Question

Which of the following is NOT a possible alternative in two mean testing for independent populations?

	Standard deviations are known.
	Standard deviations are equal to each other.
	Standard deviations are not known.
	Standard deviations are irrelevant, because the samples are very large.

I got D and

Which is NOT a true statement about F-distribution?

I got C

	F close to 1 might indicate that normal deviations are equal.
	F can be equal to 0.5.
	F equal to a significantly large number might mean that the standard deviations are not equal.
	F-distribution curve depends on both df1 and df2.

please check and explain your answers and tell me why if I am wrong or right (im just not 100% sure and would rather understand these concepts )

Answer 1

1. Standard deviations are irrelevant, because the samples are very large. ( The test statistic value for two sample t test includes standard deviation in it. The sample size being large would not change the contribution of including standard deviation in the formula. Though the samples are large, variability may be less in them and variability will surely exist. So they cannot be irrelevant)

2. F equal to a significantly large number might mean that the standard deviations are not equal. (F distribution is used to measure the probability for change in sample size or confidence. F value is greater than the corresponding pvalue mean that standard deviation are not equal.)