Something does not make sense in R. It has to do with the question of balance and unbalance. *A<-factor(c(1,1,1,1,1,1,2,2,2,2,2,2))* *B<-factor(c(1,1,2,2,3,3,1,1,2,2,3,3))* *y<-rnorm(12)* *mod<-aov(y~A+B)* I was under the impression that the design is balanced ie order does not effect the sums of squares. However, when I compute the anova R reports that the Estimated Effects are Unbalanced. I thought that when all combinations of levels of A and B have equal replications then the design is called balanced. But, R tends to think that when not all levels of A and levels of B have equal replication, then the "Estimated Effects are unbalanced".... Is this the same as the design being unbalanced? Because for the example below, where the error occured, the order does not matter (which make me think that the design is balanced). *Call:* * aov(formula = y ~ A + B)* *Terms:* * A B Residuals* *Sum of Squares 0.872572 0.025604 16.805706* *Deg. of Freedom 1 2 10* *Residual standard error: 1.296368* *Estimated effects may be unbalanced* -- Yours sincerely, Justin *I check my email at 9AM and 4PM everyday* *If you have an EMERGENCY, contact me at +447938674419(UK) or +60125056192(Malaysia)* [[alternative HTML version deleted]]
Please! "when I compute the anova R reports that the Estimated Effects are Unbalanced" It does *not* say this. It says that they **may** be unbalanced. They are not. -- Bert Bert Gunter "The trouble with having an open mind is that people keep coming along and sticking things into it." -- Opus (aka Berkeley Breathed in his "Bloom County" comic strip ) On Mon, Aug 22, 2016 at 8:15 AM, Justin Thong <justinthong93 at gmail.com> wrote:> Something does not make sense in R. It has to do with the question of > balance and unbalance. > > *A<-factor(c(1,1,1,1,1,1,2,2,2,2,2,2))* > *B<-factor(c(1,1,2,2,3,3,1,1,2,2,3,3))* > *y<-rnorm(12)* > *mod<-aov(y~A+B)* > > I was under the impression that the design is balanced ie order does not > effect the sums of squares. However, when I compute the anova R reports > that the Estimated Effects are Unbalanced. I thought that when all > combinations of levels of A and B have equal replications then the design > is called balanced. But, R tends to think that when not all levels of A and > levels of B have equal replication, then the "Estimated Effects are > unbalanced".... Is this the same as the design being unbalanced? Because > for the example below, where the error occured, the order does not matter > (which make me think that the design is balanced). > > > *Call:* > * aov(formula = y ~ A + B)* > > *Terms:* > * A B Residuals* > *Sum of Squares 0.872572 0.025604 16.805706* > *Deg. of Freedom 1 2 10* > > *Residual standard error: 1.296368* > *Estimated effects may be unbalanced* > -- > Yours sincerely, > Justin > > *I check my email at 9AM and 4PM everyday* > *If you have an EMERGENCY, contact me at +447938674419(UK) or > +60125056192(Malaysia)* > > [[alternative HTML version deleted]] > > ______________________________________________ > R-help at r-project.org mailing list -- To UNSUBSCRIBE and more, see > 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.
The problem is that you have 12 observations and 1+2+10=13 degrees of freedom. There should be 1 + 2 + 8 = 11 degrees of freedom. Probably one of your variables is masked by something else in you workspace. Protect yourself by using a data.frame> tmp <- data.frame(A=factor(c(1,1,1,1,1,1,2,2,2,2,2,2)),+ B=factor(c(1,1,2,2,3,3,1,1,2,2,3,3)), + y=rnorm(12))> mod <- aov(y ~ A+B, data=tmp) > summary(mod)Df Sum Sq Mean Sq F value Pr(>F) A 1 1.553 1.553 1.334 0.281 B 2 3.158 1.579 1.357 0.311 Residuals 8 9.311 1.164 On Mon, Aug 22, 2016 at 11:15 AM, Justin Thong <justinthong93 at gmail.com> wrote:> Something does not make sense in R. It has to do with the question of > balance and unbalance. > > *A<-factor(c(1,1,1,1,1,1,2,2,2,2,2,2))* > *B<-factor(c(1,1,2,2,3,3,1,1,2,2,3,3))* > *y<-rnorm(12)* > *mod<-aov(y~A+B)* > > I was under the impression that the design is balanced ie order does not > effect the sums of squares. However, when I compute the anova R reports > that the Estimated Effects are Unbalanced. I thought that when all > combinations of levels of A and B have equal replications then the design > is called balanced. But, R tends to think that when not all levels of A and > levels of B have equal replication, then the "Estimated Effects are > unbalanced".... Is this the same as the design being unbalanced? Because > for the example below, where the error occured, the order does not matter > (which make me think that the design is balanced). > > > *Call:* > * aov(formula = y ~ A + B)* > > *Terms:* > * A B Residuals* > *Sum of Squares 0.872572 0.025604 16.805706* > *Deg. of Freedom 1 2 10* > > *Residual standard error: 1.296368* > *Estimated effects may be unbalanced* > -- > Yours sincerely, > Justin > > *I check my email at 9AM and 4PM everyday* > *If you have an EMERGENCY, contact me at +447938674419(UK) or > +60125056192(Malaysia)* > > [[alternative HTML version deleted]] > > ______________________________________________ > R-help at r-project.org mailing list -- To UNSUBSCRIBE and more, see > 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.
