Loomio
Sun 1 Nov 2015

Abstention - Do Loomio voting options foster participation ?

FB
Fabio Balli Public Seen by 532

Today, a proposition may be read by 50 people, but only 20 vote. What if the other 30 were made visible ? Would that not increase the sense of belonging in the community, acknowledging the commitment of both the proposer and the readers of the proposal ?

The "orange light" resulting of "abstain" is not representative of very different voting intentions (undecided, not caring, blank cheque, no legitimacy, etc.). Comments, while helpful, do not impact the interpretation of the charts.

To solve this, I propose to 1) change the orange icon title to "undecided", and 2) create a white icon "abstain".

FB

Fabio Balli started a proposal Sun 1 Nov 2015

Change the orange icon title to "undecided" and create a white icon "abstain" Closed Wed 4 Nov 2015

Results
Agree - 3
Abstain - 3
Disagree - 3
Block - 3
8 people have voted (0%)
FB

Fabio Balli
Agree
Sun 1 Nov 2015

RJ

Raphaël Jadot
Abstain
Sun 1 Nov 2015

Interesting but i guess the low rate is more a global Internet syndrom than due to anything in loomio… i may change my vote later but still am not certain :)

JK

James Kiesel
Disagree
Sun 1 Nov 2015

This complicates the simple premise of the software, in my humble opinion.

I may be able to be talked into clarifying what the positions mean per proposal (which has been suggested a few times before), but not changing this core tenet of Loomio.

GC

Greg Cassel
Disagree
Sun 1 Nov 2015

Low 'voter turnout' is prevalent in most online communities, even for rare scheduled events such as elections of annual officers, etcetera. There are many reasons for low participation, which can be addressed both generically and per-community.

O

Ohrar
Agree
Sun 1 Nov 2015

Why not just test it? The rate of undecided people seems an interesting figure to me.

JK(

Joop Kiefte (LaPingvino)
Disagree
Sun 1 Nov 2015

No real need. I would go for a FB-like option x people read this proposal, and you could mark that as white in the chart, but not with a new button. Conscious choise for this would be Abstain.

LF

Luke Flegg
Agree
Tue 3 Nov 2015

Seems to me the main resistance is people don't like too many colours because it doesn't look as pretty. I think white is smart, because it's a lack of any colour (because it's a lack of any opinion) while orange becomes an active expression of :-S

RG

Rob Guthrie
Disagree
Tue 3 Nov 2015

What a great proposal. This has generated some really good ideas. Thank you OP. I love the idea that whitespace on the pie represents people who've read the proposal and did not vote

FB

Fabio Balli started a proposal Wed 4 Nov 2015

Include people who viewed a tread but did not vote as a white slice in the pie chart. Closed Sat 7 Nov 2015

Results
Agree - 7
Abstain - 7
Disagree - 7
Block - 7
19 people have voted (2%)
DU

[deactivated account]
Agree
Wed 4 Nov 2015

This is actually a very good idea.

CE

Colin England
Agree
Wed 4 Nov 2015

Looks good.

KL

Ken LeFebvre
Agree
Wed 4 Nov 2015

KL

Ken LeFebvre
Agree
Wed 4 Nov 2015

MB

Matthew Bartlett
Disagree
Wed 4 Nov 2015

Viewing a thread feels too light-weight and ambiguous an action to be concretely represented in the pie chart

GC

Greg Cassel
Disagree
Wed 4 Nov 2015

I really like this general train of thought, but I don't think we're 'there yet'. IMO Matthew said it perfectly: views feel too lightweight and ambiguous for direct representation in the pies.

HS

Hannah Salmon
Disagree
Wed 4 Nov 2015

TT

Theodore Taptiklis
Disagree
Wed 4 Nov 2015

FB

Fabio Balli
Agree
Thu 5 Nov 2015

Z

zack
Agree
Thu 5 Nov 2015

I vote for this proposal because this kind of change is needed in my view. but I would change it to people who commented not just looked

JK(

Joop Kiefte (LaPingvino)
Agree
Thu 5 Nov 2015

Much more in line with what we really want to see :).

DM

Damon Meledones
Disagree
Fri 6 Nov 2015

At a minimum this needs to be an option, not default behavior.

PF

Paul Fenwick
Disagree
Fri 6 Nov 2015

Weak disagree. I concur with the position that viewing is too weak an action to be meaningful. I'll open lots of tabs (I have 196 open right now), but that doesn't mean I've read any of them.

DS

Danyl Strype
Disagree
Fri 6 Nov 2015

Like Zack, I would agree if the proposal was based on comments, not views.

