Generalization and efficient computation of the box-meyer analysis for orthogonal designs

Box and Meyer (1986) [1] proposed a Bayesian analysis for saturated orthogonal dedigns, based on the widely-used method of examining normal plots of effects estimates. Stephenson, Hulting, and Moore (1989) [5] give an algorithm for computing this analysis, but it can be quite slow for even 2⁵ designs. In this paper we extend the technique to cover all orthogonal factorial designs, rather than just saturated ones, and we show how the computational algorithm can be greatly improved, both in terms of accuracy and speed. With these extensions and improvements the Box-Meyer method becomes viable as a technique for interactive analysis of any orthogonal factorial design, not just small, saturated ones.     

Evaluation of Gaussian orthant probabilities based on orthogonal projections to subspaces

In this paper, a new procedure is described for evaluating the probability that all elements of a normally distributed vector are non-negative, which is called the non-centered orthant probability. This probability is defined by a multivariate integral of the density function. The definition is simple, and this probability arises frequently in statistics because the normal distribution is prevalent. The method for evaluating this probability, however, is not obvious, because applying direct numerical integration is not practical except in low dimensional cases. In the procedure proposed in this paper, the problem is decomposed into sub-problems of lower dimension. Considering the projection onto subspaces, the solutions of the sub-problems can be shared in the evaluation of higher dimensional problems. Thus the sub-problems form a lattice structure. This reduces the computational time from a factorial order, where the interim results are not shared, to order \(p^{2}2^{p}\), which is faster than the procedures that have been reported in the literature.     

Three Parallel Flats Designs for Two-level Factorial Experiments

This paper investigates the properties of the class of three parallel flats designs for two-level factorial experiments. It shows that the designs constructed from this class of designs can have a very simple correlation structure. The correlation of any pair of best linear unbiased estimators of factorial effects is 0, ⅓ or ¼. Furthermore, the designs obtained also have high D-efficiency. Finally, a class of designs is generated with run-size N = 12 to illustrate the use of the theorem.     

Generation of space-filling uniform designs in unit hypercubes

In this paper a new discrepancy measure of uniformity for uniform designs (UDs) in a unit cube is presented. Alternative measures of uniformity based on distance criteria which can be applied to higher dimensions are also discussed. The good lattice point (glp) method was used to construct the uniform designs. Two approaches (generator equivalence and projection) of reducing the computational cost of the glp method are proposed and discussed. Two examples are presented in this paper.     

A generalization of the alias matrix

The investigation of aliases or biases is important for the interpretation of the results from factorial experiments. For two-level fractional factorials this can be facilitated through their group structure. For more general arrays the alias matrix can be used. This tool is traditionally based on the assumption that the error structure is that associated with ordinary least squares. For situations where that is not the case, we provide in this article a generalization of the alias matrix applicable under the generalized least squares assumptions. We also show that for the special case of split plot error structure, the generalized alias matrix simplifies to the ordinary alias matrix.     

Incomplete row-column designs with factorial treatment structure for estimating main effects with full efficiency

In this paper, we propose two methods of constructing row-column designs for factorial experiments. The constructed designs have orthogonal factorial structure with balance and permits estimation of main effects with full efficiency.     

Minimum aberration blocking schemes for 128-run designs

Several criteria have been proposed for ranking blocked fractional factorial designs. For large fractional factorial designs, the most appropriate minimum aberration criterion was one proposed by Cheng and Wu (2002). We justify this assertion and propose a novel construction method to overcome the computational challenge encountered in large fractional factorial designs. Tables of minimum aberration blocked designs are presented for N=128 runs and n=8–64 factors.     

A study of simple unbalanced factorial designs that use type ii and type iii sums of squares

Methods for analysing unbalanced factorial designs can be traced back to the work of Frank Yates in the 1930s . Yet, still today the question on how his methods of fitting constants (Type II) and weighted squares of means (Type III) behave when negligible or insignificant interactions exist, is still unanswered. In this paper, by means of a simulation study, Type II and Type III ANOVA results are examined for all unbalanced structures originating from a 2x3 balanced factorial design within homogeneous groups (design types) accounting for structure, number of observations lost and which cells contained the missing observations. The two level factor is further analysed to test the null hypothesis, for both Type II and Type III analyses, that the unbalanced structures within each design type provide comparable F values. These results are summarised and the conclusion shows that this work agrees with statements made by Yates Burdick and Herr and Shaw and Mitchell-Olds, but there are some results which require further investigation.     

