I haven’t used Polis before, but from looking at the report, I believe that Groups A and B (I don’t seem to see C) are automatically generated based on people with similar sentiments (voting similarly on groups of statements). It seems like Group A unconcerned about money and thinks more of it is good (for example, only 31% agree with “It’s not that more money is bad, but is a bit unnerving that there seems to be so much more of it” compared with 88% in group B), and Group B is concerned about money, which includes everything from a concern around a small number of funders controlling the distribution of funding in EA to believing that personal sacrifice is an important part of EA (perhaps in contrast to wanting to offer people high salaries).
Typical caveats of it being a small and potentially unrepresentative sample size, but if I’m interpreting the data correctly, I think it’s interesting that a lot of people seem to be concerned about money in general, rather than specific things. Perhaps there actually are more granular differences in opinion and there could be a way to highlight those more nuanced shared perspectives if the number of sentiment groups can be changed. Not sure what happened to Group C depicted in the screenshot.
At some times I looked at the report there were two groups, at other there were three.
What I meant about interpretation is, which opinions are correlated and why? How would the different groups be characterized in general terms? Is there some demographic/professional/etc. axis that separates them? (Obviously we can only guess about the last one).
I think I already shared this comment on facebook with you, but here I am re-upping my complaints about polis and feature suggestions for ways that it would be cool to explore groups and axes in more detail: https://github.com/compdemocracy/polis/discussions/1368 (UPDATE: I got a really great, prompt response from the developers of Polis. Turns out I was misinterpreting how to read their “bullseye graph”, and Polis actually provides a lot more info than I thought for understanding the two primary axes of disagreement.)
If you have your own thoughts after conducting various experiments with polis (I thought this was interesting and I liked seeing the different EA responses), perhaps you too should ping the developers with your ideas!
Currently the visualisation on the pol.is questions looks like this. Anyone feel free to write a summary.
And the full report is here:
https://pol.is/report/r8rhhdmzemraiunhbmnmt
This is interesting but I have no idea how to interpret it.
I haven’t used Polis before, but from looking at the report, I believe that Groups A and B (I don’t seem to see C) are automatically generated based on people with similar sentiments (voting similarly on groups of statements). It seems like Group A unconcerned about money and thinks more of it is good (for example, only 31% agree with “It’s not that more money is bad, but is a bit unnerving that there seems to be so much more of it” compared with 88% in group B), and Group B is concerned about money, which includes everything from a concern around a small number of funders controlling the distribution of funding in EA to believing that personal sacrifice is an important part of EA (perhaps in contrast to wanting to offer people high salaries).
Typical caveats of it being a small and potentially unrepresentative sample size, but if I’m interpreting the data correctly, I think it’s interesting that a lot of people seem to be concerned about money in general, rather than specific things. Perhaps there actually are more granular differences in opinion and there could be a way to highlight those more nuanced shared perspectives if the number of sentiment groups can be changed. Not sure what happened to Group C depicted in the screenshot.
At some times I looked at the report there were two groups, at other there were three.
What I meant about interpretation is, which opinions are correlated and why? How would the different groups be characterized in general terms? Is there some demographic/professional/etc. axis that separates them? (Obviously we can only guess about the last one).
I think I already shared this comment on facebook with you, but here I am re-upping my complaints about polis and feature suggestions for ways that it would be cool to explore groups and axes in more detail: https://github.com/compdemocracy/polis/discussions/1368
(UPDATE: I got a really great, prompt response from the developers of Polis. Turns out I was misinterpreting how to read their “bullseye graph”, and Polis actually provides a lot more info than I thought for understanding the two primary axes of disagreement.)
If you have your own thoughts after conducting various experiments with polis (I thought this was interesting and I liked seeing the different EA responses), perhaps you too should ping the developers with your ideas!