A revision can address more issues and the material is incorporated into the standard which makes for a more legible document. The next opportunity to request a change from an amendment to a revision is October 18 for submission with approval in December. Subsequent opportunities will occur every three months.

In this Friday's meeting I was planning to propose that I transact the process to change the standards update from an amendment to a revision, keeping the same narrow scope:

"This Revision corrects technical errors identified in the five year reaffirmation ballot of the 1484.11.1-2004 standard and defines updates for use in a Web service context. "

Alternatively, we can expand the scope of the revision to accommodate the full set of comments from Mike and others. This would have technical, procedural, and resource consequences.

Technically, it would open the door for a revision that could be significantly different from the current data model and that could reflect ADL/AICC "core" considerations.

Procedurally, it would take more time. Ballot comments are only actionable within the scope. The narrower the scope, the more rapid the standardization process because the range of actionable comments is narrower. However, speed is not the only consideration. With the current narrow scope, Mike's priority issue of "interactions need to allow for more description of the question" would be out of scope. There are also considerations of cost/benefit, demand, and value.

Resource wise, the WG11 membership is small to address the current scope. For an increased scope I would need to see some evidence of an increased commitment to participate in the WG. In particular, if individuals strongly want features added to the standard they should participate in the WG to the extent necessary to define those features.

So, on Friday I would like to get a sense of the group on going with the current scope (default), or a revised scope.

[Mike Rustici]

(tried to post this on Thursday but had a Google account mix-up)

Hi Tyde,

I will try to make the Friday meeting, but if I can, it will only be
for a
few minutes. I have two points to make in response to that last email:

1) You can count on our participation when it comes time to do the
technical
work

2) I don't understand why an element that adds a description for
learner
response and correct response is of significantly greater scope than
the
other proposed element additions. I would also add that this data
(which
provides the ability to do much better reporting on test results) is
fundamentally important and a glaring hole in the current data model.
It is
a *huge* frustration for current SCORM users. To do a revision without
including it would be downright foolish in my mind. In my experience,
the
number of people who just want to be able to do basic test reporting
would
out number the people who would require any of the other 8 changes
combined
by at least a factor of 10.

[Can you tell where I stand?]

Mike