Thanks, Rich. I didn't notice that! -- Bert Bert Gunter "The trouble with having an open mind is that people keep coming along and sticking things into it." -- Opus (aka Berkeley Breathed in his "Bloom County" comic strip ) On Mon, Aug 22, 2016 at 1:43 PM, Richard M. Heiberger <rmh at temple.edu> wrote:> The problem is that you have 12 observations and 1+2+10=13 degrees of freedom. > There should be 1 + 2 + 8 = 11 degrees of freedom. > Probably one of your variables is masked by something else in you workspace. > Protect yourself by using a data.frame > >> tmp <- data.frame(A=factor(c(1,1,1,1,1,1,2,2,2,2,2,2)), > + B=factor(c(1,1,2,2,3,3,1,1,2,2,3,3)), > + y=rnorm(12)) >> mod <- aov(y ~ A+B, data=tmp) >> summary(mod) > Df Sum Sq Mean Sq F value Pr(>F) > A 1 1.553 1.553 1.334 0.281 > B 2 3.158 1.579 1.357 0.311 > Residuals 8 9.311 1.164 > > On Mon, Aug 22, 2016 at 11:15 AM, Justin Thong <justinthong93 at gmail.com> wrote: >> Something does not make sense in R. It has to do with the question of >> balance and unbalance. >> >> *A<-factor(c(1,1,1,1,1,1,2,2,2,2,2,2))* >> *B<-factor(c(1,1,2,2,3,3,1,1,2,2,3,3))* >> *y<-rnorm(12)* >> *mod<-aov(y~A+B)* >> >> I was under the impression that the design is balanced ie order does not >> effect the sums of squares. However, when I compute the anova R reports >> that the Estimated Effects are Unbalanced. I thought that when all >> combinations of levels of A and B have equal replications then the design >> is called balanced. But, R tends to think that when not all levels of A and >> levels of B have equal replication, then the "Estimated Effects are >> unbalanced".... Is this the same as the design being unbalanced? Because >> for the example below, where the error occured, the order does not matter >> (which make me think that the design is balanced). >> >> >> *Call:* >> * aov(formula = y ~ A + B)* >> >> *Terms:* >> * A B Residuals* >> *Sum of Squares 0.872572 0.025604 16.805706* >> *Deg. of Freedom 1 2 10* >> >> *Residual standard error: 1.296368* >> *Estimated effects may be unbalanced* >> -- >> Yours sincerely, >> Justin >> >> *I check my email at 9AM and 4PM everyday* >> *If you have an EMERGENCY, contact me at +447938674419(UK) or >> +60125056192(Malaysia)* >> >> [[alternative HTML version deleted]] >> >> ______________________________________________ >> R-help at r-project.org mailing list -- To UNSUBSCRIBE and more, see >> 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 -- To UNSUBSCRIBE and more, see > 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.
On 22 Aug 2016, at 17:15 , Justin Thong <justinthong93 at gmail.com> wrote:> Something does not make sense in R. It has to do with the question of > balance and unbalance. > > *A<-factor(c(1,1,1,1,1,1,2,2,2,2,2,2))* > *B<-factor(c(1,1,2,2,3,3,1,1,2,2,3,3))* > *y<-rnorm(12)* > *mod<-aov(y~A+B)*As an aside, notice that this is one case where posting in HTML goes against your own interests (let alone the posting guide). The above cannot be cut+pasted, so fewer people will be willing to try it. -- Peter Dalgaard, Professor, Center for Statistics, Copenhagen Business School Solbjerg Plads 3, 2000 Frederiksberg, Denmark Phone: (+45)38153501 Office: A 4.23 Email: pd.mes at cbs.dk Priv: PDalgd at gmail.com