Referee

Because the production of magnesium is a large-scale iadustrial process, fast and rehable methods for magnesium analysis have been developed for the quick turnaround times necessary ia a production foundry (76,77). Referee methods which are more time consuming but have larger ranges and greater accuracy compared to the production methods have also been developed (78). 

Referee Methods. The American Society for Testing Materials (ASTM) has collected a series of standard referee methods for the analysis of magnesium and its alloys (78). These methods are accurate over a larger range of concentration than the production methods, but are time consuming ia thek apphcation. The methods are based on potentiometric titration, photometric methods, or gravimetric methods. The photometric methods are most common and are relatively straightforward. 

The bulk of the patent specification is the disclosure, the text and illustrations that describe the claimed invention in detail and explain how the claimed invention differs from the prior art. Modem patent disclosures contain a summary of the claimed invention, a description of the background of the invention, a general description of the way in which the invention is made and used, specific examples, and, where appHcable, drawings of the invention in general or specific embodiments. The technical information provided in a patent specification may be used without infringing the patent only practicing the invention defined in the claims within the term and territory of the patent grant is forbidden. Because much of the information in patent specifications is never pubHshed in refereed journals or other nonpatent media, patent disclosures are an invaluable part of the technical Hterature. 

Thin-film oven test may be used but the rolling thin-film oven test shall be the referee method. 

The test methods indicated are the approved referee methods. Other acceptable methods are indicated in Section 2 and 5.1. 

Bad laboratory analyses are not always the fault of the laboratory.. Sampling plays a big role. One plant superintendent investigated every instance of suspect analyses in his plant using elaborate around-the-clock methods over a considerable period. His results revealed that over one half of the bad analyses were not the fault of the laboratory. We are all human and bad analyses will result from time to time. Rather than resubmit samples, it may be well to spend a few minutes using the following methods as referees to evaluate the reasonableness of the results. 

The basis for evaluation of piston speed varies throughout industry. This indicates that the subject is spiced with as much emotion as technical basics. An attempt to sort out the fundamentals will be made. First, because there are so many configurations and forms of the reciprocating compressor, it would appear logical that there is no one piston speed limit that will apply across the board to all machines. The manufacturer is at odds with the user because he would like to keep the speed up to keep the size of the compressor down, while the user would like to keep the speed down for reliability purposes. As is true for so many other cases, the referee is the economics. An obvious reason to limit the speed is maintenance... 

We conclude this section with an amusing cautionary tale which illustrates the type of blunder that can still appear in the pages of a refereed journal (1975) when scientists (in this... 

Though he won the prize he did not win the outright acclaim of his referees. They accepted that Fourier had formulated heat flow correctly but felt that his methods were not without their difficulties. The use of the Fourier Series was still controversial. It was only when he had returned to Paris for good (around 1818) that he could get his work published in his seminal book, The Analycical Theoiy of Heat. 

Remember that firms will put their best face forward. Where possible, it is advisable to visit at least two jobs recently carried out, and to contact at least two referees. A banker s reference can be misleading, although perfectly accurate. A banker s reference will say that the firm is in a trading position, and sometimes that the firm is in... 

A different questionnaire needs to be sent to the referees, remembering to enclose a stamped addressed envelope. The referee is doing an unpaid favor in completing the questionnaire, and for his sake, it should enable him to return it without having to write a thesis with an attached letter. A box to be ticked for most questions will suffice, with a space at the end for general comments. Questions that should be asked of the referee are ... 

Advances in Chemistry Series was founded in 1949 by the American Chemical Society as an outlet for symposia and collections of data in special areas of topical interest that could not be accommodated in the Society s journals. It provides a medium for symposia that would otherwise be fragmented, their papers distributed among several journals or not published at all. Papers are refereed critically according to ACS editorial standards and receive the careful attention and processing characteristic of ACS publications. Papers published in Advances in Chemistry Series are original contributions not published elsewhere in whole or major part and include reports of research as well as reviews since symposia may embrace both types of presentation. 

Acknowledgements—We have worked together on this paper for several years. It in turn represents a coming together of two still older single-authored papers. One or other (or both) of the authors would like to thank each of the following for helpful comments on at least one previous version Michael Akeroyd, Nathan Brookes, Stephen Brush, Fernando Dufour, Carmen Giunta, Robin Hendry, Colin Howson, Peter Lipton, and two anonymous referees of this journal. 

A data set containing measurements on a set of known samples and used to develop a calibration is called a training set. The known samples are sometimes called the calibration samples. A training set consists of an absorbance matrix containing spectra that are measured as carefully as possible and a concentration matrix containing concentration values determined by a reliable, independent referee method. 

The data in the validation set are used to challenge the calibration. We treat the validation samples as if they are unknowns. We use the calibration developed with the training set to predict (or estimate) the concentrations of the components in the validation samples. We then compare these predicted concentrations to the actual concentrations as determined by an independent referee method (these are also called the expected concentrations). In this way, we can assess the expected performance of the calibration on actual unknowns. To the extent that the validation samples are a good representation of all the unknown samples we will encounter, this validation step will provide a reliable estimate of the calibration s performance on the unknowns. But if we encounter unknowns that are significantly different from the validation samples, we are likely to be surprised by the actual performance of the calibration (and such surprises are seldom pleasant). 

In practical terms, we can usually develop satisfactory calibrations with training set concentrations, as determined by some referee method, that are accurate to 5% mean relative error. Fortunately, when working with typical industrial applications and within a reasonable budget, it is usually possible to achieve at least this level of accuracy. But there is no need to stop there. We will usually realize significant benefits such as improved analytical accuracy, robustness, and ease of calibration if we can reduce the errors in the training set concentrations to 2% or 3%. The benefits are such that it is usually worthwhile to shoot for this level of accuracy whenever it can be reasonably achieved. 

We must often compromise between the number of samples in the training set and the accuracy of the concentration values for those samples. This is because the additional time and money required for a more accurate referee method for determining the concentrations must often be offset by working with fewer samples. The more we know about the particulars of an application, the easier it would be for us to strike an informed compromise. But often, we don t know as much as we would like. 

Our data will simulate spectra collected on mixtures that contain 4 different components dissolved in a spectrally inactive solvent. We will suppose that we have measured the concentrations of 3 of the components with referee methods. The 4th component will be present in varying amounts in all of the samples, but we will not have access to any information about the concentrations of the 4th component. 

For the purposes of this section, error is simply the difference between the value of the y variable predicted by a regression and the true value (sometimes called the expected value). Naturally, it is impossible to know the true value, so we are forced to settle for using the best available referee value for the y variable. (Note it is possible that the "best available referee values" can have larger errors than the predicted values produced by the calibration.) We will follow the common convention and name the expected value of the variable y and the predicted value of the variable y, pronounced "Y-hat." Then the error is given by p -y. We will also denote the number of samples in a data set by the letter n. 

We owe much to our continued discussions of these problems with A. G. Harrison, to whom we are also indebted for preprints prior to publication. We are grateful to a referee for drawing our attention to Ref. (30) and for seeking clarification of several points which were obviously obscure in the original manuscript. We thank the Petroleum Research Fund for a grant to support both this work and one of us (D.J.H.). 

Another indicator of the performance of AO systems is the number of refereed publications per year based on observations made using AO. This has been rising steadily over the last decade and is now >30 per year (see http //www2.keck.hawaii.edu 3636/realpublic/inst/ao/ao sciJist.html). 

The Lick AO system demonstrated the first wavefront correction using a Na LGS (Max et al., 1997), and engineering tests demonstrated LGS corrected Strehls of 0.5-0.6 at 2.2 m (Fig. 15, Gavel et al., 2003). In 2002, the LGS system was turned over to the Observatory staff for operation in science observing mode. It is used almost 100 nights per year. The first refereed science paper using a sodium LGS/AO system was published by Perrin et al. (2004). 

Xhe 1987 WINTER SYMPOSIUM of the Division of Industrial and Engineering Chemistry, hosted by the Minnesota Supercomputer Institute, was supported by grants from the American Chemical Society, the Minnesota Supercomputer Institute, Cray Research, Inc., and ETA Systems, Inc. The symposium consisted of four half-day sessions with four lectures per session and a two-part poster session with 15 poster papers. This book includes chapters by the lecturers plus five papers contributed by the session chairs. All contributions were refereed anonymously according to usual procedures of the ACS Symposium Series. 

---