Dear Wangjiao,
I couldn't see your design matrix but from what you describe I would
expect 76 columns: (2*2*3+6+1)*4 = 76.
Otherwise you are right that each regressor modelling (the amplitude of)
each of your four experimental conditions is immediately followed by two
extra regressors corresponding to the convolution of that condition by
the time and dispersion derivatives of the canonical HRF.
Best regards,
Guillaume.
On 07/01/2019 08:50, 王娇 wrote:
>
>
>> 在 2019年1月7日,下午4:41,王娇 <[log in to unmask]
>> <mailto:[log in to unmask]>> 写道:
>>
>> Hi, all,
>>
>> Thank you for answering my question~
>>
>> My question is that I have two within- group conditions, and every
>> condition has two levels. And I have four runs. In 1-st level model
>> specification, I inserted two variables in the factorial design and I
>> also added model derivatives including both time and dispersion
>> derivatives into the model. And I also put motion parameters in each
>> run when I specify every model. As a result, I obtained 74 columns
>> (please see below).
>>
>> <PastedGraphic-1.png>
>> It means that there are two columns following every column of
>> operationalized experimental condition(for there are 18 columns for
>> each run, minus 6 motion parameter. there are 12 columns for the
>> operationalized experimental conditions in four runs). *_Who can tell
>> me what are the two columns following the experimental condtion
>> represent for? I guess they represent for the time and dispersion
>> derivates…._*
>> *_
>> _*
>> all the best
>> wangjiao
>> 1.7
>> *_
>> _*
>>
>>
>>
>>
>>
>>
>
--
Guillaume Flandin, PhD
Wellcome Centre for Human Neuroimaging
UCL Queen Square Institute of Neurology
London WC1N 3BG
|