Nonestimability of missing values for $2^K$ and $3^K$Factoroial Designs

  • Jung W. Sim (Department of Computer Science and Statistics, Chonnam National University, Kwangju 500) ;
  • Park, Sung H. (Department of Computer Science and Statistics, Seoul National University, Seoul 151)
  • Published : 1984.06.01

Abstract

A method of missing value estimation for a general design is descrived. In particular, the cases of missing value estimation for $2^k$ and $3^k$ design are explored and discussed. Some examples are illustrated to show the missing value estimation and the nonestimatimable cases.

Keywords

References

  1. Technometrics v.12 no.3 Missing Values in Multiresponse Nonlinear Model Fitting Box, M.J.;Draper, N.R.;Hunter, W.G.
  2. Technometrics v.3 no.3 Missing Values in Response Surface Designs Draper, N.R.
  3. Applied Statistics v.5 Missing Values in Experiments Analysised on automatic computers Healy, M.J.;Westmacott, M.H.
  4. Technometrics v.13 no.4 Iterative Procedures for Missing Values in Experiments Analysised on automatic computers Preece, D.A.
  5. Applied Statistics v.22 Missing Data in Quantitatipe Designs Shearer, P.R.
  6. J. Roy. Statist. Soc. Ser. B v.14 The Design and Analysis of Block Experiments Tocher, K.D.
  7. Biometrics v.14 Estimation of Missing Values for the Analysis of Incomplete Data Wilkinson, G.N.