Abstract
The use of a replicated Latin square design for reference material homogeneity assessment is illustrated by application to a homogeneity study of eight high-purity organic materials certified for melting point. The design controlled for both a three-level location effect and a run effect. Variance components were extracted using mixed effects modelling using a restricted maximum likelihood method. An alternative classical ANOVA calculation is also given. The effect of appreciable numerical rounding by the instrument software was investigated and shown to be acceptable for the particular example. Estimation of the scale of location and run effects showed that in this example the location effect was both statistically and practically significant, while the run effect was not statistically significant at the 95% level of confidence. The design allowed unbiased estimates of between-unit variances in the presence of both interfering effects.
References
ISO Guide 34 (2009) General requirements for the competence of reference material producers. International Organization for Standardization, Geneva
ISO Guide 35 (2006) Reference materials—general and statistical principles for certification. International Organization for Standardization, Geneva
Ellison SLR, Barwick VJ, Duguid Farrant TJ (2009) Practical statistics for the analytical scientist. A bench guide. Royal Society of Chemistry, Cambridge, UK. ISBN 978-0-85404-131-2
Box GEP, Hunter WG, Hunter JS (1978) Statistics for experimenters. Wiley, Hoboken NJ. ISBN 0-471-09315-7
Thompson M, Ellison SLR, Wood R (2006) Pure Appl Chem 78:145–196
ISO 13528 (2005) Statistical methods for use in proficiency testing by interlaboratory comparisons. International Organization for Standardization, Geneva
R Development Core Team (2009) R: a language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. ISBN 3-900051-07-0, URL http://www.R-project.org
Bates D, Maechler M (2010) lme4: linear mixed-effects models using S4 classes. R package version 0.999375-35. http://CRAN.R-project.org/package=lme4
Chambers JM, Hastie TJ (1992) Statistical models in S. Wadsworth & Brooks/Cole, Pacific Grove, CA. ISBN 0412052911
Pinheiro JC, Bates DM (2000) Mixed-effects models in S and S-PLUS. Springer, New York, NY. ISBN 0-387-98957-9
Burr T, Hamada MS, Cremers T, Weaver BP, Howell J, Croft S, Vardeman SB (2011) Accred Qual Assur 16:347–359
Searle SR, Casella G, McCulloch CE (2006) Variance components. Wiley, Hoboken, NJ. ISBN 978-0-470-00959-8
Ellison SLR, Burke S, Walker RF, Heydorn K, Månsson M, Pauwels J, Wegscheider W, te Nijenhuis B (2001) Accred Qual Assur 6:274–277
Van der Veen AMH, Linsinger TPJ, Lamberty A, Pauwels J (2001) Accred Qual Assur 6:257–263
Author information
Authors and Affiliations
Corresponding author
Appendix: Manual calculation for a replicated Latin square design with one fixed effect
Appendix: Manual calculation for a replicated Latin square design with one fixed effect
The statistical model for the replicated Latin square design used here can be written
where y lijk is the observation for CRM unit j in run k and at location l in set i, μ denotes the population mean, L l the effect of location l on the observed melting point where S i ~ N(0, σ S) is a nominal offset for set (~N(θ, σ) denoting ‘distributed as a normal distribution with mean θ and standard deviation σ’), U ij ~ N(0, σ bb) the deviation from μ for unit j in set i, R ik ~ N(0, σ run) the deviation of run k in set i and e lijk ~ N(0, σ w) the usual residual error term. Estimates of the variance components represented by σ w and σ bb are desired for homogeneity assessment; the between-run and between-set variance and a test of the between-location difference is also potentially of interest.
The following calculations follow Box, Hunter and Hunter [4] with nomenclature modified for the present paper and with the addition of variance components for the random effects.
Let n S be the number of sets and let n L = n U = n R = n be the number of locations, the number of units per set and the number of runs per set. Then, the ANOVA table is constructed as follows:
Let \( G = \sum\nolimits_{l,i,j,k} {y_{lijk} } \) be the sum of all n S n 2 observations. Further, let
\( T_{{{\text{set}},i}} = \sum\nolimits_{l,j,k} {y_{lijk} } ,\quad i = 1, \ldots ,n_{S} \) be the vector of n S sums of observations for each set.
\( T_{{{\text{loc,}}l}} = \sum\nolimits_{i,j,k} {y_{lijk} } ,\quad l = 1, \ldots ,n_{\text{L}} \) be the vector of n L sums of observations for each location.
\( T_{{{\text{run,}}ik}} = \sum\nolimits_{l,j} {y_{lijk} } ,\quad i = 1, \ldots ,n_{\text{S}} ,\quad k = 1, \ldots ,n_{\text{R}} \) be the vector of n S n R sums of observations for each run within each set.
\( T_{{{\text{unit,}}ij}} = \sum\nolimits_{l,k} {y_{lijk} } ,\quad i = 1, \ldots ,n_{\text{S}} ,\quad j = 1, \ldots ,n_{\text{U}} \) be the vector of n S n R sums of observations for each run within each set.
Then, from
The analysis of variance table is then constructed as shown in Table 2, where M set, M unit, M loc, M run and M w are the ANOVA mean squares corresponding to the set, location, run, unit and residual effects, respectively. Note that the calculation involves subtraction of some potentially large sums of squares, which can result in serious numerical inaccuracy; it is therefore prudent to code the data (that is, to subtract a convenient amount from each value to leave only the digits that show variation) before carrying out the calculation.
Variance components for the different random effects are extracted in the usual way from the mean squares. The estimate s 2w of the residual variance σ 2w estimate is equal to M w. The estimates s 2run and s 2bb of the between-run and between-unit variances are given by
while the estimate s 2S of the between-set variance is given by
Note that because location is a fixed effect common to all sets, the mean square for location does not appear in Eq. 6.
Conventionally, variance estimates that fall below zero are set to zero for simplicity.
Rights and permissions
About this article
Cite this article
Ellison, S.L.R., Grimes, N., Fardus, F. et al. Use of a replicated Latin square design in a homogeneity test for high purity organic melting point standards. Accred Qual Assur 17, 283–290 (2012). https://doi.org/10.1007/s00769-012-0875-5
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s00769-012-0875-5