Goodness of data

The test that is used more frequently than any other to check the goodness of data is the criterion (chi square), or Pearson s x test. The x test is based on the quantity... 

Reliability The "goodness" of data measured by accuracy and precision. 

Unfortunately, good binary data are often not available, and no model, including the modified UNIQUAC equation, is entirely adequate. Therefore, we require a calculation method which allows utilization of some ternary data in the parameter estimation such that the ternary system is well represented. A method toward that end is described in the next section. 

Enthalpies are referred to the ideal vapor. The enthalpy of the real vapor is found from zero-pressure heat capacities and from the virial equation of state for non-associated species or, for vapors containing highly dimerized vapors (e.g. organic acids), from the chemical theory of vapor imperfections, as discussed in Chapter 3. For pure components, liquid-phase enthalpies (relative to the ideal vapor) are found from differentiation of the zero-pressure standard-state fugacities these, in turn, are determined from vapor-pressure data, from vapor-phase corrections and liquid-phase densities. If good experimental data are used to determine the standard-state fugacity, the derivative gives enthalpies of liquids to nearly the same precision as that obtained with calorimetric data, and provides reliable heats of vaporization. 

In many process-design calculations it is not necessary to fit the data to within the experimental uncertainty. Here, economics dictates that a minimum number of adjustable parameters be fitted to scarce data with the best accuracy possible. This compromise between "goodness of fit" and number of parameters requires some method of discriminating between models. One way is to compare the uncertainties in the calculated parameters. An alternative method consists of examination of the residuals for trends and excessive errors when plotted versus other system variables (Draper and Smith, 1966). A more useful quantity for comparison is obtained from the sum of the weighted squared residuals given by Equation (1). 

Appendix C-6 gives parameters for all the condensable binary systems we have here investigated literature references are also given for experimental data. Parameters given are for each set of data analyzed they often reflect in temperature (or pressure) range, number of data points, and experimental accuracy. Best calculated results are usually obtained when the parameters are obtained from experimental data at conditions of temperature, pressure, and composition close to those where the calculations are performed. However, sometimes, if the experimental data at these conditions are of low quality, better calculated results may be obtained with parameters obtained from good experimental data measured at other conditions. 

Neural network classifiers. The neural network or other statistical classifiers impose strong requirements on the data and the inspection, however, when these are fulfilled then good fully automatic classification systems can be developed within a short period of time. This is for example the case if the inspection is a part of a manufacturing process, where the inspected pieces and the possible defect mechanisms are well known and the whole NDT inspection is done in repeatable conditions. In such cases it is possible to collect (or manufacture) as set of defect pieces, which can be used to obtain a training set. There are some commercially available tools (like ICEPAK [Chan, et al., 1988]) which can construct classifiers without any a-priori information, based only on the training sets of data. One has, however, always to remember about the limitations of this technique, otherwise serious misclassifications may go unnoticed. 

A variety of experimental techniques have been employed to research the material of this chapter, many of which we shall not even mention. For example, pressure as well as temperature has been used as an experimental variable to study volume effects. Dielectric constants, indices of refraction, and nuclear magnetic resonsance (NMR) spectra are used, as well as mechanical relaxations, to monitor the onset of the glassy state. X-ray, electron, and neutron diffraction are used to elucidate structure along with electron microscopy. It would take us too far afield to trace all these different techniques and the results obtained from each, so we restrict ourselves to discussing only a few types of experimental data. Our failure to mention all sources of data does not imply that these other techniques have not been employed to good advantage in the study of the topics contained herein. 

Experimental values of X have been tabulated for a number of polymer-solvent systems (4,12). Unfortunately, they often turn out to be concentration and molecular weight dependent, reducing their practical utility. The Flory-Huggins theory quahtatively predicts several phenomena observed in solutions of polymers, including molecular weight effects, but it rarely provides a good quantitative fit of data. Considerable work has been done subsequentiy to modify and improve the theory (15,16). 

Hydroxides. The hydrolysis of uranium has been recendy reviewed (154,165,166), yet as noted in these compilations, studies are ongoing to continue identifying all of the numerous solution species and soHd phases. The very hard uranium(IV) ion hydrolyzes even in fairly strong acid (- 0.1 Af) and the hydrolysis is compHcated by the precipitation of insoluble hydroxides or oxides. There is reasonably good experimental evidence for the formation of the initial hydrolysis product, U(OH) " however, there is no direct evidence for other hydrolysis products such as U(OH) " 2> U(OH)" 2> U(OH)4 (or UO2 2H20). There are substantial amounts of data, particulady from solubiUty experiments, which are consistent with the neutral species U(OH)4 (154,167). It is unknown whether this species is monomeric or polymeric. A new study under reducing conditions in NaCl solution confirms its importance and reports that it is monomeric (168). 8olubihty studies indicate that the anionic species U(OH) , if it exists, is only of minor importance (169). There is limited evidence for polymeric species such as Ug(OH) " 25 (1 4). 

There is a scattered body of data in the literature on ordinary photochemical reactions in the pyrimidine and quinazoline series in most cases the mechanisms are unclear. For example, UV irradiation of 4-aminopyrimidine-5-carbonitrile (109 R=H) in methanolic hydrogen chloride gives the 2,6-dimethyl derivative (109 R = Me) in good yield the 5-aminomethyl analogue is made similarly (68T5861). Another random example is the irradiation of 4,6-diphenylpyrimidine 1-oxide in methanol to give 2-methoxy-4,6-diphenyl-pyrimidine, probably by addition of methanol to an intermediate oxaziridine (110) followed by dehydration (76JCS(P1)1202). 

The chi-square distribution can be applied to other types of apph-catlon which are of an entirely different nature. These include apph-cations which are discussed under Goodness-of-Fit Test and Two-Way Test for Independence of Count Data. In these applications, the mathematical formulation and context are entirely different, but they do result in the same table of values. 

This gives some information about the errors (i.e., the variance and standard deviation of each data point), although the goodness of fit, P, cannot be calculated. 

When considering the availability of nutrients, it is also necessary to examine the significance of nutrient re-use within the waterbody. These internal sources amount not to an additional load, but a multiplier on the recyclability of the same load. This nutrient recycling and the internal stores from which they are recycled are often misunderstood, but there is a dearth of good published data about how these recycling mechanisms operate. Microbial decomposition in the water column is one of several internal loops recognized in recent years, but these are not closed and the flux of nutrients recycled through them is delayed rather than retained. 

---