Reproducibility of results

Prior to the introduction of SEC, polymer molecular mass characterization was generally very difficult and time-consuming. SEC permits this characterization to be carried out relatively routinely. However, the variation of results between laboratories, and even within a single laboratory over a period of time, has made this characterization somewhat questionable. 

Round-robin experiments have been conducted to compare the results between laboratories [15], and the variation of results within a laboratory have been studied [16]. Large variations in calculated molecular masses have been observed, but not generally explained. 

It would be expected that, if all the extra-column aspects of an SEC system were kept constant, exactly the same calibrants were employed and the data 

It is likely that the main variations in calculated molecular masses are due to the data manipulation and the actual SEC columns. There must be potential for imposing common methods of data manipulation but, with the range of data handling systems available, this would be difficult. Unfortunately, the contribution of the SEC columns is complex and not well appreciated. It is quite easy to compound these problems by an inappropriate selection of columns. 

In SEC, frequently only the calculated average molecular masses are quoted as results, and this can lead to apparently very poor levels of reproducibility. However, if the results are quoted as the log molecular masses, it can be seen that the reproducibility is much better than originally perceived. 

---

Estimations based on statistics can be made for total accuracy, precision, and reproducibility of results related to the sampling procedure being applied. Statistical error is expressed in terms of variance. Total samphng error is the sum of error variance from each step of the process. However, discussions herein will take into consideration only step (I)—mechanical extraction of samples. Mechanical-extracdion accuracy is dependent on design reflecding mechanical and statistical factors in carrying out efficient and practical collection of representative samples S from a bulk quantity B,... 

Good heat transfer on the outside of the reactor tube is essential but not sufficient because the heat transfer is limited at low flow rates at the inside film coefficient in the reacting stream. The same holds between catalyst particles and the streaming fluid, as in the case between the fluid and inside tube wall. This is why these reactors frequently exhibit ignition-extinction phenomena and non-reproducibility of results. Laboratory research workers untrained in the field of reactor thermal stability usually observe that the rate is not a continuous function of the temperature, as the Arrhenius relationship predicts, but that a definite minimum temperature is required to start the reaction. This is not a property of the reaction but a characteristic of the given system consisting of a reaction and a particular reactor. 

FIGURE 16.8 Dextran T-SOO ( ) separated on Sephacryl S-SOO/S-IOOO (60 + 9S x 1.6 cm) sampled in 3-ml fractions normalized (area = 1.0) eluogram profiles (ev) constructed from an off-line determined mass of carbohydrates within each of the fractions flow rate 0.42 ml/min Vexd = 126 ml, V,o, = 273 ml eluent 0.005 M NaOH reproducibility of results over a period of 6 months 5% of ev maximum and 3 ml ( 1 fraction). 

Most cast irons, except those fully decarburised during malleabilising, give coatings of the chromium carbide type. In view of the great variations in composition of cast irons, reproducibility of results can be achieved only by careful control of specification. High phosphorus and sulphur contents are detrimental to the formation of non-porous coatings. 

These remarks apply as well to the treatment of the surfaces of specimens to be used in tests in corrosion research projects, except here selection of a particular method of surface preparation is required so as to achieve reproducibility of results from test to test and amongst different investigators. Methods of preparing specimens are described in ASTM Gl 1988 and ISO 7539-1 1987. 

Statistical analysis of data from tests with an apparatus by Wesley has demonstrated satisfactory reproducibility of results not only among specimens in a particular test, but also from test to test undertaken at different times. 

To favour reproducibility of results, the cycles of immersion and withdrawal must be kept the same from test to test. It is necessary to control the temperature and humidity of the atmosphere surrounding the test setup as these affect the rate of evaporation of the solution and of drying of the specimens when they are out of the solution. It is also necessary to provide for replenishing losses of water from the test solution resulting from evaporation. 

Analytical measurement Pros - Results obtained reflect well reality - Repeatability and reproducibility of results (at least between good qualified labs) - Measurements are independent of information/data sources - Multipurpose analytical methods can cover many compounds on a single run - Even the best model will ultimately need to be experimentally checked - Discovery of new emerging contaminants is possible... 

This is an important but very difficult area for research, particularly regarding the formulation and evaluation of stabiliser systems. Non-reproducibility of results is a serious problem and in commercial practice a single batch can unexpectedly give rise to damage despite using the same protective measures that had been successful with apparently similar... 

Advances in measurement of gene expression have also been phenomenal. In five years, the numbers of genes, sensitivity of the assays, and reproducibility of results have also increased significantly as has the ability to analyze the data. The measurement of gene expression has been used in several ways in cancer genetics and cancer pharmacogenomics. Gene expression has led to better ways to classify cancers and to select the appropriate therapy (Miyazato et al., 2001 Birner et al., 2001). 

The cellulose specimen under examination is refluxed in the acid-oxidant mixture and the gases formed are swept continuously into an absorption train by a carrier stream of air free of carbon dioxide. Conrad and Scroggie20 have added a number of important improvements which apparently increase the reproducibility of results. One of their modifications is a stirrer in the reaction chamber which reduces the danger of bumping caused by superheating. The latter is undesirable since the reaction is apparently quite sensitive to the temperature. 

