Statistical proper application

Proper application of this framework to any analytical situation encountered can be expected to produce the appropriate statistical database that supports acceptance of said claim by the FDA reviewers. This approach provides evidence of a full exploration of alternatives investigated, their relevance to the situation at hand, and the statistical significance levels attached to each finding. 

Some problems related to the proper application of regression analysis and of other multivariate statistical methods in QSAR studies and concerning the validity of the obtained results have recently been reviewed [403, 408, 409] (compare chapter 4.1). 

The statistical analysis of data requires a proper design of experiments to prove or disprove a certain hypothesis which has been formulated in advance. From the viewpoint of a puritanical statistician most QSAR analyses are forbidden , because they are retrospective studies and, in addition, many different hypotheses (i.e. combinations of independent variables) are tested sequentially. Indeed, many problems arise from the application of regression analysis in ill-conditioned data sets. Only in later stages of lead structure optimization are certain hypotheses, e.g. on the influence of more lipophilic, electronegative, polar, or bulky substituents in a certain position, systematically tested, now fulfilling the requirements for the proper application of statistical methods. 

From a theoretical point of view, the proper application of regression analysis requires the formulation of a working hypothesis, the design of experiments (i.e., compounds to be tested), the selection of a mathematical model, and the test of statistical significance of the obtained result. In QSAR studies, this is pure theory. Reality is different QSAR studies are most often retrospective studies and in several cases many different variables are tested to find out whether some of them, alone or in combination, are able to describe the data. In principle, there are no objections against this method because QSAR equations should be used to derive new hypotheses and to design new experiments, based on these hypotheses. Then the requirements for the application of statistical methods are fulfilled. 

It should be noted that the data collection and conversion effort is not trivial, it is company and plant-specific and requires substantial effort and coordination between intracompany groups. No statistical treatment can make up for inaccurate or incomplete raw data. The keys to valid, high-quality data are thoroughness and quality of personnel training comprehensive procedures for data collection, reduction, handling and protection (from raw records to final failure rates) and the ability to audit and trace the origins of finished data. Finally, the system must be structured and the data must be coded so that they can be located within a well-designed failure rate taxonomy. When done properly, valuable and uniquely applicable failure rate data and equipment reliability information can be obtained. 

There are two statistical assumptions made regarding the valid application of mathematical models used to describe data. The first assumption is that row and column effects are additive. The first assumption is met by the nature of the smdy design, since the regression is a series of X, Y pairs distributed through time. The second assumption is that residuals are independent, random variables, and that they are normally distributed about the mean. Based on the literature, the second assumption is typically ignored when researchers apply equations to describe data. Rather, the correlation coefficient (r) is typically used to determine goodness of fit. However, this approach is not valid for determining whether the function or model properly described the data. 

For nonequilibrium statistical mechanics, the present development of a phase space probability distribution that properly accounts for exchange with a reservoir, thermal or otherwise, is a significant advance. In the linear limit the probability distribution yielded the Green-Kubo theory. From the computational point of view, the nonequilibrium phase space probability distribution provided the basis for the first nonequilibrium Monte Carlo algorithm, and this proved to be not just feasible but actually efficient. Monte Carlo procedures are inherently more mathematically flexible than molecular dynamics, and the development of such a nonequilibrium algorithm opens up many, previously intractable, systems for study. The transition probabilities that form part of the theory likewise include the influence of the reservoir, and they should provide a fecund basis for future theoretical research. The application of the theory to molecular-level problems answers one of the two questions posed in the first paragraph of this conclusion the nonequilibrium Second Law does indeed provide a quantitative basis for the detailed analysis of nonequilibrium problems. 

Before undertaking a discussion of the mathematics involved in the determination of reaction rates is undertaken, it is necessary to point out the importance of proper data acquisition in stability testing. Applications of rate equations and predictions are meaningful only if the data utilized in such processes are collected using valid statistical and analytical procedures. It is beyond the scope of this chapter to discuss the proper statistical treatments and analytical techniques that should be used in a stability study. Some perspectives in these areas can be obtained by reading the comprehensive review by Meites [84], the paper by P. Wessels et al. [85], and the section on statistical considerations in the stability guidelines published by FDA in 1987 [86] and in the more recent Guidance for Industry published in June 1998 [87],... 

Y data. The data set used for this example is from Miller and Miller ([1], p. 106) as shown in Table 58-1. This dataset is used so that the reader may compare the statistics calculated and displayed using the formulas and figures described in this reference with respect to those shown in this series of chapters. The correlation coefficient and other goodness of fit parameters can be properly evaluated using standard statistical tests. The Worksheets provided in this chapter series can be customized for specific applications providing the optimum information for particular method comparisons and validation studies. 

