help > Contrast help
Showing 1-6 of 6 posts
Sep 9, 2019 03:09 PM | lynette looi
Contrast help
Dear Dr.Zalesky,
I wanted to check if I am using the correct contrasts:
I am interested in if there are more functional connectivity found in a control group as compared to a patient group, while adding age as covariates:
So I have used the following contrasts of [1 0 0] using a T-test?The first being control, second being patient and last being age.
Thank you so much.
Kind regards,
Lynette
I wanted to check if I am using the correct contrasts:
I am interested in if there are more functional connectivity found in a control group as compared to a patient group, while adding age as covariates:
So I have used the following contrasts of [1 0 0] using a T-test?The first being control, second being patient and last being age.
Thank you so much.
Kind regards,
Lynette
Sep 10, 2019 12:09 AM | Andrew Zalesky
RE: Contrast help
Hi Lynette,
No - this is not correct.
In your design matrix, you will need a column of all ones, a column for group (1=patient, 0=control), a column for age and a column for the interaction between group and age. The last column is simply given by the multiplication of the age and group columns.
To test for an interaction between age and group, the contrast for a t-test is then [0 0 1] or [0 0 -1].
Andrew
Originally posted by lynette looi:
No - this is not correct.
In your design matrix, you will need a column of all ones, a column for group (1=patient, 0=control), a column for age and a column for the interaction between group and age. The last column is simply given by the multiplication of the age and group columns.
To test for an interaction between age and group, the contrast for a t-test is then [0 0 1] or [0 0 -1].
Andrew
Originally posted by lynette looi:
Dear Dr.Zalesky,
I wanted to check if I am using the correct contrasts:
I am interested in if there are more functional connectivity found in a control group as compared to a patient group, while adding age as covariates:
So I have used the following contrasts of [1 0 0] using a T-test?The first being control, second being patient and last being age.
Thank you so much.
Kind regards,
Lynette
I wanted to check if I am using the correct contrasts:
I am interested in if there are more functional connectivity found in a control group as compared to a patient group, while adding age as covariates:
So I have used the following contrasts of [1 0 0] using a T-test?The first being control, second being patient and last being age.
Thank you so much.
Kind regards,
Lynette
Sep 10, 2019 08:09 AM | lynette looi
RE: Contrast help
Dear Andrew,
thank you for the prompt reply, and for clarifying my mistake.
I hypothesized that age will influence control and patient group differently and want to control for age-group interactions.
Will it be the following:
1 1 12 12
1 1 10 10
1 1 8 8
1 0 11 0
1 0 15 0
1 0 18 0
the first column being constant, second column being groups (1- Control, 0-Patient), third column being age, and the fourth to control for age-group interactions?
And to test for group effects, the contrasts to be [0, 1, 0, 0] based on prior hypothesis that control>patient?
Thanks,
Lynette
thank you for the prompt reply, and for clarifying my mistake.
I hypothesized that age will influence control and patient group differently and want to control for age-group interactions.
Will it be the following:
1 1 12 12
1 1 10 10
1 1 8 8
1 0 11 0
1 0 15 0
1 0 18 0
the first column being constant, second column being groups (1- Control, 0-Patient), third column being age, and the fourth to control for age-group interactions?
And to test for group effects, the contrasts to be [0, 1, 0, 0] based on prior hypothesis that control>patient?
Thanks,
Lynette
Sep 10, 2019 10:09 AM | Andrew Zalesky
RE: Contrast help
Yes - this is all correct
Originally posted by lynette looi:
Originally posted by lynette looi:
Dear Andrew,
thank you for the prompt reply, and for clarifying my mistake.
I hypothesized that age will influence control and patient group differently and want to control for age-group interactions.
Will it be the following:
1 1 12 12
1 1 10 10
1 1 8 8
1 0 11 0
1 0 15 0
1 0 18 0
the first column being constant, second column being groups (1- Control, 0-Patient), third column being age, and the fourth to control for age-group interactions?
And to test for group effects, the contrasts to be [0, 1, 0, 0] based on prior hypothesis that control>patient?
Thanks,
Lynette
thank you for the prompt reply, and for clarifying my mistake.
I hypothesized that age will influence control and patient group differently and want to control for age-group interactions.
Will it be the following:
1 1 12 12
1 1 10 10
1 1 8 8
1 0 11 0
1 0 15 0
1 0 18 0
the first column being constant, second column being groups (1- Control, 0-Patient), third column being age, and the fourth to control for age-group interactions?
And to test for group effects, the contrasts to be [0, 1, 0, 0] based on prior hypothesis that control>patient?
Thanks,
Lynette
Sep 12, 2019 11:09 AM | lynette looi
RE: Contrast help
Dear Andrew,
Thank you so much for your help. I just have one last question please.
Earlier you mentioned that a T-test of [0 0 0 1] or [0 0 0 -1] can be used to look at the age-group interaction, but I don't quite understand what that will mean when significance is found in either?
Thank you.
Kind regards,
Lynette
Thank you so much for your help. I just have one last question please.
Earlier you mentioned that a T-test of [0 0 0 1] or [0 0 0 -1] can be used to look at the age-group interaction, but I don't quite understand what that will mean when significance is found in either?
Thank you.
Kind regards,
Lynette
Sep 13, 2019 12:09 AM | Andrew Zalesky
RE: Contrast help
Hi Lynette,
The contrast [0 0 0 1] whether the the slope of the age-connectivity relationship is steeper in the group coded with 1, whereas [0 0 0 -1] will tester whether the slope is less steep.
In other words it is testing whether the age effect is stronger or weaker in one of the particular groups.
Andrew
Originally posted by lynette looi:
The contrast [0 0 0 1] whether the the slope of the age-connectivity relationship is steeper in the group coded with 1, whereas [0 0 0 -1] will tester whether the slope is less steep.
In other words it is testing whether the age effect is stronger or weaker in one of the particular groups.
Andrew
Originally posted by lynette looi:
Dear Andrew,
Thank you so much for your help. I just have one last question please.
Earlier you mentioned that a T-test of [0 0 0 1] or [0 0 0 -1] can be used to look at the age-group interaction, but I don't quite understand what that will mean when significance is found in either?
Thank you.
Kind regards,
Lynette
Thank you so much for your help. I just have one last question please.
Earlier you mentioned that a T-test of [0 0 0 1] or [0 0 0 -1] can be used to look at the age-group interaction, but I don't quite understand what that will mean when significance is found in either?
Thank you.
Kind regards,
Lynette