Schweser Book 1 Page 201 - Detecting Multicollinearity

Ok…here…they have an example with the table showing the F-Test value (34.6). How do they conclude that the F-Test is statiscally significant, without any calculations. No Significance value or anything is mentioned. Only P-value of <0.001. Same goes for the Concept checkers question on page 211. Quest. no. 8. How are they able to conclude that the Breusch-Pagan and the F-Statistic are statistically significant, without doing any calculations? I may be missing something here. Thanks in advance.

In other words…how can you tell that a Breusch-Pagan or F-Test is statistically significant by just looking at the table?

Sparty, Without looking at the problem, the P-value tells you the minimum value at which the null can be rejected (or something like that). If you look up the definition of hte P-value it will help. It is probably because the P-value is known to give significance at even the lowest usually significance level, 1%.

First, you can’t really tell whether an F-statistic or chi-square statistic is statistically significant without either having a computer claulcate something or looking in a table. But F’s are supposed to be numbers near 1 if you don’t have significance and for any reasonable sample size an F of 34 is wildly significant. I, too, would just write down p <0.001 and go on. “Statistically significant” almost always means “P-value less than 0.05” (or maybe 0.01).

Given small p-value, reject Ho and conclude that results are statistically significant. Small p-value needs to be defined, but as Joey said, if its below 0.05 or 0.01 that is what you need (considering confidence level you are working with 95% or 99%), if its something like 0.001 then just apply this rule (this gives you confidence level of 99.9%).