what statistical test should I use?

Question

A. the data from 18 past kentucky derby is shown with the winning horse and its sire (father) and the mean pre-derby price of each sire's offspring and a mean post-derby price for offspring. Does siring of the winner of the kentucky derby change the mean offspring price from pre-derby to post-derby?

What statitical test would I run in SPSS?

B. What test would I run if I wanted to determine if there was a relationship between pre-derby and post-derby price

C. Use the pre-derby price to predict the post derby price (what test?)

I was thinking Pearson r and a linear regression maybe?

Answer 1

If you wanted to determine whether or not the pre/post derby price had changed, you would most probably use a related samples t-test.



To determine the strength of the relationship, you would use the Pearson correlation coefficient (which, surprisingly enough can be algebraically derived from the t test). or perhaps Spearman rho (rank order correlation.) If you go with rho, recomember you are predicting ranks, not values.



Using the correlation along with a regression function would allow you to accomplish the prediction you desire.



The above is probably far more conveniently handled with excel than SPSS.



always,

tony

Answer 2

A.

If the data are one-dimensional, you may want to use the t-test.



B,C)

I was thinking Pearson r and a linear regression maybe?

(you are correct; assume you are testing for linear relationship).

Answer 3

The null decision is the two varieties are an identical So the advise could be at 5% ingredient the advise is 5.2 the alternative speculation could be that the advise are distinctive so advise<5.2 or advise>5.2 could be decision. yet no you would be waiting to desire to think of of , that the breeder isn't unterested by potential of potential of the undesirable effect so indoors the priority you soak up account in straight forward words the 2d speculation