1,721,061 research outputs found
Comparison of group screening strategies for factorial experiments
No full textFactor screening is an important first step in many industrial experiments where a large number of factors potentially influence a response. The purpose of screening is to identify those few factors which have a substantive influence (that is, are active) and therefore, require further investigation. This paper provides a simulation tool for comparing two-stage group screening strategies where both design and noise factors may be under study. The strategies investigated are classical group screening, in which only main effects are considered at the first stage of the experiment, and an alternative strategy of screening for two-factor interactions as well as main effects.An algorithm is described which allows the user to simulate, and hence to compare, the strategies under different selections of designs and different group sizes for the stage 1 experiment, and for different probabilities of active effects. A detailed example of the use of the algorithm shows how an appropriate strategy can be chosen based on two criteria. These criteria consider the proportion of active factorial effects that are incorrectly screened out at the first-stage experiment, and the average number of observations needed for the entire experiment
Crossover designs in the presence of carry-over effects from two factors.
No full textExperiments, used in the telecommunications industry and elsewhere, are considered that involve the simultaneous application of levels of two unrelated factors, treatments and stimuli, to each of several subjects in a succession of time periods. The existence is suspected of carry-over effects of treatments and stimuli, in the period immediately following the period of their application. Methods are given for the construction of separate sequences of treatments and of stimuli; these methods are based on the Latin squares of Williams and of Russell. In the resulting designs, the treatments and stimuli are either orthogonal or nearly orthogonal, and the coincidence of the direct and carry-over effects of each factor is either balanced or nearly balanced. The efficiencies of the designs are assessed by comparing the average variances of elementary contrasts in the levels of each factor with appropriate lower bounds
Efficient cross-over designs allowing a check on the assumption that direct and carry-over effects do not interact
No full textIn cross-over experiments, each subject receives a sequence of treatments, one in each of a number of consecutive time periods, and a response is measured at the end of each period. Traditional models for these experiments assume that the direct effect of the treatment applied in the current period and the carry¬over effect of the treatment applied in the previous period act additively on the response. Recently, models involving directxcarry-over interaction have been considered and universally optimal designs for investigating such an interaction have been presented. However, these designs require large numbers of subjects and time periods, and such resources are not always available. This paper presents construction methods for designs that permit a test of the interaction, but which require a small number of subjects and periods. The best designs obtained by these methods are given. The designs are also optimal or near-optimal in the absence of directxcarry-over interaction
Screening for dispersion effects by sequential bifurcation
No full textThe mean of the output of interest obtained from a run of a computer simulation model of a system or process often depends on many factors; many times however only a few of these factors are important. Sequential bifurcation is a method that has been considered by several authors for identifying these important factors using as few runs of the simulation model as possible. In this paper, we propose a new sequential bifurcation procedure whose steps use a key stopping rule that can be calculated explicitly, something not available in the best methods previously considered. Moreover we show how this stopping rule can also be easily modified to efficiently identify those factors that are important in influencing the variability rather than the mean of the output. In empirical studies, the new method performs better than previously published fully sequential bifurcation methods in terms of achieving the prescribed Type I error and high power for detecting moderately large effects using fewer replications than earlier methods. To achieve this control for midrange effects, the new method sometimes requires more replications than other methods in the case where there are many very large effect
Use of statistically designed experiments to explore sensitivities in the strength scaling of FRP composites.
No full textDesigning experiments for multi-variable B-spline models
No full textIn a range of practical applications where a response cannot be adequately described by a low order polynomial, B-spline regression models for a single variable have proved useful for prediction. In this paper identifiable models for several explanatory variables are considered which are formulated from B-spline and monomial basis functions of known degree and with specified knots. The use of search methods to find efficient designs under the V-, G- and D-optimality criteria is investigated. Two methods of constructing lists of feasible candidate points are described and compared across a variety of examples
- …
