What is 'Significance F' and what is its forumula?

#1
My prof is giving us an ANOVA table to fill in with limited data and he pointed to significance f. I thought it was nothing or could find it on my own. I can't. Any suggestions?

Also, how can I find confidence intervals without having access to excel?

This is a summer stats 2 class. Thanks.:wave:
 

bugman

Super Moderator
#6
Ok, so the f-ratio is as I mentioned before.

The significance F is your P-value (the p-value is the probability of obtaining a test statistic at least as extreme as the one that was actually observed, assuming that the null hypothesis is true) - in the case of regression: Ho: B1= 0
 
#7
Is there a way I can make the formulas visible in this worksheet? I'd like to get comfortable with how excel is doing the calculations.

Thanks
 

Dason

Ambassador to the humans
#9
Now I'm using my mac right now using Numbers 09 as opposed to Excel but it looks like the regression output was hard coded in. The summaries in the data seemed to be from formulas inside the spreadsheet but if the regression output was hard coded I don't think there's a way to recover the formulas that whoever made the spreadsheet used. I mean we could tell you what the formulas are in general (actually we'd probably just point you to a website that has the formulas). Now this might be wrong since I'm not using Excel but Numbers picks up the formulas usually.
 
#10
Yes it came from mystatlab that my prof just loves using. Not really learning much from it b/c he doesn't teach it. Which I guess he should b/c I don't think businesses use the pencil paper method much. I wish minitab was taught.
 

Dragan

Super Moderator
#11
Ok, so the f-ratio is as I mentioned before.

The significance F is your P-value (the p-value is the probability of obtaining a test statistic at least as extreme as the one that was actually observed, assuming that the null hypothesis is true) - in the case of regression: Ho: B1= 0
Hi Bugman :): The F-test, in this case, is testing the (null) hypothesis:

Ho: B1=B2=0.

The t statistics associated with X1 and X2 are separately testing the unique contribution (i.e. the squared semi-partial correlation) of each independent variable i.e. Ho1: B1=0 and Ho2: B2=0
 

Dason

Ambassador to the humans
#14
Do you mean is there a formula to calculate a p-value from an F statistic? Do you mean how do you calculate the F statistic? Are you wondering how to calculate the critical value for an F distribution? Can you elaborate on what you mean because what you're asking doesn't really make sense to me.
 

Dason

Ambassador to the humans
#16
Oh ok. Looks like you were looking for cdf values for the F distribution and you're using some sort of stats package (not sure which one). You might want to mention these things in the future to get help a little quicker.