[EM] Summary of psych/market-research studies of rating scales

[Abd ul-Rahman Lomax]

At 12:24 AM 6/4/2013, Richard Fobes wrote:
>While reading the information about score ballots, I wondered what 
>the range-voting advocate's response is to the belief that a big 
>preference gap in one ballot will have more influence than a smaller 
>preference gap in another ballot.

So, Richard wants to know what one believer will say about the belief 
of another believer?

What's the belief here?

A big preference gap on one ballot will have more influence than a 
small preference gap on *any* ballot. That's not a "belief," that's 
Score voting!

Range Voting is Approval Voting with fractional votes allowed, it's 
seriously that simple.

Thus the ultimate characteristic is that a larger fractional vote has 
more influence than a smaller fractional vote, and, in the extreme, 
that a full vote has more influence than no vote. This applies to all 
candidate pairs.

>For example, suppose one voter votes:
>
>A = 1
>B = 2
>C = 10
>
>and another voter votes:
>
>A = 1
>B = 5
>C = 10
>
>and, combined with the other ballots, the winner is C.
>
>Now, suppose the first voter changes hisher ballot to:
>
>A = 1
>B = 5
>C = 10
>
>and now B wins.

I don't like ranges that don't have a zero. Is that first vote 
one-tenth vote *for* A -- in which case it *might* cause a tie for A 
-- or is it *no* vote for A?

I'm going to assume that the first vote is actually a zero, not "1." 
It makes the matter clearer.

And I apologize for writing "Range" instead of "Score." The name of 
"Score voting" was a political decision that not only attempted to 
obsolete many usages in notable publications, it also shaded into the 
whole set of misconceptions behind "grading" systems, which take us 
away from a clear understanding of voting as a process of *choice*, 
not of absolute rating.

Above, the voter changed their vote for B from 2/10 vote to 5/10 
vote. They increased their vote for B. And this, indeed, could cause B to win.

>This implies that the big gap between B and C in the first ballot 
>has more influence than the smaller gap between B and C in the second ballot.

I.e., a larger fractional vote has more influence than a smaller 
fractional vote, just as a full vote has more influence than no vote 
(or 0/10 vote).

>How do range voting advocates resolve this apparent unfairness?

What unfairness?

*That* is a belief, or, perhaps more accurately, an "occurring."

What is unfair about it?

>I'm asking out of curiosity.
>
>("Curiosity killed the cat, but satisfaction brought him back.")

2/10 killed B, but 5/10 brought him back.