jaycee
Well-Known Member
The problem with all of this is that when we are talking about human behavior, or predicting human behavior based on past behaviors, this is all a function of probability. Scientists aren't more okay with the maybe. They just understand that the math and model is only a suggestion of reality based on the limited data we have. The models are supposed to show us how likely a given answer is if someone has the choice between yes or no, rep or dem, but they also account for random guessing and error which is why we have confidence intervals, and to me, this is more accurate. But I understand the math behind all of it.
I agree with some of this and disagree with some of this. I think what "maybe" means to you depends on what your discipline is. Variability is a description of the maybe... of the range of possibilities... and models (again discipline specific) are tools that use the past to provide reasonable guesses at the future. They are never answers. Ideally they provide reasonable information based upon sound data, but they can't tell you what's going to happen. Only what is more or less likely.
I don't fully understand modeling based on political polls, but my guess is that even with significantly large enough sample sizes there is greater "maybe" because there is a human factor involved and we are inherently (seemingly) unreliable beasts.
Regardless of any of that my thought was more that people see polling information / models, often they see this or that %, or this or that # of respondents and they don't have the knowledge to understand the fluff around those numbers and that they are not an answer (just information).
Last edited: