Precision of data

The total stream and individual component flows do not normally need to be shown to a high precision on the process flow-sheet at most one decimal place is all that is usually justified by the accuracy of the flow-sheet calculations, and is sufficient. The flows should, however, balance to within the precision shown. If a stream or component flow is so small that it is less than the precision used for the larger flows, it can be shown to a greater number of places, if its accuracy justifies this and the information is required. Imprecise small flows are best shown as TRACE . If the composition of a trace component is specified as a process constraint, as, say, for an effluent stream or product quality specification, it can be shown in parts per million, ppm. 

A trace quantity should not be shown as zero, or the space in the tabulation left blank, unless the process designer is sure that it has no significance. Trace quantities can be important. Only a trace of an impurity is needed to poison a catalyst, and trace quantities 

---

It may be mentioned in passing that the volume, and quantitative precision, of data available in this field owes much to the use of gas/ liquid chromatography for the rapid, and accurate, quantitative analysis of alkene mixtures. 

We have made references in the foregoing discussion to the precision of data, or how precise the data are. We have also made reference to the accuracy of data. Precision refers to the repeatability of a measurement. If you repeat a given measurement over and over and these measurements deviate only slightly from one another, within the limits of the number of significant figures obtainable, then we say that the data are precise, or that the results exhibit a high degree of precision. The mean of such data may or may not represent the real value of that parameter. In other words, it may not be accurate. Accuracy refers to the correctness of a measurement, or how close it comes to the correct value of a parameter. 

The report of any measurement must always include enough information to avoid misunderstanding and should contain specific details about the sample as indicated in Table 1.10. In reporting the precision of data, a value for the standard deviation implies that the result is the mean of replicate analyses of... 

It is partly the fault of statistics that experimenters have misconstrued the value of the number and precision of data points relative to the value of the location of the points. The importance of the location of the data in the model specification stage can be seen from Fig. 1, which represents literature data (M3) on sulfur dioxide oxidation. The dashed and solid lines represent the predicted rates of two rival models, and the points are the results of two series of experimental runs. It can be seen that neither a greater number of experimental points nor data of greater precision will be of major assistance in discriminating between the two rival models, if data are restricted to the total pressure range from 2 to 10 atm. These data simply do not place the models in jeopardy, as would data below 2 atm and greater than 10 atm total pressure. This is presumably the problem in the water-gas shift reaction, which is classical in terms of the number of models proposed, each of which adequately represent given sets of data. 

This abbreviation refers to quality assurance/quality control. QA/QC procedures are standardized methods used to verify the quality (accuracy and precision) of data. 

Perhaps some day the precision of data will lead to the point that a geometer can calculate in his study the phenomena of any chemical combination, in the same manner that he calculates the movement of the celestial bodies. 

Although it was Daltons atomic hypothesis that first achieved the idealized goal of a true chemical calculus, the effort to acquire the requisite precision of data called for by Lavoisier had made substantial progress before 1808. Indeed without that accumulation of the empirical data, Daltons hypothesis would have carried no persuasion at all in a climate still dominated by an inductivist epistemology. 

D. G. Larsen and P. R. Rony, "Computer Interfacing Precision of Data Transmission Analog Versus Digital Data," Am. Lab., 6, 67 (June 1974). 

The primary screen - up to thousands of reactions per day typically on a microliter scale - provides information about potential hits and serves as a tool to rule out quickly catalyst formulations that perform poorly. The primary screening filter should be suitably high-throughput to cover a selected parameter space rapidly, although to achieve this it is often necessary to tolerate certain compromises regarding the precision of data from such a screen. 

A discussion of the accuracy and precision of data has been added to Chapter 1. 

In summary with this system it has been possible to observe the transition from the osmotic to a salted and then to a collapsed brush. One might suspect that these transitions are also reflected in the pressure/area isotherms. However, we have not yet been able to relate the breaks in the isotherm slopes with parameters deduced from X-ray reflectivity. This would require an even higher precision of data analysis, and precision in density determination better than 1% is difficult to achieve. We realize that the transition at nc is accompanied by a change in relaxation times as expected for a... 

Precision is a quantitative measure of the random variation between a series of measurements from a method. Approaches for determining precision of data from the screening assay (assay response), specificity confirmation assay (percent inhibition), and titration assay (titers) are described below. 

The precision of data is often expressed in terms of the variance (r/ ). which is the square of the sian-dard deviation. Kor independent sources of random error in a system, variances arc often additive. That is, if there are n independent sources, the total variance c is given by... 

This example demonstrates that even for a deterministic system without any external loading, the response appears to be uncertain. In the real world, there are many types of unmodeled behavior/dynamics of complex physical phenomena (e.g., chaotic systems) and one possible approach is to treat them as random variables or random processes. Then, statistical moments are used to represent the overall behavior. This type of error is regarded hereafter as a modeling error. Another main source of uncertainty is due to the finite amount of information carried by the data. Due to the finite amount of the measurement, and hence the finite amount of information, identification results can be determined up to finite precision so uncertainty gets into the picture. Finally, due to the finite precision of data acquisition, measurement error is induced, including electrical noise and quantization error. 

The accuracy and precision of data from any technique is of vital importance. The experimental results would be accurate if they show their proximity to the true values while the precision is connected with the reproducibility of the results under repeated measurements. 

---