F statistic Quant

[sunman9]

This might be a stupid question but:

if the F stat is 2.355 and the F significance in the ANOVA table is 0.1285 ---- this comparison is not helpful for assessing the explanatory power of the regression correct?

I know that we look at if F stat is significant and tstats arent, then good chance of multicollinearity. So in this case F > Fcrit? OR is this saying that 12.85% chance that when F > Fc when in fact it isnt signifcance - ie at the 5% signifincace level, we cant say this F stat is significant.

Am i looking at that correctly?

The CFAI quant results section made a statement that kinda threw me off.

[tickersu]

@sunman9

This might be a stupid question but:

if the F stat is 2.355 and the F significance in the ANOVA table is 0.1285 ---- this comparison is not helpful for assessing the explanatory power of the regression correct? Significance of F (or t) is a p-value. Compare it to alpha. Alpha > p-value implies reject Ho. It’s the same as comparing F(observed) to F(critical). The F-test in this case tells us that the model is not statistically useful (FTR H0 for all slopes = 0). It also tells us (since p-value> alpha) that F(observed) < F(critical), even though they didn’t give us the critical F-value.

I know that we look at if F stat is significant and tstats arent, then good chance of multicollinearity. So in this case F > Fcrit? OR is this saying that 12.85% chance that when F > Fc when in fact it isnt signifcance - ie at the 5% signifincace level, we cant say this F stat is significant. It says the p-value = 0.1285 and alpha = 0.05 – alpha < P-value implies Fail to reject (FTR) H0. The F-test is not significant. The 0.1285 tells us that there is about a 12.85% chance we see a test result at least as favorable to the alternative hypothesis, assuming H0 is true.

Am i looking at that correctly?

The CFAI quant results section made a statement that kinda threw me off.

[sunman9]

perfect. thanks a bunch

[tickersu]

Glad to help