Show HN: PolliticalScience – Anonymous daily polls with 24-hour windows

You're right on that. That is one reason I am avoiding scientific polling and calling it "public sentiment".

The binary, with all its flaws, is dead simple. I see a statement, do I agree or disagree with that statement. If I don't fully support one or the other, the discussion section I intended for them to make their case.

I am trying to keep the same similar format for each poll. X should Y or X was right to Y in hopes people get used to the wording. On the backend I have a bias rating I give each poll as well as a temperature. The bias is LeansLeft, Center, LeansRight. The temperature is Cool, Warm and Hot. For most warm to hot polls, the statement has to be worded in the affirmative towards one side or the other. I try to keep the wording bias somewhat balanced so I don't post too many in one direction.

I am playing around with the idea of having two steps though, Agree/Disagree (required) then an optional follow-up to rate where on the scale you fall. I want to keep the simplicity as the primary target, but it could be interesting to see where people fall on a range. It would still be anonymous but rather this vote was agreed and they leaned 75% towards that side or something.

Hey absolutely! I checked out your website and looks great, I guess our ideas are validating each other! I would definitely be interested in learning more about your choices and implementation on demographics/verification, as I started down that path before shying away.

I had originally considered using protected session storage in the browser for storing demographic information so it never got stored on my server and stayed with the user, but decided for now to go as simple as possible with two big buttons. My main concern was, even if the vote was anonymous (especially when starting up without a large user base), if I had the users demographics stored on their account and if there were not many votes from say their region, you could in theory narrow it down to who that person was (email, ethnicity, age, location). If you had one user who was from Wyoming with those sets of demographics and you had one vote from Wyoming, well... I could make an educated guess who that was who voted and what they voted. I may have been overthinking it, but I just didn't feel personally confident enough to move forward with that, at least at the time.

I do also really like how you give the users context during their voting (pros/cons). I went with context after voting to get the gut reaction then drive discussion based on context. Really interesting to see!

I see your point. My main goal with this was simplicity, not reductivism. I am not sure the average person wants to dive into the nuances of every poll just to participate (I do like it as an option though). I also don't want to exclude people because it is too complex or there are too many steps to contribute.

The binary choice can be polarizing. But it can also yield very interesting results. Since it is anonymous, you can vote exactly how you feel with zero repercussions. People might actually find they are not as alone as they think. It also might allow someone on one side to choose something on the other side they normally wouldn't in something more formal.

It is really hard to say at this point though as I don't have enough data to make real conclusions (and not sure I ever will with this type of anonymous voting). But I do find it interesting and there have been some pretty good discussions so far. People have been explaining their thoughts without any dissolution into personal attacks which is great.