Stats questions
So, I just had a killer final, and I'll be glad if anyone can help me answer some questions I have about hypothesis testing.
1)
Say I have 2 classes, and I have a sample from one of them, Class A. I want to find out if the average number of class units the students are taking in Class A are different from another class, Class B, or whether it's just chance error. Note that I DON'T have the sample of Class B, so I can't do a 2-sample z-test. essentially, I have sample data from Class A, and only the population data for Class B.
I know it's possible to compare Class A sample to Class A population using z-test, but is it possible to compare Class A sample to Class B population and conclude whether or not the difference in data is due to chance error?
2) A survey is conducted, and basically the test is to see whether the wording of the survey affects the responses, or is the difference in responses due to chance error. So, for Survey A, 500 were sampled and 49% answered Yes, while 51% answered No. For Survey B (where questions are worded differently), 520 people were sampled, and 42% answered Yes, 58% answered No.
I used a chi-square test for this, but my friends said it should be 2-sample z-test. So....which is it?
Thanks!!
1)
Say I have 2 classes, and I have a sample from one of them, Class A. I want to find out if the average number of class units the students are taking in Class A are different from another class, Class B, or whether it's just chance error. Note that I DON'T have the sample of Class B, so I can't do a 2-sample z-test. essentially, I have sample data from Class A, and only the population data for Class B.
I know it's possible to compare Class A sample to Class A population using z-test, but is it possible to compare Class A sample to Class B population and conclude whether or not the difference in data is due to chance error?
2) A survey is conducted, and basically the test is to see whether the wording of the survey affects the responses, or is the difference in responses due to chance error. So, for Survey A, 500 were sampled and 49% answered Yes, while 51% answered No. For Survey B (where questions are worded differently), 520 people were sampled, and 42% answered Yes, 58% answered No.
I used a chi-square test for this, but my friends said it should be 2-sample z-test. So....which is it?
Thanks!!