Dear Doug Ward,
Thanks for your elaborate answers. Now, I fully understand the df calculation itself.
As to your A2, ¡®This does NOT mean that the statistical result itself is independent of the number of trials of each stimulus.¡¯ I read sections 1.2.11 and 1.4.4. again and I am still not aware of how the statistical result IS dependent of the number of trials. As I have learned, only adequate trials could generate a good estimated response function. Is it something like this?
Same as above, see example 1.4.4.5. We want to compare the difference between areas under the curves of two stimuli. As the statistics textbook has told us, the statistical result will depend on the sample size, say here, the number of curves of each stimulus. What is the case here?
Thanks again.
Yufeng
