Probability plot techniques

This section will cover the potential difficulties that may be encountered when using hazard and probability plotting paper. It will also look at how to use the plotting paper for the most effective results. Much of the discussion applies equally to both hazard or probability plotting, especially where good plotting techniques are concerned. 

Normal Probability Plot The normal probability plot is a graphical technique for assessing whether or not a data set is approximately normally distributed. The data are plotted against a theoretical normal distribution in such a way that the points form an approximate straight line. Departures from this straight line indicate departures from normality. The normal probability plot is important for quality process improvement since many other tools require the normality assumption. A normal... 

Ideally, at the end of a determination that has made use of hundreds or thousands of observations and is thus amenable to statistical analysis at a reasonably high level, one should not merely take comfort in satisfactory values of parameters such as R and ax but rather undertake an overall statistical analysis of the residuals to ascertain whether the uncertainties in the data were properly assessed and whether there is an opportunity to detect errors in the model or systematic errors in the data. In the past this has rarely been done, in part because it is a lot of work and in part because most crystallographers do not really know how to go about it. Recently, Abrahams and Keve17 (later also Hamilton and Abrahams18 ) showed that the objectives sought can be largely achieved by use of an (normal probability plot. In the application of this technique to the problem at hand, the normalized residual... 

A number of statistical techniques exist for assessing whether a given distribution departs from normality. Historically, the most commonly used method in crystallography is the normal probability plot [29, 30]. In this procedure, the n observations in a sample are normalised to zero mean and unit variance, and then ranked. The resulting ranked quantities are plotted against the order statistics expected for a random sample of size n taken from the standard normal distribution. Departures from a straight-line plot indicate non-normality. Various type of departure (e.g. bowed lines or S-shaped curves) are characteristic of particular sorts of deviations from normality. 

Constant Crystal Growth Model. In this instance, crystals have an inherent constant growth rate, but the rate from crystal-to-crystal varies. The modeling of this phenomenon must be accomplished by use of probability transform techniques due to the presence of a growth rate distribution. The complete solution for the population density yields a semilogarithmic population density plot that is concave upwards similar to size-dependent growth (Berglund and Larson 1984). Since it is relatively difficult to handle, a moment approximation was developed for an MSMPR crystallizer (Larson et al. 1985). 

When performing strength testing, it is critical that the flaws leading to failure in the test specimens are the same as those that cause failure in service. Spurious results will be obtained, for example, if flaws are produced in the machining of test specimens. Additionally, flaws in the component may appear during service (e.g., by oxidation or corrosion) and these may be different than those in test specimens. In some cases, several types of flaws may be present in the same specimen and this leads to non-linearities in the Weibull plots. Techniques have been put forward for extracting the Weibull parameters from such plots. In many situations, Weibull plots need to be extrapolated to failure probability values well... 

In the first instance, when the results were analyzed by simple mean and standard deviation analysis, Amico et al. [16-18] got large relative standard deviatiOTi, indicating limitatimi of this method for the proper characterizatiOTi of the diameter. Then, they used Weibull probability density and cumulative distribution functions [20,56,58] to estimate two parameters, the characteristic life and a dimensionless positive pure number, which were supposed to determine the shape and scale of the distribution curve. For this, they adopted two methods, the maximum likelihood technique, which requires the solution of two nonlinear equations, and the analytical method using the probability plot as mentioned earlier for coir fibers. 

Of the four electrochemical techniques discussed in this chapter, the errors of the Tafel-plot technique have been the least evaluated, in the sense that every reported error analysis is only qualitative. Consequently, it is difficult to compare this technique to the others. The expected errors may be larger, under some conditions, than for the other techniques, but this is, at least partially, compensated by the fact that the error will often be accompanied by a distortion of the Tafel line (deviation from linearity) that will alert the user to the problem. Considering all the disadvantages enumerated in Section II.2, it is probably the least desirable alternative among the four techniques discussed. 

