R help - May 2012 - Outcome~predictor model evaluation, repeated measurements

Dear all,


I have simple question regarding how to fit a model (i.e. linear) to the
data.

Say I have 10 subjects with different phenotypes (dependent var Y,
identical for a particular subject) and one predictor variable measured 3
times for each subject (X). By other words:

Y Subj X
1 1 1.2
1 1 1.3
1 1 0.7
3 2 2.1
3 2 2.5
3 2 4
5 3 3
5 3 4
5 3 4
...
20 10 12
20 10 13
20 10 12.5

Subj is a grouping variable.




I would like know the correlation of Y with X (Y~X) and the effect of
within subject variance on this correlation. And thus, overall significance
and correlation.

Will it be valid to fit lm to all combinations of x and y and take an
average values of p and R-squared?

Usually, I estmate the correlation using simple lm between outcome and
averaged predictor (1-to-1, i.e. 20 outcomes versus 20 predictors).
However, I would like to take in account variations associated with
replicated measurements (i.e. the same 20 outcomes versus 20 predictors
replicated say 3 times), and, therefore, evaluate slope and intercept
variabilities. Do mixed model regression analysis suitable for such an
analysis for example using lme function from nlme package? If not, what
kind of analysis is most appropriate? Thank you.

	[[alternative HTML version deleted]]

Dear all,

I have simple question regarding how to fit a model (i.e. linear) to the
data.
Say I have 10 subjects with different phenotypes (dependent var Y,
identical for a particular subject) and one predictor variable measured 3
times for each subject (X). By other words:

Y Subj X
1 1 1.2
1 1 1.3
1 1 0.7
3 2 2.1
3 2 2.5
3 2 4
5 3 3
5 3 4
5 3 4
...
20 10 12
20 10 13
20 10 12.5

Subj is a grouping variable.




I would like know the correlation of Y with X (Y~X) and the effect of
within subject variance on this correlation. And thus, overall significance
and correlation.

Will it be valid fitting lm to all combinations of x and y and take an
average values of p and R-squared?

Usually, I estmate the correlation using simple lm between outcome and
averaged predictor (1-to-1, i.e. 20 outcomes versus 20 predictors).
However, I would like to take in account variations associated with
replicated measurements (i.e. the same 20 outcomes versus 20 predictors
replicated say 3 times), and, therefore, evaluate slope and intercept
variabilities. Do mixed model regression analysis suitable for such an
analysis for example using lme function from nlme package? If not, what
kind of analysis is most appropriate? Weighted least squares?Thank you.

	[[alternative HTML version deleted]]

Dear nameless,

A mixed model seems reasonable for your kind of data. lme() from nlme or lmer()
from lme4 are good starting points.

Please note that there is R-sig-mixed-models: a R mailing list dedicated to
mixed models.

Best regards,

ir. Thierry Onkelinx
Instituut voor natuur- en bosonderzoek / Research Institute for Nature and
Forest
team Biometrie & Kwaliteitszorg / team Biometrics & Quality Assurance
Kliniekstraat 25
1070 Anderlecht
Belgium
+ 32 2 525 02 51
+ 32 54 43 61 85
Thierry.Onkelinx at inbo.be
www.inbo.be

To call in the statistician after the experiment is done may be no more than
asking him to perform a post-mortem examination: he may be able to say what the
experiment died of.
~ Sir Ronald Aylmer Fisher

The plural of anecdote is not data.
~ Roger Brinner

The combination of some data and an aching desire for an answer does not ensure
that a reasonable answer can be extracted from a given body of data.
~ John Tukey


-----Oorspronkelijk bericht-----
Van: r-help-bounces at r-project.org [mailto:r-help-bounces at r-project.org]
Namens rad mac
Verzonden: vrijdag 11 mei 2012 12:38
Aan: r-help at r-project.org
Onderwerp: [R] Outcome~predictor model evaluation, repeated measurements

Dear all,

I have simple question regarding how to fit a model (i.e. linear) to the data.
Say I have 10 subjects with different phenotypes (dependent var Y, identical for
a particular subject) and one predictor variable measured 3 times for each
subject (X). By other words:

Y Subj X
1 1 1.2
1 1 1.3
1 1 0.7
3 2 2.1
3 2 2.5
3 2 4
5 3 3
5 3 4
5 3 4
...
20 10 12
20 10 13
20 10 12.5

Subj is a grouping variable.




I would like know the correlation of Y with X (Y~X) and the effect of within
subject variance on this correlation. And thus, overall significance and
correlation.

Will it be valid fitting lm to all combinations of x and y and take an average
values of p and R-squared?

Usually, I estmate the correlation using simple lm between outcome and averaged
predictor (1-to-1, i.e. 20 outcomes versus 20 predictors).
However, I would like to take in account variations associated with replicated
measurements (i.e. the same 20 outcomes versus 20 predictors replicated say 3
times), and, therefore, evaluate slope and intercept variabilities. Do mixed
model regression analysis suitable for such an analysis for example using lme
function from nlme package? If not, what kind of analysis is most appropriate?
Weighted least squares?Thank you.

        [[alternative HTML version deleted]]

______________________________________________
R-help at r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
and provide commented, minimal, self-contained, reproducible code.
* * * * * * * * * * * * * D I S C L A I M E R * * * * * * * * * * * * *
Dit bericht en eventuele bijlagen geven enkel de visie van de schrijver weer en
binden het INBO onder geen enkel beding, zolang dit bericht niet bevestigd is
door een geldig ondertekend document.
The views expressed in this message and any annex are purely those of the writer
and may not be regarded as stating an official position of INBO, as long as the
message is not confirmed by a duly signed document.