 |
 |
On 13 May 2019, at 13:31, Rosalia Dacosta Aguayo <[log in to unmask]> wrote:
Hi Matthew,
I have seen the problem. Because I am interested in the different effect of gender in a continuos variable. Instead of gender I have included the scores of this variable for male (in one column and female in another column). Age as a covariate. I would like to know if the variable X have different slopes and if these slopes interact. Then, the same for age. Does make it sense?
As I had the matrix, I was looking at gender differences in white matter adjusting for age. This is not what I want. So, I have included the scores in variable X splitted in two columns (scores for males and scores for females).
Rosalia
Hi Rosalia,What interaction do you want to test for specifically? The only potential interaction in your current model seems to be GenderxAge.
Kind RegardsMatthew
--------------------------------
Dr Matthew Webster
FMRIB Centre
John Radcliffe Hospital
University of Oxford
On 13 May 2019, at 11:11, Rosalia Dacosta Aguayo <[log in to unmask]> wrote:
Hi Matthew,
Thank you. What about testing if there is an interaction?
Rosalia
Hi Rosalia,If I understand your design correctly, then your first two contrasts test for what you are interested in ( if there is any gender difference, and the direction of that gender difference ). There is no need to have a column for the dependent variable itself - that is provided by the data!
Kind RegardsMatthew
--------------------------------
Dr Matthew Webster
FMRIB Centre
John Radcliffe Hospital
University of Oxford
On 13 May 2019, at 04:22, Rosalia Dacosta Aguayo <[log in to unmask]> wrote:
Dear FSL experts,
I have a group of 90 Healthy Controls with DTI data and I would like to test if gender has a different effect in one continuous dependent variable with age as a covariate.
I have been looking to the different designs proposed in the FSL wiki and in the mail list. It seems that depending on if I split gender in two EVs or not, I could get some answer but not others. I would like to test if there is any differences in white matter integrity in relation to gender for the dependent variable. If the difference is more in females than in males in relation to the dependent variable and if there is main effect of one of the gender or any interaction between gender and the dependent variable.
Setting the design of the matrix this way:
Group Male Female Variable X Covariate 1 1 0 1 1 0 1 0 1 1 0 1 1 0 1 1 1 0 1 1 0 1 1 0 1 0 1 1 0 1
C1: male>female 1 -1 0 0 I am worried about the fact the Variable X is my dependent variable and I am setting to 0 as it was a covariate. So, should I add a 1 for the dependent variable and a + 0.5 for females and -0.5 for males?C2: female> male -1 1 0 0 Same question here....C3: interaction male/female? I do not know how I could test the interaction.
Would it be possible in the same model to know if the covariate is significant? I do not think so, I should run the model with and without covariate...?
Thank you for your time,Rosalia
To unsubscribe from the FSL list, click the following link:
https://www.jiscmail.ac.uk/cgi-bin/webadmin?SUBED1=FSL&A=1
To unsubscribe from the FSL list, click the following link:
https://www.jiscmail.ac.uk/cgi-bin/webadmin?SUBED1=FSL&A=1
To unsubscribe from the FSL list, click the following link:
https://www.jiscmail.ac.uk/cgi-bin/webadmin?SUBED1=FSL&A=1
To unsubscribe from the FSL list, click the following link:
https://www.jiscmail.ac.uk/cgi-bin/webadmin?SUBED1=FSL&A=1
To unsubscribe from the FSL list, click the following link:
https://www.jiscmail.ac.uk/cgi-bin/webadmin?SUBED1=FSL&A=1
To unsubscribe from the FSL list, click the following link:
https://www.jiscmail.ac.uk/cgi-bin/webadmin?SUBED1=FSL&A=1