MDB

Mathijs de Bruin
Disagree
Fri 6 Nov 2015

The idea that viewing a resource contributes a change violates the expectations (and RFC's) concerning HTTP and web applications.

DU

[deactivated account]
Disagree
Fri 6 Nov 2015

agree with Matthew and Zack

CZ

Chris Zumbrunn
Disagree
Fri 6 Nov 2015

Hitting a URL does not mean someone actually read the content on that page. Even commenting in the discussion does in my opinion not constitute abstaining.

NK

Neofytos Kolokotronis
Disagree
Fri 6 Nov 2015

I think the pie should only include people that engaged with the voting.
Alternatively, there could be sub-options for the those voting 'abstain' to explain their reasoning. Possible scenarios would be 'don't care, undecided, no time to follow, etc.

RJ

Raphaël Jadot
Agree
Fri 6 Nov 2015

Maybe not necessarily on the chart itself but I find the idea of having an amount of lurking users awesome.

BK

Benjamin Knight
Disagree
Fri 6 Nov 2015

I love the idea of displaying engagement stats, but don't think it should be presented in the pie graph.

JW

Jochen Walter
Agree
Fri 6 Nov 2015

Perhaps the "viewers" should be shown as "not decided yet" or "still not decided".

JW

Jochen Walter
Disagree
Fri 6 Nov 2015

Changed because there is enough information about the not decided. They have to accept that democracy means to be part of a decision. And the group have to improve the contact to those for that they take part.

MB

Matthew Bartlett Sun 1 Nov 2015

@fabioballi — I really support your wish to 'increase the sense of belonging in the community'. However it's hard for me to see how changing the voting options or colours will help that. Maybe there are other ways we could achieve that?

GC

Greg Cassel Sun 1 Nov 2015

I think that online communities can benefit from peer feedback/ sociometric tools which are much more finely granulated and nuanced than stuff such as "like", "upvote/downvote" and "agree/abstain/disagree/block"; however, I think that a four point system is quite strong for formal decision process.

FB

Fabio Balli Sun 1 Nov 2015

Thanks for these first feedbacks. :)

@raphaeljadot @gregorycassel For me, it is not so much about having more people take part to the discussion but about making the one who invest time in reading the topic visible (immaterial contribution).

Let's say we propose a topic. It has a totally different signification if we see 10 votes and have no clue about the others, or if we see them, but also know that 30 people have taken the time to read the proposal. Maybe that could be a "people who read" automatic number @matthewbartlett ?

@jameskiesel maybe that is just about wording, not adding an element ? From what I understand, the orange "abstain" is in fact "neither agree, nor disagree", and the "undecided" - which I saw is already used for non voters -, is "non voters".

MB

Matthew Bartlett Mon 2 Nov 2015

Ah — I hadn't understood that. I love the idea of making lurking more visible.

GC

Greg Cassel Sun 1 Nov 2015

It's not possible to directly register whether or not someone has taken the time to properly read a proposal. If someone has indeed read a proposal carefully and they feel undecided, they can use Abstain along with a comment such as "I'm undecided." I think that this is a good communication tactic, but I don't think it's been strongly cultivated in groups which I belong to.

RJ

Raphaël Jadot Sun 1 Nov 2015

@fabioballi wdyt if then would be automatically added the amount of
* unique clics on a proposal
* connected users who opened the proposal

It certainly can be done automatically :)

TT

Theodore Taptiklis Mon 2 Nov 2015

@fabioballi I think one of the strengths of Loomio's present format is that 'lurking' is legitimate...it's possible to feel part of the conversation without actively participating. Ambiguity can be a strength not a weakness here. I'm sympathetic to your concern (How many people are watching but not contributing?) but I don't think trying to crytallise this uncertainty is necessarily the solution.

RG

Rob Guthrie Mon 2 Nov 2015

Being able to see who has viewed a proposal might help the group understand more about how a proposal was received. And it would not require the user take any extra actions.

DS

Danyl Strype Tue 3 Nov 2015

The four options (agree, disagree, abstain, block) are based on a consensus process which has a long tradition of use in face-to-face meetings. I agree that changing the buttons around would be very disruptive without adding much of value. However, I really do like the idea of using white space in the pie chart to proportionally represent the group members who have at least glanced at the thread (as opposed to being completely disinterested), but not taken a position on the proposal. Another variant could be white space for those who have commented, indicating some degree of active interest, but not taken a position yet.

KL

Ken LeFebvre Wed 4 Nov 2015

I like the idea of, perhaps, counting those who comment without taking a position. In fact, the more i think about it, i think I'd recommend that there be a line of text (as someone else suggested), that says "x people have viewed this proposal" and put a white wedge in the pie for unique commenters without a position.

