Quote:
Originally Posted by wilsonmw04
So what this data tells me is that a small number of teams really hate this idea and are very vocal about it. This seems to jive with what typically happens here on CD.
|
More than a quarter of the respondents are as against this proposal as it is possible to register on this scale. That's closed to 2,000 people in itself (numerically 1912.3 people). 55% of respondents are against the proposal; that's more than 4045 people. CD is vocal, but even assuming the relationship between CD and the survey sample (which is a weird assumption when n=7355), opposition by definition is not the minority opinion. And despite the scale shift, the "strongly opposed" outnumbers all those who voted 10, 9, and 8
combined. More people voted for 1 or 2 than voted for
anything above 5.
Does anyone know if there's a standard method of "centering" a scale like this? (The true center is at 5.5, the average of 1 and 10). I don't have a statistical method of turning 4 buckets into 5, but I think the worst-case scenario would be that everyone who voted 1 would've voted 0, and everyone in 2 took 1 (no one votes 4). This creates a new weighted average of 3.92, which represents the low end of possibility: thus the average is somewhere between 3.92 and 4.47 when centered about 5. Did I handle that correctly?