Question

which choices listed below indicate that a linear model is not the best fit for a data set? select all that apply.
a) scatterplot shows a strong linear pattern
b) scatterplot shows a curve pattern
c) residual plot shows a curve pattern
d) residual plot shows no pattern
e) correlation coefficient is close to 1 or -1
f) coefficient of determination is close to 1 or -1
g) unexplained variation is close to 1 or -1

Answer 1

hhhhhhhhhhhhhhhhhhhhhiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwdwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwwww

Answer 2

@BTaylor

Answer 3

C is the best answer. If a residual shows a strong curve, then it cannot be a good fit for the data. Usually, that implies a power or log model.

Answer 4

is that the only answer that applies though? it says this question is worth 3 points :/

Answer 5

@BTaylor

Answer 6

B is another good answer. If the data forms a curve, a line won't approximate well.

Answer 7

would g be one too?

Answer 8

I think so.

Answer 9

I'm sure so. We don't want unexplained variation near 1, since then all the variation is unexplained.