FACTORS INFLUENCING THE EFFICIENCY OF INCOMPLETE BLOCK DESIGNS IN SENSORY EVALUATION EXPERIMENTS

The efficiencies of incomplete block designs were investigated by comparing two hundred and twenty eight analyses from eleven trials using hedonic scales with corresponding randomized complete block analyses. Of the ten explanatory factors examined, only the panelist, the product type, the number of samples per session and the average score of the data had an effect on the efficiency of incomplete block designs. The effect of product type was attributed to influences of produce consumed outside the trial, and the effect of the data mean reflected decreased conscientiousness with products the panelists disliked. With three and four samples per session, incomplete block designs were 31 % and 2 % more efficient, respectively, than randomized complete block designs. When five or more samples were tested, the incomplete block designs were markedly less efficient. The practical implications of all these effects on experimental design are discussed.     

Dual Designs and Their Application in Genetical Experiments

The dual of incomplete block designs has been studied with th́eir applications in genetical experiments. Partial diallel crosses (PDC) of type I have been constructed using balanced incomplete block (BIB) designs, partially balanced incomplete block (PBIB) designs and their dual designs. Simplified analysis of PDC has been presented using the dual property of these designs. List of optimal PDC having simple analysis has been given.     

Construction of Partially Balanced Incomplete Block Designs with Nested Rows and Columns

A number of methods of construction of partially balanced incomplete block designs with nested rows and columns are developed and new balanced incomplete block designs with nested rows and columns are obtained as a by-product.     

A Simple Approach to Analyse Non-Orthogonal Data

The present paper deals with an alternative and simple procedure to analyse the non-orthogonal data. The procedure is general in nature but has some advantages for the non-orthogonal data due to some missing observations. The procedure is applied to (i) two way classification with unequal number of observations per cell; (ii) randomized block designs with some missing observations and (iii) balanced incomplete block designs and also illustrated with the help of numerical examples.     

Construction of Lattice Square Designs

Khare and Federer (1981) presented a simple method for constructing incomplete block designs for any number of treatments. Their procedure is extended to constructing lattice square designs. Using variety cutting, lattice square designs are available for any number of treatments.     

Analysis of Partial Diallel Crosses in Incomplete Blocks

With a large number of lines in a diallel cross experiment, the number of crosses becomes unmanageable to be accommodated in homogeneous blocks. To overcome this problem, a sample of crosses, known as partial diallel cross (PDC) is often used. The selection of a PDC is based on the criterion of high efficiency for the estimation of general combining ability (gca) effects. Even with a moderately large number of crosses, the use of incomplete blocks is necessary to obtain homogeneous experimental units. The analysis of data from a general PDC grown in general incomplete block designs is being described. An iterative scheme is being developed for obtaining a generalized inverse of the information matrix used in estimating gca effects. Properties such as connectedness and efficiency of mating designs embedded in environment designs are being examined. The paper also examines the universal optimality of some designs in a class of designs. An illustration of the numerical procedure is also presented.     

A Simple Procedure for Constructing Experiment Designs with Incomplete Blocks of Sizes 2 and 3

In order to maximize control of heterogeneity within complete blocks, an experimenter could use incomplete blocks of size k = 2 or 3. In certain situations, incomplete blocks of this nature would eliminate the need for such spatial types of analyses as nearest neighbor. The intrablock efficiency factors for such designs are relatively low. However, with recovery of interblock information, FEDERER and SPEED (1987) have presented measures of design efficiency factors which demonstrate that efficiency factors approach unity for certain ratios of the intrablock and interblock variance components. Hence with recovery of interblock information, even incomplete block designs with k = 2 or 3 have relatively high efficiency factors. The reduction in the intrablock error variance over the complete block error variance in many situations will provide designs with high efficiency. A simple procedure for constructing incomplete blocks of sizes 2 and 3 is presented. It is shown how to obtain additional zero-one association confounding arrangements when v = 4 t, t an integer, and for v = pk, k ≤ p. It is indicated how to do the statistical analysis for these designs.     

