Thanks Nathan! People seem to like it so we might use it again in the future. If you or anyone else has feedback that might improve the next iteration of it, please let us know! You can comment here or just dm.
But I think thereâs work to do on the display of the aggregate.
I imagine there should probably be a table somewhere at least (a list of each person and what they say).
This might show a distribution, above.
There must be some way to just not have the icons overlap with each other like this. Like, use a second dimension, just to list them. Maybe use a wheat plot? I think strip plots and swarm plots could also be options.
Really appreciate all the feedback and suggestions! This is definitely more votes than we expected. đ
I implemented a hover-over based on @Agnes Stenlundâs designs in this PR, though our deployment is currently blocked (by something unrelated), so Iâm not sure how long it will take to make it to the live site.
I may not have time to make further changes to the poll results UI this week, but please keep the comments comingâif we decide to run another debate or poll event, then we will iterate on the UI and take your feedback into account.
The orange line above the circles makes it look like thereâs a similar number of people at the extreme left and the extreme right, which doesnât seem to be the case
I donât think it would help much for this question, but I could imagine using this feature for future questions in which the ability to answer anonymously would be important. (One might limit this to users with a certain amount of karma to prevent brigading.)
I note some of my confusion that might have been shared by others. I initially had thought that the option from users was between binary âagreeâ and âdisagreeâ and thought the method by which a user could choose was by dragging to one side or another. I see now that this would signify maximal agreement/âdisagreement, although maybe users like me might have done so in error. Perhaps something that could indicate this more clearly would be helpful to others.
Thanks Brad, I didnât foresee that! (Agree react Bradâs comment if you experienced the same thing). Would it have helped if we had marked increments along the slider? Like the below but prettier? (our designer is on holiday)
Yeah, if there were markers like âneutralâ, âslightly agreeâ, âmoderately agreeâ, âstrongly agreeâ, etc. that might make it clearer.
After the decision by the user registers, a visual display that states something like âyouâve indicated that you strongly agree with the statement X. Redrag if this does not reflect your view or if something changes your mind and check out where the rest of the community falls on this question by clicking here.â
Thanks Nathan! People seem to like it so we might use it again in the future. If you or anyone else has feedback that might improve the next iteration of it, please let us know! You can comment here or just dm.
I think itâs neat!
But I think thereâs work to do on the display of the aggregate.
I imagine there should probably be a table somewhere at least (a list of each person and what they say).
This might show a distribution, above.
There must be some way to just not have the icons overlap with each other like this. Like, use a second dimension, just to list them. Maybe use a wheat plot? I think strip plots and swarm plots could also be options.
Iâm excited that we exceeded our goals enough to have the issue :)
I would personally go for a beeswarm plot.
But even just adding some random y and some transparency seems to improve things
document.querySelectorAll(â.ForumEventPoll-userVoteâ).forEach(e â e.style.top = `${Math.random()*100-50}px`);
document.querySelectorAll(â.ForumEventPoll-userVoteâ).forEach(e â e.style.opacity = `0.7`);
Really appreciate all the feedback and suggestions! This is definitely more votes than we expected. đ
I implemented a hover-over based on @Agnes Stenlundâs designs in this PR, though our deployment is currently blocked (by something unrelated), so Iâm not sure how long it will take to make it to the live site.
I may not have time to make further changes to the poll results UI this week, but please keep the comments comingâif we decide to run another debate or poll event, then we will iterate on the UI and take your feedback into account.
Looks great!
I tried to make it into a beeswarm, and while IMHO it does look nice it also needs a bunch more vertical space (and/âor smaller circles)
Also adding a little force works too, eg here. There are pretty easy libraries for this.
The orange line above the circles makes it look like thereâs a similar number of people at the extreme left and the extreme right, which doesnât seem to be the case
I donât think it would help much for this question, but I could imagine using this feature for future questions in which the ability to answer anonymously would be important. (One might limit this to users with a certain amount of karma to prevent brigading.)
I note some of my confusion that might have been shared by others. I initially had thought that the option from users was between binary âagreeâ and âdisagreeâ and thought the method by which a user could choose was by dragging to one side or another. I see now that this would signify maximal agreement/âdisagreement, although maybe users like me might have done so in error. Perhaps something that could indicate this more clearly would be helpful to others.
Thanks Brad, I didnât foresee that! (Agree react Bradâs comment if you experienced the same thing).
Would it have helped if we had marked increments along the slider? Like the below but prettier? (our designer is on holiday)
Yeah, if there were markers like âneutralâ, âslightly agreeâ, âmoderately agreeâ, âstrongly agreeâ, etc. that might make it clearer.
After the decision by the user registers, a visual display that states something like âyouâve indicated that you strongly agree with the statement X. Redrag if this does not reflect your view or if something changes your mind and check out where the rest of the community falls on this question by clicking here.â
Another idea could be to ask, âHow many EA resources should go do this, per year, for the next 10 years?â
Options could be things like,
â$0âł, â$100kâ, â1Mâ, â100Mâ, etc.
Also, maybe there could be a second question for, âHow sure are you about this?â
Interesting. Certainty could also be a Y-axis, but I think that trades off against simplicity for a banner.