Membership Policy

[Ian McIntosh]

Brad issued an e-mail making the following proposals:

Policy for SJTAG Core Member Attendance:

1. To become a core member, you must attend at least two consecutive meetings
2. As a core member, you must attend the meetings either
   • by physically attending the meeting on the conference bridge
   • respond to the draft minutes with feedback as proxy information to be added to the minutes
3. Voting privileges will be granted for members who have attended the past two consecutive meetings or who have at most one excused absence for the past two consecutive meetings and one meeting attendance
4. Members must send email to the chairman of the group at least 24 hours in advance of a meeting indicating they will not be able to attend to be considered an excused absence from the meeting. This email also provides an indication that the person is interested in attendance and the work of the group, but has business or personal conflicts with the meeting time. This person, while excused, is responsible to provide feedback to the draft minutes of the meeting to complete the qualification of the excused absence.
5. A draft release of the meeting minutes will only be sent to the meeting attendees and to those who indicated they need an excused absence via an email message to the chairman.
6. Persons who have more than four consecutive unexcused absences will no longer be considered part of the core group and will need to begin the process of membership again to reenter the group

While I am in agreement with Brad's intention to constrain Core Group membership to people who are actively supporting the initiative, I was concerned that the detail of the proposal would lead to a continually fluctuating membership: It is my belief that we need a relatively stable Core Group membership in order to move forward and to encourage members to take ownership of tasks within the group.

So, I have created a graphic (linked here, PDF file), using attendance data from the 27 meetings held to date, to show how membership would be affected if Brad's scheme had been in place. I also present a counter-proposal which I believe gives a more stable Core Group membership.

Firstly, there are assumptions in this analysis:

1. That the liklihood of any person attending a meeting is not affected by any membership criteria.
2. That the level of proxy contributions will also remain unchanged.
   I think these are both incorrect assumptions, but I have no way of telling what effect invoking a membership criteria will have on these.

Sheet 1 of the graphic show's Brad's proposal in place. The light green boxes indicate the meetings at which the individual would be considered a Core Group member given the qualification rule in point 1 and disqualification rule in point 6 of Brad's e-mail. The red bars indicate changes in the Core Group membership.
On average, membership changes every 2-3weeks and there are periods where there are weekly changes.

I feel that an alternative method would be to assess contributions at three monthly intervals. The people having attended a set percentage of meeting during the preceeding 3 months would then form the Core Group for the following three months. I initially had the thought that the qualification would be attending something like two-thirds or maybe three-quarters of the meetings, but when I set out to chart this it was apparent that very few people would qualify! As result, I set the threshold to 50%, and the effect is shown on sheet 2. This still did not seem satisfactory as during the first quarter of 2008 we would only (realistically) have five people in the Core Group. Sheet 3 drops the qualification to attending 25% of meetings and gets closer to what I intuitively felt was the "right" membership.

Brad's proposal effectively requires that someone attends at least one meeting in four to maintain membership, so my counter-proposal is not really much different - it just has more hysteresis! However there is a further implication of using a threshold as low as this: The logistics of approving meeting minutes could become very difficult, if attendances are commonly this infrequent.
So perhaps we ought to "set the bar a little higher" to encourage more consistent attendance?

[Peter.Horwood]

I have no preference on the methods proposed by Brad or Ian, my view is that to be called a core member you must participate in some fashion.

[Tim Pender]

I follow Ian proposal but did not think the graphics agree with the requirement.
Are Proxy submissions indicated with a "1" for attendance or is it hidden.
My assumptions that are proxy attendance is shown as a 1 in the spreadsheet.

In the pdf link, Ian V1 (member atleast 50%)Center grouping I expected Carl N and Carl W to be highlighted green since the attendance was 50% or greater.
AL H was highlighted but he only attended 1 out of 12.
TIm P was a new joiner on 3/10 but not highlighted.

In the pdf link, Ian V2 (member atleast 25%)
Highlighted Jim W with only 1 of 12.

[Ian McIntosh]

For the purposes of the analysis, I counted proxy contributions, where they had been noted in the minutes, as if it were full attendance, as that was part of Brad's proposal. So your assumption is correct.

For "Ian V1" and "Ian V2" it is the attendance in the preceding three months that sets the qualification for the following three months, so to use Jim W in V2 as an example, between 5/10 (European date format!) and 17/12, Jim W attended 4 meetings out of a possible 11 (>25%), so qualified for the following three months. During that second quarter, only one meeting was attended, so at the end of the quarter the qualification was lost. On the same chart, Tim P joined on 10/03 but had yet to qualify for membership, but then attended all possible meetings up to the end of the quarter, so qualified as a member from 07/04 (and until 30/06, were the chart to go that far).

My proposal isn't complete, because I'm unsure where the threshold should be set: The imposition of a membership policy might influence attendance levels and so change the dynamic of the whole thing, but I suspect that, for those who attend or contribute by proxy fairly regularly, it won't have any observable effect.