Hello to everybody, I have a data frame with 100 measures of quality for 3 variables: A, B and C. These quality variables are measured in diferent times along the productive process. My data comes from 5 experiments (5 replicates with 20 measures for replicate). I also have a final measure (Z) but just one measure for each unit, that is, for the 20 units that are measured on each replica. My objetive is to study the relationships between the 3 quality parameters with the last measure, that is: lm(Z ~ A+B+C, data=mydata) I have found significant differences between replicas for each qualite parameters (A, B and C) and I would like to include the replica effect as a random effect: lme(Z ~ A+B+C, data=mydata, random=~1|replica) And here is my problem. I know that there are signifficant diferences between replicas but since the final measure, Z, is the same for each replica I do not know how to deal with. Can you help me? How could I take into account the variability due to the replica when I want to study the effects of variables A, B and C on the final result of a productive process? Thank you in advance. ----- Manuel Ram?n Fern?ndez Group of Reproductive Biology (GBR) University of Castilla-La Mancha (Spain) mramon at jccm.es -- View this message in context: http://www.nabble.com/How-to-deal-with-this-random-variable--tp24684341p24684341.html Sent from the R help mailing list archive at Nabble.com.
This sounds way too complicated for this forum, which is designed to provide help to users on the use of the R language, not remote statistical consulting. While you may receive replies, I would argue that you would do better to find a local statistical expert with whom to work -- not least because they should probably have a deep understanding of how your experiment was conducted, data gathered, measurements made, etc. to be able to give you worthwhile advice. Long distance consulting based on incomplete understanding is very risky. Caveat emptor! Bert Gunter Genentech Nonclinical Biostatistics -----Original Message----- From: r-help-bounces at r-project.org [mailto:r-help-bounces at r-project.org] On Behalf Of Manuel Ramon Sent: Monday, July 27, 2009 9:54 AM To: r-help at r-project.org Subject: [R] How to deal with this random variable? Hello to everybody, I have a data frame with 100 measures of quality for 3 variables: A, B and C. These quality variables are measured in diferent times along the productive process. My data comes from 5 experiments (5 replicates with 20 measures for replicate). I also have a final measure (Z) but just one measure for each unit, that is, for the 20 units that are measured on each replica. My objetive is to study the relationships between the 3 quality parameters with the last measure, that is: lm(Z ~ A+B+C, data=mydata) I have found significant differences between replicas for each qualite parameters (A, B and C) and I would like to include the replica effect as a random effect: lme(Z ~ A+B+C, data=mydata, random=~1|replica) And here is my problem. I know that there are signifficant diferences between replicas but since the final measure, Z, is the same for each replica I do not know how to deal with. Can you help me? How could I take into account the variability due to the replica when I want to study the effects of variables A, B and C on the final result of a productive process? Thank you in advance. ----- Manuel Ram?n Fern?ndez Group of Reproductive Biology (GBR) University of Castilla-La Mancha (Spain) mramon at jccm.es -- View this message in context: http://www.nabble.com/How-to-deal-with-this-random-variable--tp24684341p2468 4341.html Sent from the R help mailing list archive at Nabble.com. ______________________________________________ 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.
Thank you for your replay Bert. You are right, is complicated to get a good response when people do not know how the experiment was conducted, etc. The main problem, maybe, is that this experiment has a wrong design being complicated to get some good conclusion from it. I read this forum frequently and I found a lot of useful information on it. For that reason I decided to ask to the forum; maybe someone can help us. Thank you again for your response Bert. Bert Gunter wrote:> > This sounds way too complicated for this forum, which is designed to > provide > help to users on the use of the R language, not remote statistical > consulting. While you may receive replies, I would argue that you would do > better to find a local statistical expert with whom to work -- not least > because they should probably have a deep understanding of how your > experiment was conducted, data gathered, measurements made, etc. to be > able > to give you worthwhile advice. > > Long distance consulting based on incomplete understanding is very risky. > Caveat emptor! > > Bert Gunter > Genentech Nonclinical Biostatistics > > > -----Original Message----- > From: r-help-bounces at r-project.org [mailto:r-help-bounces at r-project.org] > On > Behalf Of Manuel Ramon > Sent: Monday, July 27, 2009 9:54 AM > To: r-help at r-project.org > Subject: [R] How to deal with this random variable? > > > Hello to everybody, > I have a data frame with 100 measures of quality for 3 variables: A, B and > C. These quality variables are measured in diferent times along the > productive process. My data comes from 5 experiments (5 replicates with 20 > measures for replicate). I also have a final measure (Z) but just one > measure for each unit, that is, for the 20 units that are measured on each > replica. > > My objetive is to study the relationships between the 3 quality parameters > with the last measure, that is: > > lm(Z ~ A+B+C, data=mydata) > > I have found significant differences between replicas for each qualite > parameters (A, B and C) and I would like to include the replica effect as > a > random effect: > > lme(Z ~ A+B+C, data=mydata, random=~1|replica) > > And here is my problem. I know that there are signifficant diferences > between replicas but since the final measure, Z, is the same for each > replica I do not know how to deal with. > > Can you help me? How could I take into account the variability due to the > replica when I want to study the effects of variables A, B and C on the > final result of a productive process? > > Thank you in advance. > > ----- > Manuel Ram?n Fern?ndez > Group of Reproductive Biology (GBR) > University of Castilla-La Mancha (Spain) > mramon at jccm.es > -- > View this message in context: > http://www.nabble.com/How-to-deal-with-this-random-variable--tp24684341p2468 > 4341.html > Sent from the R help mailing list archive at Nabble.com. > > ______________________________________________ > 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. > > ______________________________________________ > 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. > >----- Manuel Ram?n Fern?ndez Group of Reproductive Biology (GBR) University of Castilla-La Mancha (Spain) mramon at jccm.es -- View this message in context: http://www.nabble.com/How-to-deal-with-this-random-variable--tp24684341p24695050.html Sent from the R help mailing list archive at Nabble.com.