Statistical Methods, Experimental Design, and Scientific Inference

There is an increasing awareness of the added value that efficient design of experiments can provide to simulation studies. However, many practitioners lack the proper training required to design, execute and analyze a well-designed simulation experiment. In this study, we present the initial stages of a model-driven engineering based tool for managing simulation experiments. Underlying our approach are an experiment ontology and a feature model that capture the statistical design of experiments expert knowledge which users might be lacking. In its current state, the tool provides support for the design and execution of simple experiments. Using a web-based interface, the user is guided through an experiment design wizard that produces an experiment model that can be exported as an XML file containing the experiment's description. This XML can be used to synthesize scripts that can be run by a simulator and shared across different platforms.

As part of the development of a dry powder inhalation system for the treatment of asthma and chronic obstructive pulmonary disease, this work specifically aimed at the systematic, however, cost-effective and efficient development of an inhalation device. Based on theoretical design considerations and an initial inhaler prototype, the concept of a modular inhaler was developed. The modular inhaler was used for the systematic evaluation of the influence of the inhaler's inner dimensions on the resistance to the air flow and the in-vitro deposition characteristics of the inhalation system by using statistical design of experiments and cascade impaction analysis. A reliable statistical model enabled the accurate prediction of the device resistance of any combination of inner dimensions of the inhaler. In conjunction with results from in-vitro deposition studies, this allowed for the definition of optimised inner dimensions of the inhaler to maximise the fine particle fraction and minimise oropharyngeal deposition within the desired range of the inhaler's resistance to air flow. The concept of the modular inhaler and statistical design and evaluation of experiments proved to be important tools for an efficient and successful product development. Eventually, the approaches described and the knowledge obtained enabled the cost-effective development and design of a technically feasible and competitive dry powder inhaler.

The prime object of this book is to put into the hands of research workers, and especially of biologists, the means of applying statistical tests accurately to numerical data accumulated in their own laboratories or available in the literature.

29 - MISCELLANEOUS CONSIDERATIONS REGARDING ANALYSIS OF VARIANCE

Editorial/Letters

Statistical Methods for Research Workers

IN the review of Dr. R. A. Fisher's “Statistical Methods for Research Workers” in NATURE of June 8, the point is made that a careless reader may get the impression that the various methods outlined therein will give exact results when applied to the ordinary small “sample’, although we have, in general, no proof, or even expectation, that the sample is drawn from a ‘normal’ population, to which alone the tables can be exactly applied. That this is so is clear from the fact that an American writer has stated that the English school of statisticians claims to have produced tables which may be used for samples, however small, drawn from any conceivable population; but of course Dr. Fisher would be the first to cry out against the foolishness of making any such claim.

Statistical Methods for Research Workers.

Statistical Methods for Research Workers

Statistical Methods for Research Workers (Seventh Edition).

HereditasVolume 99, Issue 1 p. 151-152 Open Access Maternal effects on recombination in Drosophila melanogaster JACK VALENTIN, JACK VALENTIN Department of Genetics, University of Göteborg, Stigbergsliden 14, S-414 63 Göteborg, SwedenSearch for more papers by this author JACK VALENTIN, JACK VALENTIN Department of Genetics, University of Göteborg, Stigbergsliden 14, S-414 63 Göteborg, SwedenSearch for more papers by this author First published: September 1983 https://doi.org/10.1111/j.1601-5223.1983.tb00740.xAboutPDF ToolsExport citationAdd to favoritesTrack citation ShareShare Give accessShare full text accessShare full-text accessPlease review our Terms and Conditions of Use and check box below to share full-text version of article.I have read and accept the Wiley Online Library Terms and Conditions of UseShareable LinkUse the link below to share a full-text version of this article with your friends and colleagues. Learn more.Copy URL Share a linkShare onFacebookTwitterLinkedInRedditWechat No abstract is available for this article. Literature cited Broadhead, R. S. and Kidwell, J. F. 1975. A note on the distribution of the recombination fraction in Drosophila melanogaster. – J. Hered. 66: 309– 310. Fisher, R. A. 1954. Statistical Methods for Research Workers (12 ed.). – Oliver & Boyd, Edinburgh . Lake, S. 1982. Recombination frequency in temperature shocked reciprocal hybrids of inbred lines of Drosophila melanogaster. – In Advances of Genetics, Development and Evolution of Drosophila ( S. Lakovaara), Plenum Publ. Corp., p. 55– 64. Landner, L. 1971. Genetic recombination in Neurospora crassa affected by changes in the supply of cellular energy. – Hereditas. 69: 315– 319. Luning, K. G. 1981. Genetics of inbred Drosophila melanogaster. I. Induction of marker genes and preliminary recombination tests. –Hereditas. 95: 181– 188. Luning, K. G. 1982. Genetics of inbred Drosophila melanogaster. VIII. The effects of inbreeding on recombination. – Hereditas. 96: 291– 294. Thoday, J. M. and Boam, T. B. 1956. A possible effect of the cytoplasm on recombination in Drosophila melanogaster. – J. Genet. 54: 456– 461. Volume99, Issue1September 1983Pages 151-152 ReferencesRelatedInformation