Construction of Resolvable Designs with Nested Treatment Structure

The construction of resolvable incomplete block and row-column designs is investigated when the treatments have a nested structure. Some theoretical results are derived for lattice designs. Efficient designs for unequal-sized treatment groups are obtained by defining a multiple objective function and carrying out a computer search using an interchange algorithm.     

New Constructions of BIB Designs

A method of construction of certain balanced incomplete block (BIB) designs is defined from which we get new series of BIB designs.     

Fractional Plans to Estimate Main Effects and Two Factor Interactions Inclusive of a Specific Factor

Balanced incomplete block designs are used to construct non‐geometric 2ⁿ fractional factorial plans to estimate all n main effects and n – 1, 2 factor interactions with a specific factor included in each interaction. When the balanced incomplete block design is of Family (A), the resulting fractional factorial plan has the same number of runs as a fold‐over Hadamard matrix giving same variances for the estimates; however, some new designs are shown to be non‐isomorphic to the fold‐over Hadamard matrix plans.     

Small Sample Properties of Rank Tests for Incomplete Unbalanced Designs

A rank test is presented for analysis of incomplete unbalanced designs, i.e. for designs that may have been originally planned to be either balanced or unbalanced and where some observations may be missing at random. This test is a modification of the procedure of Benard and van El-teren (1953) based on a generalization of block weights proposed by Prentice (1979). It is compared with the tests of Haux, Schumacher, and Weckesser (1984) and Rai (1987). For incomplete or unbalanced designs with more than two treatments the quadratic forms proposed by these authors are proven to be invalid for small sample sizes, except for special cases. A necessary condition is given for test statistics to be valid also for small samples.     

A Simple Construction Procedure for Eesolvable Incomplete Block Designs for any Number of Treatments

A simple, straightforward procedure, which requires no special tables or generators, is presented for constructing resolvable incomplete block designs for v=pk, v=p²k, …, treatments, for k≥p, in incomplete blocks of size k. Also, it is shown, how to obtain incomplete block designs for any v in blocks of size k and k+1. The procedure allows construction of balanced incomplete block designs for p = k a prime number. For p = n not a prime number, incomplete block designs can be obtained by the procedure, but are not balanced. However, for p_s being the smallest prime factor of n, p_s + 1 for v = n², p_s²+ p_s + 1 for v = n³, …, arrangements can be obtained for which the occurrence of any treatment pair in the blocks is either zero or one. This is called a zero-one concurrence design. Procedures are described for obtaining additional zero-one concurrence arrangements. It is shown that the efficiency of these designs is maximum. Both intra-block and inter-block analyses are described.     

Construction of resolvable spatial row-column designs

Resolvable row-column designs are widely used in field trials to control variation and improve the precision of treatment comparisons. Further gains can often be made by using a spatial model or a combination of spatial and incomplete blocking components. Martin, Eccleston, and Gleeson presented some general principles for the construction of robust spatial block designs which were addressed by spatial designs based on the linear variance (LV) model. In this article we define the two-dimensional form of the LV model and investigate extensions of the Martin et al. principles for the construction of resolvable spatial row-column designs. The computer construction of efficient spatial designs is discussed and some comparisons made with designs constructed assuming an autoregressive variance structure.     

Characterisation of Certain Split-Block Designs with a Control

The paper deals with an incomplete split-block design in which one or two factors are split into two parts, the first one containing test treatments and the second one — a control treatment. The aim of the experiment carried out in such design is to compare the average of the test treatment effects with the control treatment effect and the test treatment effects with the control treatment effect individually. All those comparisons can be expressed by contrasts, elementary or basic ones. The aim of the paper is to characterise particular cases of the incomplete split-block designs with respect to general balance and to efficiency factors of the design with respect to the contrasts. In the paper we restrict our attention to the designs in which each block has two rows or/and two columns only. These designs are suitable for certain agricultural experiments.     

On the PBB Designs for Multiresponse Experiments

We present the idea of using multiresponse incomplete block designs when not all responses can be observed in all experimental units. For a special class of such designs, in which partial designs are PBB designs, a method for estimating natural treatment contrast is given. We also consider the problem of testing the hypotheses concerning the natural and any estimable treatment contrasts. For testing this hypothesis the Wald statistics, being asymptotically chi-square distributed, is proposed.     