MB

Matthew Bartlett Tue 3 Nov 2015

@strypey feel like mocking up some examples of how the pie change idea could look?

GC

Greg Cassel Tue 3 Nov 2015

This is really off-topic, but I think that 'Block' is a less appropriate descriptor than 'Stop' for Loomio's general use-cases, which in my experience rarely involve formal consensus decision. This detail is the only reason that I'm not 100% smitten with your awesome comment above @strypey . :)

JK(

Joop Kiefte (LaPingvino) Tue 3 Nov 2015

I would recommend to make the vote more specific...

PLG

Purple Library Guy Wed 4 Nov 2015

I'm currently, ah, undecided. I don't mind the "clutter" as such. The one worry I have about this is that if the whitespace from "looked but haven't (yet) registered an opinion" tended to occupy a big chunk of the pie, it would make it less intuitive to judge roughly how much support or opposition a proposal had, what with both support and opposition being squinched over into a small part of the circle.

TT

Theodore Taptiklis Wed 4 Nov 2015

@purplelibraryguy What a great observation! So maybe the geography of the 'white space' (if this is technically possible) would need to be altered. Maybe "looked but haven't registered an opinion" is not actually the same kind of thing as 'agree/disagree/abstain/block" and should be represented outside the pie or in a different way somehow.

FB

Fabio Balli Wed 4 Nov 2015

Here is the new proposal which I refocused on seeing the participating. For changing the wording, please refer to the new thread : https://www.loomio.org/d/S3I0HTq5/abstention-part-2-make-wording-in-votes-more-congruent

DS

Danyl Strype Fri 6 Nov 2015

I like Fabio's visualisation, although I would swap the wording:
* orange remains 'abstain' - formally bow to the will of the group on this decision
* white is 'undecided' - commented (thus interested) but no position taken yet, not even to abstain

DS

Danyl Strype Wed 4 Nov 2015

I'd really like to, and I know its easier for coders to discuss implementation once they see how the suggested feature could look in the interface, but I'm juggling some deadlines at the moment and I really don't have time to wrestle with graphics software (I'm a total noob at graphics). Is there anyone else who thinks they grok what I'm suggesting, and could have a go at visualizing it?

JK

James Kiesel Wed 4 Nov 2015

The thing I like most about all of this is the idea of tracking views of proposals, which would give coordinators / group members an additional layer of engagement to consider (and it's, of course, a thing we already do when calculating what thread activity you've read.)

N

naught101 Wed 4 Nov 2015

What about having one large pie (the current one) for the vote results (agree, disagree, abstain, block, undecided), and a separate graph (pie/stacked bar) for participation (voted, viewed but didn’t vote, part of group, but haven’t visited).

Maybe something like this:

KL

Ken LeFebvre Thu 5 Nov 2015

I really like this idea!

LF

Luke Flegg Wed 11 Nov 2015

How about offering this as a feature to people who want it. Like me + thousands of people.
Then people who don't like it don't have to use it & everyone is happy?

GC

Greg Cassel Thu 5 Nov 2015

Another general possibility is to scale the size of the pie based on the percentage of position-registration. I've considered this for other types of sociometrics. I generally recommend a logarithmically scaled visual representation, instead of a linear scale.

This is a somewhat thorny concept, which I have not yet explored exhaustively, but it seems relevant here.

DU

[deactivated account] Thu 5 Nov 2015

Hmm perhaps the viewing/commenting slice should sit vertically symmetrical at the top of the pie, while block, disagree, abstain, agree slices would be tracked clockwise around the pie in the remaining space

DU

[deactivated account] Thu 5 Nov 2015

That way we can visually see the "balance" - the "gravity" of the votes

DS

Danyl Strype Fri 6 Nov 2015

I agree with @hannahsalmon that a large white space could make it harder to see the first few positions taken, but what would be gained is to see at a glance that most members engaged enough to comment are still undecided. I worry that with the current setup, where the first person to take a position fills the whole pie with their position, this may give a disproportionate impression of how much support that position has, and create an unintended network effect dragging the group in that direction. To mitigate this, I usually don't take a position on proposals I've created until at least a few other people have.

JW

Jochen Walter Fri 13 Nov 2015

How about offering this kind of decision options to your group/s. Telling, that not taking part of the proposals will mean undecided and the abstain means abstain. Then You probably need to make an effort that decisions for proposals need at minimum 50 % of voices otherwise it is not decided. Therefore You don't need a change of the voting system.
What do You think?

