Search All of the Math Forum:
Views expressed in these public forums are not endorsed by
Drexel University or The Math Forum.
|
|
|
|
What's the number for weird?
Posted:
Nov 13, 2011 3:25 AM
|
|
|
|
I'm thinking the class needs to grapple with the questions below. I've attached a worksheet and an example of the TinkerPlots document I think they will end up using. Any thoughts?
Questions to grapple with:
1) What happens to the chi-square statistic as the number of trials (?repeats? in TinkerPlots) changes?
2) Why do we have to add the observed values when computing the chi-square statistic?
3) What?s the likelihood of getting different chi-square values assuming their is no bias (that the null-hypothesis is true)?
4) At what point does it feel like we can rule out the likelihood of their being no bias, that we would feel compelled to reject the null hypothesis? In other words, for what value of chi-square is the point at which you say ??liar??
5) What sort of chi-square values would I get if their was a bias in the situation; if the situation were unfair?
6) Can you tell by looking at data if a situation is unfair?
|
|
|
|
