help interpreting SPSS readout

[George_Foreman]

if you guys could help me understand what all this stuff i just made SPSS spit out means, i'd really appreciate it. it's part of a project i'm doing analyzing the affectiveness of pre-season projections from different sources.



many thanks!

[Oat Soda]

I have no way of figuring out what you're talking about from that screen shot. Do you have a link to the actual page so we can scroll around to see what you're talking about?

[George_Foreman]

the page is on my computer... so no, no link.

i'm just looking for help with what the different categories mean, not the actual data outputs.

for instance, i know that an f-score has something to do with the distirbution of the data, and that a lower f-score means that the model more accurately predicts the dependent variable, but i don't know precicely what it means or how to interpret a generic f statistic. i know that the category "sig." means "significance", and i assume that it means something along the lines of "significance of the independent variable with regards to predicting the dependent variable", but again, i'm not sure specifically what it refers to. for instance, what does a positive "sig." mean?

hope this helped to clairfy. thanks for the response.

[LBJackal]

Just search the net for an FAQ on the program you're using. That's what I did for the Excel extension I was using to run regressions in order to make my pitcing formula. There should be something showing step-by-step what everything means.

[matthew18]

The sig should range from .000 to .999...........the lower the sig score the better. If a sig score is too high (number depends on how exact you are trying to be), then you can just throw the variable out of the equation because it does nothing to predict the dependent variable.

[DCNats]

Typically, though it depends entirely on the situation, you want your f-prob (overall significance of the model) and your p-values (significance of the individual independent variables) to be less than .05.

But you shouldn't throw variables out of the equation just because they are not significant, because excluding them can introduce bias. Think about the theory behind the variables (should they influence the dependent variable) when deciding whether to include them or not.