A conditionally-unbiased estimator of population size based on plant-capture in continuous time

This paper proposes an estimator of the unknown size of a target population to which has been added a planted population of known size. The augmented population is observed for a fixed time and individuals are sighted according to independent Poisson processes. These processes may be time-inhomogeneous, but, within each population, the intensity function is the same for all individuals. When the two populations have the same intensity function, known results on factorial series distributions suggest that the proposed estimator is approximately unbiased and provide a useful estimator of standard deviation. Except for short sampling times, computational results confirm that the proposed population-size estimator is nearly unbiased, and indicate that it gives a better overall performance than existing estimators in the literature. The robustness of this performance is investigated in situations in which it cannot be assumed that the behaviour of the plants matches that of individuals from the target population.     

Characterization of information matrices for balanced two-level fractional factorial designs of odd resolution derived from two-symbol simple arrays

We consider a balanced fractional 2^m factorial design of resolution 2?+1 which permits estimation of all factorial effects up through ?-factor interactions under the situation in which all (?+1)-factor and higher order interactions are to be negligible for an integer satisfying [m/2]<lE;?m, where [x] denotes the greatest integer not exceeding x. This paper investigates algebraic structure of the information matrix of such a design derived from a simple array through that of an atomic array. We obtain an explicit expression for the irreducible matrix representation based on the above design and its algebraic properties. The results in this paper will be useful to characterize the designs under consideration.     

A HEURISTIC TECHNIQUE FOR TESTING FOR THE PRESENCE OF INTERACTION IN NONREPLICATED FACTORIAL EXPERIMENTS

Abstract  This paper suggests univariate and multivariate techniques for investigating interaction in nonreplicated factorial experiments. The tests can detect certain types of interaction, but they are not powerful against all possible alternative hypotheses. The two-way factorial experiment is discussed in some detail and an example is used to demonstrate the procedure. The procedure is compared to other tests for interaction. These comparisons show that the procedure can detect some types of interaction which other tests cannot. Likewise other tests can detect interaction this procedure cannot.     

Catalogue of Group Structures for Three-level Fractional Factorial Designs

Taguchi (1959) introduced the concept of split-unit design to sort the factors into different groups depending upon the difficulties involved in changing the levels of factors. Li et al. (1991) renamed it as split-plot design. Chen et al. (1993) have given a catalogue of small designs for two- and three-level fractional factorial designs pertaining to a single type of factors. Aggarwal et al. (1997) have given a catalogue of group structure for two-level fractional factorial designs developed under the concept of split-plot design. In this paper, an algorithm has been developed for generating group structure and possible allocations for various 3^n-k fractional factorial designs.     

On aliasing and generalized defining relationships of factorial arrangements

The concepts of defining contrast (DC), generalized defining relationship (GDR) and aliasing structure (AS) are now well established in the terminology of regression analysis and factorial design theory. There is no complete agreement in the literature about the meaning of regular and irregular fractional factorial designs. This paper provides a workable definition of a regular fraction from a symmetrial prime-powered factorial. It characterizes the uniqueness of the GDR for fractions from the most general factorial. Results are also présentés on the uniqueness of the GDR for regular designs, on orthogonality aspects of regular and irregular designs, and on group-theoretic generation of the complete aliasing structure. Examples are provided to illustrate the developments.     

On sum composition of fractional factorial designs

Combinatorial extension and composition methods have been extensively used in the construction of block designs. One of the composition methods, namely the direct product or Kronecker product method was utilized by Chakravarti [1956] to produce certain types of fractional factorial designs. The present paper shows how the direct sum operation can be utilized in obtaining from initial fractional factorial designs for two separate symmetrical factorials a fractional factorial design for the corresponding asymmetrical factorial. Specifically, we provide some results which are useful in the construction of non-singular fractional factorial designs via the direct sum composition method. In addition a modified direct sum method is discussed and the consequences of imposing orthogonality are explored.     