HK

Harry Knight Thu 26 Nov 2015

@gregorycassel I wonder if the white could represent 'understanding' as opposed to just 'viewed it'. I like to gain understanding of a topic by reading a thread or researching and sometimes I just want to understand it. I don't feel the need necessarily to contribute because i don't feel strongly pulled either way. But it shows i've engaged with the discussion on a critical level, maybe later i will come back to it after i've decided. I find during meetings I spend a lot of time listening and trying to gain deeper understanding of what is being talked about. Often meetings are too noisy.

GC

Greg Cassel Thu 26 Nov 2015

Thanks for describing your perspective, @wikihouseharry ! I have some similar experiences.

How could Loomio help participants to specify 'understanding' instead of just 'having viewed' a topic? RIght now, I only perceive that happening with an additional, actively chosen position/'voting' option. I'm not strongly opposed to that, but I agree with Loomio developers that a four-position system-- reflecting the history of formal consensus-- is more accessible and generally helpful.

I've always found it reasonably easy to explain my abstentions when I register my position. It could be helpful for Loomio to add some suggestions along these lines to its Help materials, and perhaps to highlight those ideas with a blog post on "the art of abstention". Actually I wouldn't be surprised if @richarddbartlett has already written something like that. (Sorry Rich I don't have time to mine your wealth of helpful writings right now. :) )

GC

Greg Cassel Thu 26 Nov 2015

Now I'm going to deliberately muddy the waters a bit, or maybe it actually adds clarity from some perspectives.

I think that assessment/ratings/feedback systems can often be improved by allowing users to expand their options when desired. For instance, in theory, a four-point position system could allow people to simply select "Abstain", but to also click something like "Advanced Options" or "It's Not that Simple", and then pick from a submenu.

I'm reluctant to point this out because I'm sure that Loomio devs have their hands full, and poorly chosen or presented submenu options could practically defeat the purpose of consensus-oriented process.

In the long-long-term, I reckon that some carefully chosen (or community-generated) submenu options would be useful. In the short term, however, I think there are much higher priorities:

  • education on consensus-oriented decision process
  • education on the advantages of online asynchronous decision process (as opposed to traditional meetings and votes)

With such urgent priorities, I think that advanced options would perhaps mainly add noise and confusion.

LF

Luke Flegg Thu 10 Dec 2015

www.represent.cc
and
www.vocaleyes.org
are perhaps useful case studies here?

AW

Aaron Wolf Thu 10 Dec 2015

Those examples from Represent.cc and VocalEyes from @lukeflegg are superb. In fact, although this is tangential, their use of score voting is key. The providing of extra details for abstention is nice too, but score voting is really nice, much better resolution. Consider http://scorevoting.net for the complete compelling thing about score voting.

My own primary reason that I hesitate to really use and promote Loomio is specifically because I want to use a system that supports that sort of score voting.

GC

Greg Cassel Thu 10 Dec 2015

I agree strongly that score voting is a key form of sociometric. That's a key focus for me personally. On the other hand, I'm far from convinced that it's more appropriate for collective decision-making than the agree/abstain/disagree/block protocol which many groups have used in formal consensus process for decades.

My perception is that gradient metrics are preferable for general discussion,and for some sorts of organically 'soft decision making'-- but Loomio style seems better for those rare but important times that people need to collectively make key decisions with long-term consequences.

AW

Aaron Wolf Thu 10 Dec 2015

@gregorycassel We actually agree here. I think there are different cases: 1. facilitated consensus, 2. fallback voting; and there's the distinct case 3. general take the temperature sort of thing.

For both temperature and fallback voting (and also any form of elections for representatives!) score voting is essential. For the consensus process (which is always preferable to a vote), it makes sense to just have a yes or no on whether everyone consents (but I don't think marking the numbers or letting a majority win is healthy when the point is checking for consensus). Finally, a hard block option should be available for any vote separate from the score vote; but the block needs to have a justification / reconciliation process.

Final note: score voting when the question is a yes/no type of decision (as opposed to an election with candidates) can require a higher bar than just average-approval. One option would be requiring an average of 4 on a 5-point scale, for example (or 3.5 or it could be a 6-point scale with no neutral score, abstention has no score count etc). That is actually a very high bar because it requires strong approval from one voter just to balance a neutral vote from another, and even weak disapproval can kill the measure. So, score voting here can be compatible with keeping high bar for long-term decisions.

