Statistical analisys of EEG data

Forums: Spss Eeg Anova
Email this Topic • Print this Page

I have 2 seconds epochs of EEG recordings in eyes closed and eyes opened condition for 80 participants (two groups of 40), but a number of epochs is not the same for every participant. I will compute spectral power values for alpha rhythm. Also, there are two verbal abilities tests that were applied.

My first confusion is about a number of epochs per subject, per group and conditions. I don't have an even number, the range is between 18 and 120. How can I compute is there a significant difference between groups, conditions or subjects? Do I even need that if I use repeated measures ANOVA? Or should I take a mean value for every subject? Because of the too many data, I would prefer to take a mean for u subject, but can I do that if there is a different number of epochs between subjects, groups or conditions?

Topic Stats
Top Replies
Link to this Topic

Type: Question • Score: 0 • Views: 230 • Replies: 4

No top replies

@ninka,

The ANOVA will work fine with multiple measurements per subject even if the number is different. You can also take an average per subject if you want. You have enough samples per subject that the differences in mean confidence is minimal.

1 Reply

@engineer,

Thank you for your answer!

If I take an average per subject, does it mean that I don't need to calculate is s there a statistical difference between groups and conditions in a number of epochs?

1 Reply

@ninka,

I would think not, but you should examine the data inside your epochs to see what your distribution looks like. Do you have extreme outliers? Do the numbers inside an epoch look consistent or does it look like a random number generator?

1 Reply

@engineer,

Ther are no consistencies in a number of epochs, they are really random. But if I enter number of epochs in SPSS and check for normality of distribution, the Shapiro-Wilk test shows me that data has a normal distribution... And when I apply Independent sample T-test on that data, I get that there is no difference in a number of epochs between groups...But...I can't write that in my paper...
This is an example from one paper:

"The remaining number of segments for analyses did
not differ between conditions (F(1, 48) = 1.79, p = .19) and groups (F(1, 48) = 2.00, p = .16)"

0 Replies

Forums
» Statistical analisys of EEG data

Read Question
Reply to All

Statistical analisys of EEG data

Quick Links

My Account

able2know