D‐optimal minimax fractional factorial designs

The D‐optimal minimax criterion is proposed to construct fractional factorial designs. The resulting designs are very efficient, and robust against misspecification of the effects in the linear model. The criterion was first proposed by Wilmut & Zhou (2011); their work is limited to two‐level factorial designs, however. In this paper we extend this criterion to designs with factors having any levels (including mixed levels) and explore several important properties of this criterion. Theoretical results are obtained for construction of fractional factorial designs in general. This minimax criterion is not only scale invariant, but also invariant under level permutations. Moreover, it can be applied to any run size. This is an advantage over some other existing criteria. The Canadian Journal of Statistics 41: 325–340; 2013 © 2013 Statistical Society of Canada     

Factorial hypercube designs for spatial correlation regression

SUMMARY The problem of generating a good experimental design for spatial correlation regression is studied in this paper. The quality of fit generated by random designs, Latin hypercube designs and factorial designs is studied for a particular response surface that arises in inkjet printhead design. These studies indicate that the quality of fit generated by spatial correlation models is highly dependent on the choice of design. A design strategy that we call 'factorial hypercubes' is introduced as a new method. This method can be thought of as an example of a more general class of hybrid designs. The quality of fit generated by these designs is compared with those of other methods. These comparisons indicate a better fit and less numerical problems with factorial hypercubes.     

Graph based isomorph-free generation of two-level regular fractional factorial designs

We provide a new necessary and sufficient check for testing the isomorphism of two 2-level regular fractional factorial designs. The approach is based on modeling fractional factorial designs as bipartite graphs. We employ an efficient canonical graph labeling approach to compare two designs for isomorphism. We then improve upon the existing non-isomorphic fractional factorial design generation algorithm by reducing the number of candidate designs from which isomorphs need to be removed. Not only does our method generate non-isomorphic designs much faster, it is also able to generate designs with run sizes of 2048 and 4096 runs, which were not generated by the existing methods.     

Catalogue of group structures for two-level fractional factorial designs

SUMMARY Taguchi introduced the concept of split-unit design to sort factors into different groups with respect to difficulties involved in changing the levels of factors. Li et al. have developed all possible group structures for eight factors in an L16 orthogonal array for resolution IV with split-plot design. Chen et al. have searched for a best design, according to the various criteria for two-level fractional factorial design and have presented a catalogue. In this paper, we have developed an algorithm for generating group structure and possible allocations for various 2n- k fractional factorial designs that correspond to the designs given by Chen et al.     

The Two-Scale Plot: An Exploratory Display of Data with Heterogeneous Variances

The history of the analysis of unbalanced factorial designs is traced from Yates's original papers (Yates 1933, 1934) to the beginning of the computational revolution in the 1960s. Emphasis is placed on putting the methods proposed during this period in perspective in view of our present understanding.     

Choosing appropriate covariance matrices in a nonparametric analysis of factorials in block designs

The standard nonparametric, rank-based approach to the analysis of dependent data from factorial designs is based on an estimated unstructured (UN) variance–covariance matrix, but the large number of variance–covariance terms in many designs can seriously affect test performance. In a simulation study for a factorial arranged in blocks, we compared estimates of type-I error probability and power based on the UN structure with the estimates obtained with a more parsimonious heterogeneous-compound-symmetry structure (CSH). Although tests based on the UN structure were anti-conservative with small number of factor levels, especially with four or six blocks, they became conservative at higher number of factor levels. Tests based on the CSH structure were anti-conservative, and results did not depend on the number of factor levels. When both tests were anti-conservative, tests based on the CSH structure were less so. Although use of the CSH structure is concluded to be more suitable than use of the UN structure for the small number of blocks typical in agricultural experiments, results suggest that further improvement of test statistics is needed for such situations.