Half normal plots give a visual indication of which factors are statistically significant. The technique is useful when there are few or no degrees of freedom available for a residual mean square in ANOVA or regression. Half normal plots are therefore useful when the experimental design is saturated and all effects are of interest. The half normal plot is a type of probability plot where a numerical value for the factor effects is plotted on the vertical axis against the expected normal order statistics on the horizontal axis (see Grove and Davis, 1997). 

The main problem in Eas0 vs. correlations is that the two experimental quantities are as a rule measured in different laboratories with different techniques. In view of the sensitivity of both parameters to the surface state of the metal, their uncertainties can in principle result of the same order of magnitude as AX between two metals. On the other hand, it is rare that the same laboratory is equipped for measuring both single-crystal face is not followed by a check of its perfection by means of appropriate spectroscopic techniques. In these cases we actually have nominal single-crystal faces. This is probably the reason for the observation of some discrepancies between differently prepared samples with the same nominal surface structure. Fortunately, there have been a few cases in which both Ea=0 and 0 have been measured in the same laboratory these will be examined later. Such measurements have enabled the resolution of controversies that have long persisted because of the basic criticism of Eazm0 vs. 0 plots. 

Room-temperature fluorescence (RTF) has been used to determine the emission characteristics of a wide variety of materials relative to the wavelengths of selected Fraunhofer lines in support of the Fraunhofer luminescence detector remote-sensing instrument. RTF techniques are now used in the compilation of excitation-emission-matrix (EEM) fluorescence "signatures" of materials. The spectral data are collected with a Perkin-Elraer MPF-44B Fluorescence Spectrometer interfaced to an Apple 11+ personal computer. EEM fluorescence data can be displayed as 3-D perspective plots, contour plots, or "color-contour" images. The integrated intensity for selected Fraunhofer lines can also be directly extracted from the EEM data rather than being collected with a separate procedure. Fluorescence, chemical, and mineralogical data will be statistically analyzed to determine the probable physical and/or chemical causes of the fluorescence. 

If 2-D NMR techniques are really useful then 3-D ones must be even more so... shouldn t they A number of 3-D experiments have been devised which are in fact, produced by merging two, 2-D experiments together. The results could never be plotted in true 3-D format since etching them into an A3-sized block of glass would not be practical and viewing them as some sort of holographic projection, would probably not be feasible In essence, 3-D spectra have to be viewed as slices through the block which effectively yield a series of 2-D experiments. It is possible to combine techniques to yield experiments such as the HMQC-COSY and the HSQC-TOCSY. 

The first is to normalize the data, making them suitable for analysis by our most common parametric techniques such as analysis of variance ANOYA. A simple test of whether a selected transformation will yield a distribution of data which satisfies the underlying assumptions for ANOYA is to plot the cumulative distribution of samples on probability paper (that is a commercially available paper which has the probability function scale as one axis). One can then alter the scale of the second axis (that is, the axis other than the one which is on a probability scale) from linear to any other (logarithmic, reciprocal, square root, etc.) and see if a previously curved line indicating a skewed distribution becomes linear to indicate normality. The slope of the transformed line gives us an estimate of the standard deviation. If... 

Quantification of ORD and CD Data. In principle ORD and CD can be used to calculate the amounts of a, / , and random conformations in protein, but in practice such estimates are subject to large errors. The Moffitt-Yang plot is probably the best estimate of percentage of a-helicity, but it is unable to distinguish between the ft and random structures. A detailed analysis of CD bands and their resultant Cotton effects, combined with infrared data, is the most promising approach even here the limits of error are large (82). Traditional estimates have been based on combinations of a-helix and random coil, and attention has been centered upon estimation of helical content. Consideration of j3 structure has been introduced more recently. The technique must be calibrated empirically with synthetic polypeptides of known conformation, and the proper choice of reference is not obvious. The /3 structure seems to be particularly variable in its rotational properties (27, 82). 

A presentation of the final calculated "curve" for the vanadium values, plotted against nickel (age) is shown in Figure 11. Although we have admittedly taken more liberties with our empirical fit than the probable data quality permits, we have intended an example of the types of analysis that become possible using the combination of modern image processing techniques and... 

---