The quality of data

Two factors matter most in gaming knowledge from data first, the quality of the data and secondly, the method one applies to the data, and by which one learns from them. 

First, we have an initial, and probably utterly crude, dataset. Genuine data pre-processing has only just started. The task is to assess the quality of the data. One of the topics for discussion in this chapter is the methods by which one finds out the potential drawbacks of the dataset. 

Thorinm-232 is the only non-radiogenic thorium isotope of the U/Th decay series. Thorinm-232 enters the ocean by continental weathering and is mostly in the particulate form. Early measurements of Th were by alpha-spectrometry and required large volume samples ca. 1000 T). Not only did this make sample collection difficult, but the signal-to-noise ratio was often low and uncertain. With the development of a neutron activation analysis " and amass spectrometry method " the quality of the data greatly improved, and the required volume for mass spectrometry was reduced to less than a liter. Surface ocean waters typically have elevated concentrations of dissolved and particulate 17,3 7,62... 

The methods discussed in this section extend the original concept of deriving structures from experimental NMR data in two ways. First, during the structure calculation, part of the assignment problem is solved automatically. This allows specification of the NOE data in a fonn closer to the raw data, which makes the refinement similar to X-ray refinement. Second, the quality of the data is assessed. The methods have been recently reviewed in more detail [64,67]. 

The well-known difficulties in calculating tliree-dimensional structures of macromolecules from NMR data mentioned above (sparseness of the data, imprecision of the restraints due to spin diffusion and internal dynamics) also make the validation of the structures a challenging task. The quality of the data [88] and the energy parameters used in the refinement [89] can be expected to influence the quality of structures. Several principles can be used to validate NMR structures. 

The trends begun with the general introduction of FTIR technology will undoubtedly continue. It is safe to say that the quality of the data being produced far exceeds our ability to analyze it. In fact, for many current applications, the principle limitations are not with the equipment, but rather with the quality of the samples. Thus, the shift from qualitative to quantitative work will proceed, reaching high levels of sophistication to address the optical and matrix interference problems discussed above. 

Chapter 6, Data Collection and Incident Analysis Methods, examines the pitfalls involved in collecting data on human error and suggests possible approaches to improving the quality of the data. 

The selection of data from the resources available required decisions about the acceptable quality of the data and applicability of the data to the CCPS Taxonomy. Data from a resource was rejected by SAIC and the CCPS Subcommittee when ... 

If the data quality was acceptable, they were then evaluated for their relevance and fit to the CCPS Taxonomy. The data in the SAIC data base were fitted to taxonomy levels that best correlated with nuclear plant equipment and operational environments. CPI resources were reread thoroughly to understand the equipment subtypes, operating modes, and process severities represented by the data points and to identify as many relevant taxonomy levels as possible. SAIC data analysts made preliminary judgments on the applicability of data points to taxonomy levels and on the quality of the data. The majority of the data applied to high taxonomy levels (x.x) and a smaller amount was applicable to lower levels (x.x.x.x). The data were assigned to the lowest level possible. 

FIGURE 11.22 Control charts and outliers, (a) pEC50 values (ordinates) run as a quality control for a drug screen over the days on which the screen is run (abscissae). Dotted lines are the 95% c.l. and the solid lines the 99.7% c.l. Data points that drift beyond the action lines indicate significant concern over the quality of the data obtained from the screen on those days, (b) The effect of significant outliers on the criteria for rejection. For the data set shown, the inclusion of points A and B lead to a c.l. for 95% confidence that includes point B. Removal of point A causes the 95% limits to fall below points B, causing them to be suspect as well. Thus, the presence of the data to be possibly rejected affects the criteria for rejection of other data. 

Other aspects such as the quality of the data, the model precision, and the number of possible alternatives presented to the user as possible solutions of the problem or the degree of confidence of the final user on the DSS... 

The structure was refined with block diagonal least squares. In cases of pseudo-symmetry, least squares refinement is usually troublesome due to the high correlations between atoms related by false symmetry operations. Because of the poor quality of the data, only those reflections not suffering from the effects of decomposition were used in the refinement. With all non-hydrogen atoms refined with isotropic thermal parameters and hydrogen atoms included at fixed positions, the final R and R values were 0.142 and 0.190, respectively. Refinement with anisotropic thermal parameters resulted in slightly more attractive R values, but the much lower data to parameter ratio did not justify it. 

The tail of the FID contains very little information rather, most of the relevant information is in the initial large-volume portion of the FID envelope. The loss of the tail of the FID should not, therefore, significantly affect the quality of the data. Another manipulation to compensate for the lost information is exponential multiplication, in which the FID is multiplied by a negative exponential factor. 

In order to apply RBL or GRAFA successfully some attention has to be paid to the quality of the data. Like any other multivariate technique, the results obtained by RBL and GRAFA are affected by non-linearity of the data and heteroscedast-icity of the noise. By both phenomena the rank of the data matrix is higher than the number of species present in the sample. This has been demonstrated on the PCA results obtained for an anthracene standard solution eluted and detected by three different brands of diode array detectors [37]. In all three cases significant second eigenvalues were obtained and structure is seen in the second principal component. 