There is a recent trend towards simultaneous CE separations of several classes of food additives. This has so far been applied to soft drinks and preserved fruits, but could also be used for other food products. An MEKC method was published (Lin et al., 2000) for simultaneous separation of intense sweeteners (dulcin, aspartame, saccharin and acesulfame K) and some preservatives (sorbic and benzoic acids, sodium dehydroacetate, methyl-, ethyl-, propyl- and isopropyl- p-hydroxybenzoates) in preserved fruits. Ion pair extraction and SPE cleanup were used prior to CE analysis. The average recovery of these various additives was 90% with good within-laboratory reproducibility of results. Another procedure was described by Frazier et al. (2000b) for separation of intense sweeteners, preservatives and colours as well as caffeine and caramel in soft drinks. Using the MEKC mode, separation was obtained in 15 min. The aqueous phase was 20 mM carbonate buffer at pH 9.5 and the micellar phase was 62 mM sodium dodecyl sulphate. A diode array detector was used for quantification in the range 190-600 nm, and limits of quantification of 0.01 mg/1 per analyte were reported. The authors observed that their procedure requires further validation for quantitative analysis. 

Apart from tests carried out on single crystals, also fractions of uniform particle size should be tested in order to obtain precise readings and to ensure reproducibility of results. For runs in vacuo the use of particle sizes under 60 (0.25 mm) mesh should be avoided, since the gaseous products in the case of thermal decomposition may eject particles of the test material from the crucible. 

Always use only HPLC-Grade solvents for better precision and reproducibility of results. 

For numerous research intentions, an average scale will be chosen, considering on the one hand, the costs of chemicals and on the other, the easier experimental work and better reproducibility of results, using a cell of medium - but not too small - dimensions. 

Reproducibility of the experimental conditions is fairly easy to obtain. However, reproducibility of results in one set of circumstances is, by itself, no guarantee of their validity, since, especially with a vacuum line, the impurity effects can often be reproduced extremely well. 

Attempts to estimate the probable hazard of handling a new compound were generally based on correlation of available test data with that of materials which had been processed successfully. Unfortunately, many of the test techniques used were designed for reproducibility of result rather than for interpretation of hazard, and they applied primarily to correlation within the industry where they were developed. 

Ruggedness measure for variability (reproducibility of results obtained under variety of conditions) expressed as %RSD (interlaboratory)... 

Figure 4. Quantum yield of Norrish type II reaction as a function of temperature for 0.5 mm PS-PVK films irradiated with monochromatic radiation of 313 mu wavelength as measured by GPC. Error bars represent errors ( 20%) in reproducibility of results.

Single crystals, cut so that a certain crystal face is exposed to the solution, offer better definition and therefore reproducibility of results. By using electron diffraction measurements (which require a vacuum), one can determine the crystal face exposed to the solution before and after electrochemical measurements and hence ascertain if any change in crystal orientation has occurred as a result of contact with the solution. Such techniques, introduced by Hubbard in the 1970s, began a seminal change in electrode kinetics, the full fruits of which are still to be obtained. 

Polyacetylene attracts constant attention as an excellent simple model of the polyconjugated polymer on which the main optical and electrical properties can be verified. The possibility of achieving metallic conductivities by doping opens real perspectives of practical application of conducting polymers. The complication is the strong interaction with oxygen. The reproducibility of results strongly depends on the synthesis and measurement conditions. 

As experienced by previous investigators (5, 17), the reproducibility of results was poor unless successive runs on pure hydrocarbons (in the absence of halogen compounds) were carried out before examining the... 

The chemical yield of each element was usually about 70%. The rubidium content of the 12 County Durham coals was 6-30 ppm, and the cesium content in 10 of the coals was 0.8-3.7 ppm. Reproducibility of results was very good for cesium the relative standard deviation of the mean result was usually less than 5%. Rubidium results were slightly less reproducible, the largest relative standard deviation being 22%. 

Operator skill and experience are necessary to obtain consistent results using a Durometer. Speed of load application, dwell time, and sample thickness can affect reproducibility of results. Durometer calibration prior to each test series is done using a test block provided with the instrument. When large numbers of tests are required, improved consistency of results are obtained if the Durometer is used with the accessory vertical stand rather than hand held. 

The same process was studied using NPSV in acetonitrile—Bu4NBF4 (0.1 M). The results are summarized in Table 30. In this case, five different electrolyte solutions were used for measurements at 285 K. The overall results was k° equal to 0.98 0.03 cm s-1 and, in this case, the overall standard deviation was less than that for the individual solutions. Aalstad and Parker [46] concluded that NPSV is a superior method for the study of rapid heterogeneous charge transfer reactions and that the acetonitrile— Bu4NBF4 medium is less problematic than DMF—Bu4NBF4 as far as reproducibility of results is concerned. 

The action of the abrasive material on an area about 6 mm in diameter allows a hole of various depths to be made depending on the hardness of the material under test (Figs. 4.4.8 and 4.4.9). The readiness of measurement, the possibility of duplicating holes successively while conserving the measuring evidence, and the high reproducibility of results combined with the possibility of hardness dilferentiation in a wide range—all favour this device rather than Bohme s disc. 

Reproducibly of results, however, was questioned, because the total oxidation reaction was underestimated [84]. Possibly, these discrepancies derive from the observation of structural changes of the Pd-based membrane during direct hydroxylation of benzene to phenol [85]. The surface state of Pd during the reaction could be divided into two major regions the oxidized region near the gas... 

This method is extremely sensitive to pH changes which can lead to inconsistent transfection efficiencies, especially when using homebrew transfection buffers. To some extent, this sensitivity can be limited by the use of commercially available kits containing chemicals and buffers that have undergone quality control procedures, ensuring better reproducibility of results and less lot-to-lot variation. Although the costs per transfection for this method are unrivaled, the attractiveness of calcium phosphate precipitation has declined over the past 15 years, partly due to the trickiness of the method itself, the limited transfection efficiencies, and the narrow cell spectrum for which it is suitable, and partly because more modem and efficient DNA delivery methods have emerged. 

---