@Spike, also note that even though the Sphericity has been violated, each of the additional options to test significance after accounting for the altered degrees of freedom still result in a significant F statistic for the Repeated ANOVA. You are correct, however that this box can't actually be ignored and, as stated previously, the error is one of my expectations for my intended (but not actual) audience. I will update this one though to account for this.
@Spike, T-tests are fine if you rejected the null as they are just as post hoc test. You run Fischer's LSD and Tukey's the same way after a One Way Between Subjects ANOVA. You can always use Bonferoni corection to make sure your Type 1 stays low (but note you don't do that in a Tukey's or Ficher's test either).
@Spike, you are correct, but when I made these videos, I intended them to only be for my students in my Introductory class, whom I do not require to understand that particular window in the SPSS output. I do, however, have this on my list of new/updated videos since I have found others have been using the test.
I don't understand why the Mauchly test has been ignored. Isn't this the stat that allows us to assume whether or not the assumption of spheracity has been met? So if you ignore this how can you feasibly use the stats that assume it's been met when it infact hasn't (as is the case with the example you use). Also surely doing numerous repeated t-tests this increases the likelihood of type I errors? I thought the point of doing a repeated measures ANOVA was to avoid doing that exact thing?
@Spike, also note that even though the Sphericity has been violated, each of the additional options to test significance after accounting for the altered degrees of freedom still result in a significant F statistic for the Repeated ANOVA. You are correct, however that this box can't actually be ignored and, as stated previously, the error is one of my expectations for my intended (but not actual) audience. I will update this one though to account for this.
bernstmj 3 months ago
@Spike, T-tests are fine if you rejected the null as they are just as post hoc test. You run Fischer's LSD and Tukey's the same way after a One Way Between Subjects ANOVA. You can always use Bonferoni corection to make sure your Type 1 stays low (but note you don't do that in a Tukey's or Ficher's test either).
bernstmj 3 months ago
@Spike, you are correct, but when I made these videos, I intended them to only be for my students in my Introductory class, whom I do not require to understand that particular window in the SPSS output. I do, however, have this on my list of new/updated videos since I have found others have been using the test.
bernstmj 3 months ago
sphericity*
spike854 3 months ago
I don't understand why the Mauchly test has been ignored. Isn't this the stat that allows us to assume whether or not the assumption of spheracity has been met? So if you ignore this how can you feasibly use the stats that assume it's been met when it infact hasn't (as is the case with the example you use). Also surely doing numerous repeated t-tests this increases the likelihood of type I errors? I thought the point of doing a repeated measures ANOVA was to avoid doing that exact thing?
spike854 3 months ago
helped a lot thanks
twf102447 3 months ago
@1just2another3girl4, Great! I am happy to help. =)
bernstmj 3 months ago
Great video, really helped me out!
1just2another3girl4 3 months ago