Group screening for rare events based on incomplete block designs

Fields such as, diagnostic testing, biotherapeutics, drug development, and toxicology among others, center on the premise of searching through many specimens for a rare event. Scientists in the business of “searching for a needle in a haystack” may greatly benefit from the use of group screening design strategies. Group screening, where specimens are composited into pools with each pool being tested for the presence of the event, can be much more cost-efficient than testing each individual specimen. A number of group screening designs have been proposed in the literature. Incomplete block screening designs are described here and compared with other group screening designs. It is shown under certain conditions, that incomplete block screening designs can provide nearly a 90% cost saving compared to other group screening designs such as when prevalence is 0.001 and screening 3876 specimens with an ICB-sequential design vs. a Dorfman design. In other cases, previous group screening designs are shown to be most efficient. Overall, when prevalence is small (≤0.05) group screening designs are shown to be quite cost effective at screening a large number of specimens and in general there is no one design that is best in all situations. © 2018 American Institute of Chemical Engineers Biotechnol Progress, 35: e2770, 2019.     

Planning incomplete block experiments when treatments are genetically related

Selection trials in plant and animal breeding, in incomplete blocks, are described by linear models with random effect parameters associated with treatments with known genetic covariance structure. It is now well known that the information on relatives can improve the analysis and many extensions of this model have been proposed, but no studies have been done on the consequences of this genetical relatedness among treatments for the optimality of block designs. Using a suitable optimality criterion, we show that the knowledge on relatedness may imply that the optimal design is not in the class of designs which are optimal for unrelated treatments. Implications for practical applications are discussed.     

Efficient two-sample designs for microarray experiments with biological replications

In the last years, biostatistical research has begun to apply linear models and design theory to develop efficient experimental designs and analysis tools for gene expression microarray data. With two-colour microarrays, direct comparisons of RNA-targets are possible and lead to incomplete block designs. In this setting, efficient designs for simple and factorial microarray experiments have mainly been proposed for technical replicates. But for biological replicates, which are crucial to obtain inference that can be generalised to a biological population, this question has only been discussed recently and is not fully solved yet. In this paper, we propose efficient designs for independent two-sample experiments using two-colour microarrays enabling biologists to measure their biological random samples in an efficient manner to draw generalisable conclusions. We give advice for experimental situations with differing group sizes and show the impact of different designs on the variance and degrees of freedom of the test statistics. The designs proposed in this paper can be evaluated using SAS PROC MIXED or S+/R lme.     

On the Construction of Balanced and Partially Balanced Ternary Design

A method of constructing balanced and partially balanced ternary designs from balanced and partially balanced incomplete block designs, respectively, and two methods of constructing partially balanced ternary designs from association schemes are obtained. Two new and efficient balanced ternary designs having K < V and R ≦ 20 are obtained by the first method.     

The potential use of artificially produced monozygotic twins for comparative experiments

The uniformity of twins has been examined by assembling estimates of the intraclass correlation coefficient (rho(I)) available in the literature for a variety of parameters studied in cattle monozygotic twins and human dizygotic and monozygotic twins. The values of rho(I) vary considerably between parameters. In human monozygotic twins rho(I) is always larger compared to that found in dizygotic twins. There is insufficient evidence to determine whether artificial monozygotic twins are more uniform than natural monozygotic twins. A new measure of twin uniformity, given by T (3) = 1 (1-rho (I)) , is introduced. In practice 2T(3) gives the number of animals chosen at random that one member of a twin pair can replace without loss of statistical efficiency. A useful class of experimental designs for the exploitation of twin uniformity is incomplete block designs. These designs are defined by (v, k, b), where v is the number of treatments to be compared, k = 2, and b is the number of twin pairs. Each design has an associated efficiency (E). Provided rho(I)>1-E, an incomplete block design will be advantageous. In general, when only a few twin pairs are available, this relation will only hold for monozygotic and not dizygotic twins. Suitable arrangements of treatment comparisons for designs (3,2,8), (4,2,9), (5,2,10), (6,2,11) are presented.