Accuracy, instrument reading

Random errors arise in all measurements and are inevitable, no matter what the experiment, the quality of the instrument, or of the analyst. They are a consequence of the limitations of experimental, observations. For example, an instrument reading can only be taken within the limits of accuracy of the scale, as read by a particular observer. The position of the pointer between two division marks may be estimated to one fifth of a division by a skilled experimenter, but only to one half a division by another. Such skill may be improved with practice, but will never be totally perfected. Random errors cannot be eliminated, but can be reduced by using more sensitive measuring instruments or an experienced experimenter. The magnitude of random errors can be estimated by repeating the experiment. 

The accuracy of any measurement will depend upon the calibration of the instrument used. The calibration of an instrument determines its response to a known amount or concentration of radioactivity. This allows a correlation to be made between the instrument reading and the actual amount or concentration present. A range of activities of radium-226 standard reference materials (SRM) is available from the U.S. Department of Commerce, National Bureau of Standards (NBS) as solutions for calibrating detection systems. Also, an elevated radon atmosphere may be produced in a chamber, and samples drawn and measured in systems previously calibrated by radon emanation from an NBS radium-226 SRM. Other radon detectors may then be filled from or exposed in the chamber and standardized based on this "secondary" standard (NCRP 1988). Analytical methods for measuring radon in environmental samples are given in Table 6-2. These methods provide indirect measurements of radon i.e., the activity emitted from radon and radon progeny is detected and quantitied. 

For neutron dosimeters there is very often a significant dependence of the reading on the neutron energy. The spectral distribution of the neutrons used for calibration and the spectral distribution of the neutrons to be measured may affect the accuracy of dose determination considerably. If the energy dependence of the instrument reading and the spectral distribution of the neutrons to be measured are known, a corresponding correction factor may be used. 

Were there any differences between the accuracy of instrumentation readings in the control room compared with the requirements in the procedure ... 

If the instrument is to be direct reading, the second (or cold ) junction must be kept at a constant reference temperature. If high temperatures are to be measured then the terminals of the detector can be used as the cold junction without an unacceptable loss of accuracy. 

The commercial units have a very low thermal capacity and very high response speeds. Some are available with several independent channels and a common cold junction. Each channel is scanned in turn by the instrument, and the readings either displayed or stored for future recovery. Accuracies of better than 0.2 per cent are possible. Thermocouples are available to cover a very wide range of temperatures, their cost is low and they have a small mass, so minimizing the intrusive effect on the surface at the point where the temperature is being measured. The output characteristics (output voltage versus temperature) are reasonably linear but the measurement accuracy is not particularly high. 

Valve voltmeters were widely used in the past, but have been replaced by transistor voltmeters. With instruments of this type it is possible to achieve an input resistance of 50 MQ or more, the current required to operate the instrument being of the order of 10" A. The early instruments had a tendency to zero drift on the lower ranges, but this has been overcome in the modern transistor types. Such instruments are most often used to make potential readings in extremely high-resistance electrolytes. The accuracy of such instruments is of the order of 2% full-scale deflection. It is necessary to ensure that both types are so designed that they do not respond to alternating currents. 

The size, speed, noise levels, precision, accuracy, and cost vary among the instruments. Higher cost does not necessarily mean better performance. Unlike typical spectroscopy, where the sample is reduced and, often, diluted in a non-interfering matrix, samples in NIR are read as is and one size instrument does NOT fit all. The application will determine the instrument (an idea that escapes many instrument salespeople). 

The RQ flex test kit (Merck) which uses specific test strips is useful for the semi-quantitative determination of several analytes. D(+) ascorbic acid can be determined in fortified food products with an accuracy of 85-115% (unpublished data), however the procedure cannot be applied to coloured food products. Added iron salts may be extracted from food products with dilute sulphuric acid and adjusted to pH2 with NaOH solution. Fe3+ is reduced to Fe2+ with ascorbic acid. Fe2+ reacts with Ferrospectral to form a red-violet complex. An internal calibration is provided on a barcode which is read by the RQ-flex reflectometer prior to any measurements. This avoids the need to calibrate the instrument with standard solutions. 

A problem almost universally encountered in continuous-flow systems is that the instrument response for a given sample assay-value tends to vary with time. This effect, known as drift, affects the accuracy of results. It may be due to several causes, in particular variable performance of analyser components and variations in chemical sensitivity of the method used. It is manifest in two forms, baseline drift and peak-reading drift, which is due to sensitivity changes. The baseline drift may be detected visually if a... 

Wavelength accuracy is defined as the deviation of the wavelength reading at an absorption band or emission band from the known wavelength of the band. The wavelength deviation can cause significant errors in the qualitative and quantitative results of the UV-Vis measurement. It is quite obvious that if the spectrophotometer is not able to maintain an accurate wavelength scale, the UV absorption profile of the sample measured by the instrument will be inaccurate. The true Amax and A.min of the analyte cannot be characterized accurately. 

---