In practical application [117] a problem in statistical mechanics is approached by setting up the proper ensemble to represent the available information about the state of the system. The average values of various quantities can then be determined by averaging over the ensemble. 

The proper conduct of complex exposure studies requires that the quality of the data be well defined and the statistical basis be sufficient to support rule making if necessary. These requirements, from study design through chemical analysis to data reduction and interpretation, focused our attention on the application of chemometric techniques to environmental problems. 

The proper evaluation and assessment of the calculated or graphically determined values of the kinetic parameters requires the application of statistical analysis . This is also true when looking for possible patterns in the various plots (e.g., parallel lines V5-. intersecting lines). When reporting kinetic values, the error limits should always be provided. Programs are available that statistically evaluates kinetic data. See Statistics A Primer)... 

In many cases it may be impossible to separate fluctuations in ambient HO from background fluctuations or other sources of noise. The proper uncertainty in the net HO signal, averaged over any chosen time interval, is the standard error of the mean calculated from standard statistical formulas from the net data only. This calculation is independent of the applicability... 

In industrial plants, large numbers of process variables must be maintained within specified limits in order for the plant to operate properly. Excursions of key variables beyond these limits can have significant consequences for plant safety, the environment, product quality and plant profitability. Statistical process control (SPC), also called statistical quality control (SQC), involves the application of statistical techniques to determine whether a process is operating normally or abnormally. Thus, SPC is a process monitoring technique that relies on quality control charts to monitor measured variables, especially product quality. 

Furthermore, applications in the case of neutral atoms are more difficult because of the lack of electrostatic moments in the atom for describing the interaction with the environment. A proper treatment of liquid systems should consider its statistical nature [6, 7] as there are many possible geometrical arrangements accessible to the system at nonzero temperature. Thus, liquid properties are best described by a statistical distribution [8-11], and all properties are obtained from statistical averaging over ensembles. Thus, in this direction, it is important to use statistical mechanics, with some sort of computer simulation of liquids [6, 7], combined with quantum mechanics to obtain the electronic property of interest. 

In order to evaluate properly the transition from the older formulation of kinetic theory to the still sketchy statistical formulation we need a comparison with other branches of science in which the methods of probability theory have found an application. All these show a similar process of development. 

The more concentrated is the system, the better statistics must be involved. Here (for higher electrolyte concentrations) a family of HNC approximation approximations may be mentioned, which demonstrated very good agreement with the Monte Carlo results. They have also been checked with experiments on the interaction between two mica sheets in electrolytes (for CaCb electrolyte see [37]), and found a variety of applications. For example, a proper account of the ion distributions allows to explain such phenomenon as a clay swelling in the presence of an electrolyte, while the standard DLYO description fails [36]. Also these ideas have been utilized when studying different biological systems [38-40]. 

The number m of observations exceeds (usually by an order of magnitude or more) the number n of adjustable parameters. Thus the mathematical problem is overdetermined. (For a situation in which m= n, the corresponding set of simultaneous equations could in principle be solved exactly for the parameters aj. But this exact fit to the data would provide no test at all of the validity of the model.) In least squares as properly applied, the number of observations is made large compared to the number of parameters in order (1) to sample adequately a domain of respectable size for testing the validity of the model, (2) to increase the accuracy and precision of the parameter determinations, and (3) to obtain statistical information as to the quality of the parameter determination and the applicability of the model. 

Another important recent contribution is the provision of a good measurement of the precision of estimated reactivity ratios. The calculation of independent standard deviations for each reactivity ratio obtained by linear least squares fitting to linear forms of the differential copolymer equations is invalid, because the two reactivity ratios are not statistically independent. Information about the precision of reactivity ratios that are determined jointly is properly conveyed by specification of joint confidence limits within which the true values can be assumed to coexist. This is represented as a closed curve in a plot of r and r2- Standard statistical techniques for such computations are impossible or too cumbersome for application to binary copolymerization data in the usual absence of estimates of reliability of the values of monomer feed and copolymer composition data. Both the nonlinear least squares and the EVM calculations provide computer-assisted estimates of such joint confidence loops [15]. 

Multiple detection 2+ detectors, proper calibrants No segment-segment interactions, No neighbor-group eflFects Bulk composition and composition distribution. Broad applicability, Intrinsic generation of calibration curve. No additional sample preparation work Statistical copolymers. Densely grafted chains 2, 3, 5, 8... 

---