Evaluation of the results Evaluation of the results consists of (i) technical scrutiny of the consistency and of the quality of the data the acceptance, on technical (not statistical) grounds, of data to be used to calculate the certified value and its uncertainty, (2) the calculation (using the appropriate statistical techniques) of the certified value and its uncertainty. The approach indudes technical discussion of the results among all cooperators, rejection of outliers, statistical evaluation, and calculation of the certified value and uncertainties. 

An organized document filing system must be maintained. This could be a paper file, an electronic document file, or a mixture of both. The equipment inventory system contains key information on the components of each system, their performance criteria and maintenance and calibration status. All documents including installation and performance documentation, as with other documents necessary to demonstrate the quality of the data, include SOPs for the qualification procedures, calibration, maintenance, personnel training, etc. If the manufacturer s operating, service, or maintenance manuals are used or cited in the operating procedures, copies of these manuals should be maintained. To facilitate retrieval, documentation should be stored in a central location and be indexed for easy retrieval. 

How clean the data must be depends on the importance of the data. Critical analysis variables must be clean, so this is where the site and data management groups should focus their resources. If the data are dirty at the time of statistical analysis, many inefficient and costly workarounds may need to be applied in the statistical programming, and the quality of the data analysis could suffer. However, if a variable is not important to the statistical analysis, then it is better to save the expense of cleaning that variable. 

We have not included Atom Probe Microanalysis in this scheme. It constitutes the ultimate in local analysis - in that individual atoms can be selected and identified by TOF spectroscopy. Chapter 1 gives an account of the range of applications of the technique at the present time the development in atom-probe methods has allowed the continuing increase of both the volume of material that can be mapped at the atomic scale and the quality of the data obtained. 

Two sensibly priced commercial databases for solubility exist [366,507], An article in the journal Analytical Profiles of Drug Substances carries solubility data [496]. Abraham and Le [508] published a list of intrinsic aqueous solubilities of 665 compounds, with many ionizable molecules. It is difficult to tell from published lists what the quality of the data for ionizable molecules is. Sometimes, it is not clear what the listed number stands for. For example, Sw, water solubility, can mean several different things either intrinsic value, or value determined at a particular pH (using buffers), or value measured by saturating distilled water with excess compound. In the most critical applications using ionizable molecules, it may be necessary to scour the original publications in order to be confident of the quality of reported values. 

The quality of the data collected from 2% DOPC membranes is unmatched by any other system we have explored. It s not uncommon to see interplate reproducibility <5% and intraplate even better than that (1-3% SD). As will be seen later, lipid model 1.0 does not predict GIT absorption as well as some of the newer pION models. However, this may not be the case when BBB models are explored in detail. 

The advent of CCD detectors for X-ray diffraction experiments has raised the possibility of obtaining charge density data sets in a much reduced time compared to that required with traditional point detectors. This opens the door to many more studies and, in particular, comparative studies. In addition, the length of data collection no longer scales with the size of the problem, thus the size of tractable studies has certainly increased but the limit remains unknown. Before embracing this new technology, it is necessary to evaluate the quality of the data obtained and the possible new sources of error. The details of the work summarized below has either been published or submitted for publication elsewhere [1-3]. 

The quality of the intensity data obtainable has been assessed from an experiment on oxalic acid obtained at 100 K with a CCD detector. In this experiment the contamination of 712 to the measured intensities was eliminated by appropriate choice of the generator voltage. Various criteria for judging the quality of the data are discussed below [2],... 

An interesting aspect of the present arrangement arises in connection with the poor quality of the data set and at the same time the reliability of H-atom positions. These are included in the scattering model with a fair amount of ambiguity in their positions, more than usual in X-ray experiments. Certain abnormalities in the geometry of the carboxyl groups may be understood as a result of conformational... 

In assessing animal data, careful attention must be paid to the quality of the data, the incidence of spontaneous tumors in the control population, consistency if more than one study is available, and statistical validity. If the exposure route and experimental regimen employed do not agree with the most likely mode(s) of human exposure (e.g., intramuscular injection), the data must be interpreted cautiously. Consideration should be given to data on metabolism of the compound by the animal species tested, as compared with metabolism in humans if this information is known. If only in vitro data are available, only qualitative estimates may be possible because of uncertainties regarding the association between in vitro results and human or animal effects. The availability of associated pharmacokinetic data, however, may allow development of a rough quantitative estimate. 

Whilst sample preparation may not be the most interesting aspect of NMR spectroscopy, it is nonetheless extremely important as it will have a huge bearing on the quality of the data obtained and therefore on your ability to make logical deductions about your compounds. This is particularly true when acquiring the most straightforward 1-D proton spectra. The most typical manifestation of sub-standard sample preparation is poor line shape. It is worth remembering that in terms of 1-D proton NMR, the devil can be very much in the detail . Detail , in this context, means fine structure and fine structure is always the first casualty of poor sample preparation. 

---