Loomio

[product idea] Convergence options

AI Alanna Irving Public Seen by 64

DISCLAIMER: This is a musing discussion about things I know are not a priority this quarter, but I'm putting this out there because it's been on my mind and I think it could be useful in the longer term.


I have long resisted ideas that I feel could distract, muddy, or try to shortcut the full Loomio consensus decision process, because I know from experience that to get the full value you have to bring a measure of discipline to it and see it through. Simultaneous proposals, multiple-choice polls, etc, all fall into this category.

For complex consent-based decision-making, anything that comes to premature convergence is likely to oppress minority voices, miss diverse viewpoints, create alienation, and leave good ideas on the table. My resistance to deviation from the core Loomio decision-making process came from knowing that won't get people where they need to go.

However, we have heard requests for these types of features for years. Sometimes I would still argue that users are just looking for an easy way out of something that's not really meant to be so easy - but other times it's coming from a very valid desire to do something other than a complex consent-based decision.

Here's my view of the structure of a Loomio discussion and decision, which mirrors important truths about effective group communication, and the divergence/convergence "diamond" that enables collective action.

The thing is, though, there are different types of convergence.

If we made it really clear that there are other types of converge that are not a decision, I think we could make room for other very useful functionality, following the same divergence/convergence flow, while making use of all the architecture of a Loomio group, (such as discussion threads, group memberships, notifications, deadlines, outcomes statements, etc) - without muddying the waters when a group does want to make a real consensus decision.

We already know a fair amount about different kinds of convergence groups find useful... We know some "recipes" that users could call on to use as processes.

Guiding users to pick between types of converge is a powerful signalling system about what kind of convergence it is, and what sort of mandate it has. This can help groups to understand that while a certain kind of convergence could give them useful information, it might not hold definitive power as a conclusion. I imagine we'd still want to have things like outcome statements, to leave room for interpretation of results.

Here are a few convergence processes. I imagine they would show up on the "convergence" side of your Loomio thread, like proposals do now. The first one is our current proposal process.

One thing I took away from the Voting Methods course I did is that there are no good voting methods. They all have flaws and biases built in. We could use the convergence pane to offer some quite creative interfaces that reflect this, such as this design I did for an interface combining Borda Count, Condorcet Count, Approval Voting, Ranked Choice, and Majority Voting.

There's a lot of scope for creative designs in displaying results as well, that make it clear to people that all data is up for interpretation, and they have to view voting as a human process that has values implications.

Combining flexibility about the convergence pane with plugins could open up all kinds of possibilities for groups to design and share useful types of convergence, while keeping it in our structure with things like context, discussion topics, comments, deadlines, outcomes statements, etc - practical things that help convergence be successful and lead to collective action.

SB

Susan Basterfield Wed 17 Aug 2016 11:36PM

ooo this makes my tummy all excited

S

Simon Thu 18 Aug 2016 9:25PM

With in-person groups that have finished exploring issues from multiple perspectives (divergence) and are ready to start converging, I often ask participants to write down all their ideas for actions/recommendations as individuals. Once this is done, participant can share their ideas and same/similar ideas can be clustered together. Areas of common ground can be made visible and attention can be given to some of the less common ideas.

I spoke with @susanbasterfield about this convergence process yesterday and she knows of a online tool or two that can do this sort of work online. I think it would be a useful if this Loomio could facilitate this sort of process.

AI

Alanna Irving Tue 18 Oct 2016 1:25AM

Example of a convergence hack of Loomio.

S

Simon Tue 18 Oct 2016 8:08AM

I've seen people hack Loomio like this before and it works fine as long as you have less than 5 options. To see how messy it can get when there are many options , check out the following thread from Christchurch - https://www.loomio.org/d/ZveRHtWV/strategy-action-options-what-could-we-do-. I doff my hat to this group for their persistence!

How could the Christchurch convergence process be done better and how could the platform support this? Could you use something like https://Pol.is for this phase (if you had enough participants)? The real time analysis and visualisations could be helpful. Group Map - https://www.groupmap.com/ - is another tool that could help participants see areas of agreement and disagreement as well as outliers.

BH

Bob Haugen Tue 18 Oct 2016 11:15AM

Thanks for this discussion, may it converge! It looks like it could be the seed of the long-awaited but still unformulated group conversation for action protocol.

JK

James Kiesel Tue 18 Oct 2016 4:09PM

Simplest thing to do from our end is support multiple choice polling (NB: it's not that simple). I think @richarddbartlett had a design floating around for that at some point, and I have it on my list of things to play around with next year.

GC

Greg Cassel Tue 18 Oct 2016 5:30PM

Regarding multiple choice polling, per @gdpelican 's comment above: I think facebook's simple "Create Poll" feature works pretty well. (And I'm quite critical of fb's admin tools and participant tools in general.) We've used fb polls exclusively for 'Loomio-like' decisions in my main discussion group for over 1 1/2 years, because my fellow admins are (1) practically allergic to technology and (2) rightfully resistant to adding (and regularly visiting) a second organizational platform.

One key thing IMO is for poll creators to allow or not allow participants to select multiple options. That's a deeply important aspect of poll design. It's a decision which poll creators often mess up, but I consider that a cultural/awareness issue which tech can't solve.

RDB

Richard D. Bartlett Tue 18 Oct 2016 7:05PM

Yep I think this design is 90% complete.

S

Simon Thu 20 Oct 2016 7:42PM

An issue with multi-choice polling is that a seemingly clear 'result' can obscure differences, which might bite later. A multi-choice polling tool, IMO, can help a group deliberate on what to do and for that reason giving is important. Of course, a group can (and should) decide to discuss the results of a poll to clarify its meaning but I liked to see any multi-choice polling tool prompting this to happen. A feature that could encourage this is a prompt to give reasons for each choice - just like the current Loomio voting tool - and a way to collate reasons for each responses so its easy for participants to see differences, etc as they decide what to do next.

What I like about the Group Map tool - https://www.groupmap.com/ - is that it lets group moderators select a range of templates/structures to collate/visualise a group's responses/choices (e.g. using SWOT format)

AI

Alanna Irving Thu 20 Oct 2016 10:48PM

Yes these considerations are exactly why we haven't released polling features up to now