PLG

Purple Library Guy Thu 10 Dec 2015

I had a quick look over at the score voting link Mr. Wolf gave. I don't really understand what's so great about it, especially for the kind of electoral contests it envisions. It says it avoids various problems of existing systems, but it seems to me that in reality it would rapidly degenerate.
If I'm voting in a scoring system, and I know that the winner is the candidate with the highest average, I have an incentive to give my favoured candidate perfect marks and everyone else zero (even if I kind of like them) so as to maximize the difference in average that I'm handing my favoured candidate. Doing anything else gives me, in effect, less than a full vote. Immediately it's in effect not a scoring system any more, just a vote. I suppose if I'm undecided it at least gives me the ability to in effect vote for two people, giving both of them top and everyone else zero. So there's a slight flavour of "affirmation ballot" to it.
Even in a question about "what to do", I think there would be a tendency for people to become partisan and score things more extremely than their real beliefs. I see it when people rate movies or manga or whatever and you see an aggregate score--people will give something a "5/5" just because its current score seems low to them, even though there's nothing at stake whatsoever.

The thing that scoring does bring in, to my mind, is the simple ability to compare multiple options. One thing that I've long believed and have noticed in this very thread is that consensus voting, traditional "Robert's Rules" style voting, and typical "proposition" voting share a fundamental weakness: They are all basically about evaluating a single proposal on a "take it or leave it" basis. People have various different ideas, often closely related but still distinct, and all those systems do not have an effective built-in mechanism for choosing between such multiple related options to find the one people like the most. Arguably the closest to having such a thing is actually the old fashioned "Robert's Rules" meeting with its amendments--but the process sucks. People working with consensus-based processes, such as here in Loomio, typically make a conscious effort to incorporate different people's contributions. You can see it operating in the thread. But there's no actual process for it; there's no way of telling how much support different suggestions have relative to each other, no real way to tell what is just an articulate individual and what represents a modification a lot of people want. In this case, it seems like the so far the end result was the primary proposal (and a couple of ad hoc modified versions) ended up with insufficient support to result in anything, while other ideas sort of hung around in limbo. And I've seen this happen a few times, it seems like a common result.

In other cases with more urgency, I can see situations in which something like the first proposal made shapes the discussion, implicitly ruling out other approaches because something must be done, and this is something.

I might argue for a process which went something like this:
--Problem or issue is brought up, with a solution or action suggested (or even more than one).

--People suggest other possible solutions. Interface allows you to look at the issue under discussion and a list of the proposals to deal with it, and signal your opinion in some way, whether with positives and negatives and abstinence and blocking, or with ranking, or what have you. Perhaps with scoring, although as I say I don't really get that.
--Loomio tracks, maybe with pie charts and such, relative positive votes for different ones
--some attempt to integrate related proposals
--Whichever proposal is "winning" (and optionally not-blocked) after a length of time then moves on to consensus process.

This would have the disadvantage of being slower, but I think it would broaden the alternatives that could be seriously considered and would have a better chance of reaching a conclusion, so might actually be more efficient in the long run. You could be pretty sure by the time you sought consensus on something that it was generally acknowledged to be pretty broadly supported.

AW

Aaron Wolf Thu 10 Dec 2015

@purplelibraryguy I don't want to derail this thread, just note that strategic voting with score voting is mild degeneration that leaves a result that is still the best form of voting. More details: http://www.rangevoting.org/HonStrat.html

Not everyone is strategic, and if you choose to downvote all your other options to help your favorite, that can be risky because you have just lost your vote entirely if your favorite doesn't win. Most people will find that they would rather be honest so that even if their favorite doesn't win, they'd rather have influence on who does win than to throw away their vote. Besides that, in reality many elections people don't have a strong favorite at all. Score voting allows people flexibility and the worst case of strategic voting isn't even bad anyway.

Yes, "approval voting" is very connected to score voting, it's score voting with lowest resolution of only two score options. It has most of the benefits of score voting except the better resolution and flexibility.

Anyway, I agree with you overall: the best form of consensus is one that actively works to incorporate everyone's concerns and not just go with a lowest-common-denominator but actually work for win-win solutions that include everyone's issues. In that process, the purpose of a score system is merely informational, not decisive, and most people in such a consensus-driven situation will be honest. It's possible to be strategic in any context though, such as not consenting to something you actually are okay with just because you think that failed consent will make others compromise and move to your favored view. That's why we need both good facilitation and fallback voting (so that consensus-blocking can't be tyrannical and impossible to get past).

GC

Greg Cassel Fri 11 Dec 2015

I think you @purplelibraryguy have a valid point when range voting is used strictly for selection of 'one from many'; and I see other ways to reduce the influence of strategic voters, but this goes beyond my interest here (this thread) in Loomio's consensus-oriented decision process.