Most of the examples in the CFAI text regarding Regression analysis use a large number for ‘df’ in the question. Do i need to expect that on the exam or there won’t be such thing. Thank you

What difference does it make if it’s a large or small number of df ?

Wouldn’t df be significant for rejecting or failing to reject the null hypothesis?

anishcandy Wrote: ------------------------------------------------------- > Wouldn’t df be significant for rejecting or > failing to reject the null hypothesis? it’s a relevant contributing factor but it’s not a sole determinant. you compare your test value with the critical value and make a decision depending on the type of test. you’d have to look at the df to pick the correct critical value from the chart.

Keep in mind that that DF contributes to whether the sample (assuming a random unbiased selection of candidates) represents the population. When running a statistical test you will be using the DF all the time, regardless of whether it’s an F-test or T-Test. The only other wrinkle to keep in mind is if the sample is over 30 (once agains assuming a random unbiased selection of candidates from the pop.) it’s going to approximate the pop. and you can use the Z-test.

thank you guys.