search

LIMO-EEG-Toolbox / limo_tools

Hierarchical Linear Modelling for MEEG data
https://limo-eeg-toolbox.github.io/limo_meeg/
Other
59 stars 27 forks source link

Two within categorical conditions in LIMO #20

Closed neurolinguist closed 6 years ago

neurolinguist commented 6 years ago

Dear Dr. Pernet,

Let me ask the format of category specification. My experiment is designed with 2-by-2 conditions with two levels for each. The number of event-markers for epoching are eight with two markers corresponding to each level. I wrote "category-variables.txt" with two columns and with two rows ending the first row with ";." However, the 22 design is not recognized is design matrix, and the matrix is represented as 88. Could you give me any advice? Thank you in advance.

Sincerely,

Shingo Tokimoto

CPernet commented 6 years ago

Hi Shingo

For 'standard' analyses, all you need is to specify you N conditions, so in your case one column with 8 markers. I created a wiki page to help understand that see https://github.com/LIMO-EEG-Toolbox/limo_eeg/wiki/Defining-conditions

neurolinguist commented 6 years ago

Dear Dr. Pernet,

Thank you for the reply and the instruction page. Yes, I intended to make the category.txt for "N-conditions with/without interactions." I have 24 participants and have made a STUDY structure including them. I consulted "trialinfo" in "STUDY.datasetinfo" in VARIABLE. The beginning of the 'type" field is as follows (The event marker is recorded by BrainVisionRecorder).

'S 4' 'S 2' 'S 4' 'S 2' 'S 4' 'S 2' 'S 4' 'S 3' 'S 4' 'S 3' 'S 2' 'S 4' 'S 1'

The experimental design is 2-by-2, and thus I re-coded the event markers as follows. My intention here is that S1 and S2 correspond to "1" and S3 and S4 to "2" for the 2 levels of the first factor respectively and that S1 and S3 correspond to "3" and S2 and S4 to "4" for the 2 levels of the second factor respectively again.

2 4 1 4 2 4 1 4 2 4 1 4 2 4 2 3 2 4 2 3

However, the design matrix appears as 88 not as 22. Could you point out what is wrong with me? Thank you in advance again.

Sincerely,

Shingo Tokimoto

CPernet commented 6 years ago

your design matrix should have only 4 columns, 2 factor 1 and 2 factor 2 -- why would you expect 22?

neurolinguist commented 6 years ago

Dear Dr. Pernet,

Thank you for the reply again. Yes, I am expecting 4 columns, but Design matrix has 8 columns, as attached below.

2-by-2_design

Is this correct for the intended analysis? Thank you again.

Shingo Tokimoto

CPernet commented 6 years ago

yes, assuming you specified full factorial -- now remember that this is useful for single subject analyses; very unlikely that you can do any group analysis with such design.