THE REGRESSION of a set of quantities with unequal frequencies on corresponding equally spaced values of the independent variable may be computed by the ingenious method given in R. A. Fisher's Statistical Methods for Research Workers, Ed. 7, pp. 168-176. This method, in it requires a number of successive additions, involves a great amount of writing. While it is true, as stated on page 170, that much labour is saved by choosing a 'working zero', it is of interest to eliminate all manual labor possible. Since a method making use of orthogonal polynomials is to be preferred to one requiring cumulative frequencies, it is assumed orthogonal polynomials are used in problems with equal frequencies. With unequal frequencies, short series with low degree regressions might possibly be written out without too much effort. With a large number of very long sets of observations, punch card equipment might be employed for obtaining the cumulative frequencies. However, without this equipment a calculating machine, especially one with multiplying keys, may be put to use most advantageously. The following table gives factors may be used with the values of the dependent variable and the frequencies to obtain successive cumulative sums without the intermediate recording is shown in Fisher's tables 30.3 and 30.4. The method of constructing the table is apparent from the numbers themselves. A formal illustration with a simplified case as, for example, with four quantities above the working zero and four below follows. By writing out the successive additions above and below the zero line, quantities are obtained which are seen to be sums of products of the original values and the factors shown in the table.

An attractive feature of many simulation packages is their availability on desktop computers and their potential for allowing the user to run a simulation model under different conditions in a highly interactive way. Such a way of studying a system is attractive because of its immediacy and the direct control it offers the user. Design of Experiments is a statistical technique for quickly optimizing performance of systems. It starts with a screening experimental design test plan involving all of the known factors that are suspected to affect the system’s performance (or output). When the number of input variables or test factors is large, the primary experimental objective is to pare this number down into a manageable few. This is usually followed by another designed experiment design or test plan with the objective of optimizing the system’s performance. For an easy and interactive use of the design of experiments technique, a new tool called DOET (which stands for ‘Design Of Experiments Tool’) has been developed. This paper aims to illustrate the design of experiments technique using the DOET.In this paper we consider the implementation of the methodology of the design of simulation experiments in the form of interactive tool attached to the simulation model called DOET (which stands for ‘Design Of Experiments Tool’). The DOET receives the output from the simulation model and automatically processes this output using the DOE technique. The DOET thus can be used interactively either to give advice to the user conducting the simulation study, or else to directly control the conduct of the runs.

Electrospinning bears great potential for the manufacturing of scaffolds for tissue engineering, consisting of a porous mesh of ultrafine fibers that effectively mimic the extracellular matrix (ECM) and aid in directing stem cell fate. However, for engineering purposes, there is a need to develop material-by-design approaches based on predictive models. In this methodological study, a rational methodology based on statistical design of experiments (DOE) is discussed in detail, yielding heuristic models that capture the linkage between process parameters (Xs) of the electrospinning and scaffold properties (Ys). Five scaffolds made of polycaprolactone are produced according to a 22-factorial combinatorial scheme where two Xs, i.e., flow rate and applied voltage, are varied between two given levels plus a center point. The scaffolds were characterized to measure a set of properties (Ys), i.e., fiber diameter distribution, porosity, wettability, Young’s modulus, and cell adhesion on murine myoblast C1C12 cells. Simple engineering DOE models were obtained for all Ys. Each Y, for example, the biological response, can be used as a driver for the design process, using the process-property model of interest for accurate interpolation within the design domain, enabling a material-by-design strategy and speeding up the product development cycle. The implications are also illustrated in the context of the design of multilayer scaffolds with microstructural gradients and controlled properties of each layer. The possibility of obtaining statistical models correlating between diverse output properties of the scaffolds is highlighted. Noteworthy, the featured DOE approach can be potentially merged with artificial intelligence tools to manage complexity and it is applicable to several fields including 3D printing.

A statistical design of experiments (DOE) approach was used to determine if specific build orientation parameters impacted mechanical strength of fabricated parts. A single platform (10inch by 10 inch cross-section) on the 3D Systems Viper si2 machine was designed to hold 18, ASTM D-638 Type I samples built in six different orientations (called Location) with three samples built for each location. The DOE tested four factors: Location, Position, Axis, and Layout. Each sample within a Location was labeled as Positions 1, 2, or 3 depending on the distance from the center of the platform with Position 1 being the closest to the center. Samples were fabricated parallel with the x-axis, y-axis, or 45 to both axes (called Axis 1, 2, and 3, respectively) and were fabricated either flat or on an edge relative to the x-y plane (called Layout 1 and 2, respectively). The results from the statistical analyses showed that Axis, Location, and Position had no significant effect on UTS or E. However, Layout (or whether a sample was built flat or on an edge) was shown to have a statistically significant effect on UTS and E (at a 95% level of confidence). This result was not expected since a comparison of the average UTS for each Layout showed only a 1.2% difference (6966 psi versus 7050 psi for samples built flat and on an edge, respectively). Because of the small differences in means for UTS, the statistical differences between Layout most likely would not have been identified without performing the DOE. Furthermore, Layout was the only factor that tested different orientations of build layers (or layer-to-layer interfaces) with respect to the sample part, and thus, it appears that the orientation of the build layer with respect to the fabricated part has a significant effect on the resulting mechanical properties. This study represents one of many to follow that is using statistical analyses to identify and classify important fabrication parameters on mechanical properties for layer manufactured parts. Although stereolithography is the focus of this work, the techniques developed here can be applied to any layered manufacturing technology.