pull down to refresh

Take the Perception Gap Quiz, Let SN Know Your Score!perceptiongap.us/
618 sats \ 13 comments \ @SimpleStacker 8 Oct Politics_And_Law
Following this conversation, #714785 with @kepford, I thought it would be fun for Stackers to see how their political perceptions about the opposing side stack up to reality. This gap is called the Perception Gap, and in my opinion explains the political paralysis the U.S.A., (and likely many other countries), are experiencing.
You can take a quiz to measure your own perception gap here: https://perceptiongap.us/
I'm proud to say I got a gap of only 6%.
write
preview
related posts
47 sats \ 2 replies \ @Scroogey 8 Oct
As a foreigner, I picked Independent and got a score of -1% for Democrat and -20% for Republican.
Why is a large underestimate plus an equally large overestimate the same as two perfect estimates? I.e. shouldn't it sum the absolute errors?
reply
43 sats \ 0 replies \ @SimpleStacker OP 8 Oct
I think a -20% means that you think Republicans are more moderate than they actually are.
Their choice of how to present positive/negative is actually pretty confusing and I wish they had chosen a more intuitive approach.
reply
1 sat \ 0 replies \ @Undisciplined 8 Oct
I think it's trying to measure how extremely partisan you view each side, so averaging out an overly extreme perception with an underly extreme perception makes sense.
Using absolute values would give an interesting measure of how precisely you understand each side.
reply
20 sats \ 1 reply \ @siggy47 8 Oct
Cool. As an independent I thought Dems were a little more extreme in reality than I did Reps.
reply
43 sats \ 0 replies \ @kepford 8 Oct
It is interesting. I mean we are trusting their data but since I scored fairly well I guess the data they have aligns with what I think...
Kinda funny how these things work.
reply
14 sats \ 2 replies \ @Undisciplined 8 Oct
That was neat.
Both of mine were negative: -1% for Democrats and -7% for Republicans.
reply
59 sats \ 1 reply \ @SimpleStacker OP 8 Oct
I scored the most poorly on the gun control question. I thought more democrats would support banning guns even for law abiding citizens.
reply
0 sats \ 0 replies \ @Undisciplined 8 Oct
I nailed that one, but was way off on how many Republicans want more gun control.
reply
31 sats \ 0 replies \ @Coinsreporter 8 Oct
reply
10 sats \ 0 replies \ @DarthCoin 8 Oct
only slaves vote
reply
2 sats \ 0 replies \ @3c12153c35 8 Oct freebie
-1% for Democrats and -7% for Republicans, non US citizen.
0 sats \ 0 replies \ @kepford 8 Oct
I kinda hurried my test (over VPN btw) so I think I could have done better but I was -1% on D and -6% on R.
I really think the perception of the two sides even by those that are not on one of the sides is off. Primary due to the fact we overweight what we hear and see online and in media. I used to fall into this trap more than I do now. Once I became more aware of how media amplifies things and distorts things and started to focus more on the people I know and what they actually do and say I started to worry less and get far less concerned about things.
We are still new to this Internet thing. Prior to the Internet I just don't think we were exposed to so many people and how nuts they can be. But on the whole people aren't all that extreme.
reply
0 sats \ 0 replies \ @Shillohpopi_ 12 Oct freebie
Your question seems to be more about statistical analysis and error calculation rather than the specifics of your Independent, Democrat, or Republican party affiliations.
In statistical terms, a large underestimate and an equally large overestimate can cancel each other out when calculating the mean or average error. This is because they have opposite signs, one positive and one negative. However, when considering absolute errors, the two large errors wouldn't cancel out.
Error Calculation Methods
  • Mean Absolute Error (MAE) considers the average magnitude of errors, without regard to direction. This method would sum the absolute values of errors ¹.
  • Mean Squared Error (MSE) and Root Mean Squared Error (RMSE) consider both the magnitude and direction of errors. These methods would treat underestimates and overestimates differently.
Why Average Errors Cancel Out
When calculating the average error, underestimates (negative errors) and overestimates (positive errors) can balance each other. For instance:
True value: 10
Estimate 1: 8 (underestimate, error = -2) Estimate 2: 12 (overestimate, error = +2) Average error: (-2 + 2) / 2 = 0
In this case, the average error is zero, suggesting no overall bias.
However, using absolute errors or squared errors would provide a different picture:
Mean Absolute Error: (|-2| + |2|) / 2 = 2 Mean Squared Error: ((-2)^2 + 2^2) / 2 = 4
These metrics highlight the magnitude of errors rather than cancelling them out