| BA.est {MethComp} | R Documentation |
A variance component model is fitted to method comparison data with replicate measurements in each method by item stratum. The purpose is to simplify the construction of a correct Bland-Altman-plot.
BA.est( data, linked=TRUE, exch=!linked, MI=TRUE )
data |
A data frame representing method comparison data with replicate
measurements, i.e. with variables meth, item,
repl and y. |
linked |
Are the replicated linked within item across methods? |
exch |
Are the replicated exchangeable within (item,methods)? This argument is just for convenience. |
MI |
Logical. Should the method by item interaction be included in the model. |
The model fitted is:
y=alpha_m + mu_i + c_mi + a_ir + e_ir, var(c_mi)=tau_m^2, var(a_ir)=nu^2, var(e_mir)=sigma_m^2
We can only fit separate variances for the tau's if more than
two methods are compared (i.e. nM > 2).
NOTE: Currently, the author is too ignorant about the ins and outs of lme
to find out how to get different tau's when nM>2.
Therefore the estimates in this case are equal i.e. from the model postulated.
A list with four elements:
Bias |
Vector of estimates of alpha_m, the first element is always 0. |
Var.comp |
Two-column matrix of sds and variances of the variance components,
nM tau's,
one nu if linked, and
nM sigma's.
Only those in the model specified are included. |
LoA |
Four-column matrix with mean difference, lover and upper limit of agreement and prediction SD. Each row in the matrix represents a pair of methods. |
Rep.coef |
Two-column matrix of repeatability SDs and repeatability coefficients (2.77*SD). |
Bendix Carstensen
Carstensen, Simpson & Gurrin: Statistical models for assessing agreement in method comparison studies with replicate measurements, Preprent, Departmet of Biostatistics, University of Copenhagen, 2007, http://www.pubhealth.ku.dk/bs/publikationer/Research_report_07-14.pdf/.
data( ox ) BA.est( ox ) BA.est( ox, exch=TRUE ) data( sbp ) BA.est( sbp ) BA.est( sbp, exch=TRUE )