Data collection accuracy required

It is important to note that as data collection accuracy has increased, so have the requirements for accuracy of test equipment and data collection methods set forth in test methods. For example ASTM D638 and ISO 527 (International Standards Organization), both tensile property methods, require the use of extensometers to measure strain for modulus determinations. Grip separation and other methods of measuring strain have been determined to be inadequate for the measurement accuracy required at the low strains necessary for elastic modulus calculations. 

Written procedures and practices must be in place within the laboratory to verify the accuracy of manually entered and electronically transferred data collected on automated systems. The primary documentation for data entry requirements is an audit trail. Laboratories must ensure that an audit trail exists and is maintained. This audit trail must indicate date and time stamps for each record transmitted and the source instrument for each entry. 

The precision and accuracy of the overall method was assessed by collecting and analyzing three sets of samples from test atmospheres of known concentration. An overall coefficient of variation of 10% for all analytical data and accuracy of 10% was required for method validation. 

To obtain such samples, the chemical and physical characteristics of the site(s) must be considered, as well as possible reactions between the compound(s) of interest and soil components and the degree of variability (i.e., variation in soil profiles) within the sampling site. With these data, the site(s) can then be divided into homogeneous areas and the required number of samples can be collected. The required number of samples depends on the functions of variance and degree of accuracy. Once the correct procedure has been determined, sampling can proceed. 

The required data are reviewed in detail in Section 1.4. In order to be economic and efficient, the data collection is accompanied by their interpretation in terms of risks. This allows adapting the amount and accuracy of the data to the risk. This procedure is illustrated in the example of thermal data in Section 3.4. 

In the diffusive interfacial transport-refractive index (DIT-NDX) method, compositions are determined using precise refractive index data (8). Refractive index data valid to +/- 0.00005 are obtainable using the DIT apparatus vithin an area of 30 ym2 in a sample approximately 25-ym thick (0.75 picoliter volume). Data collection and analysis require 9 seconds. The accuracy, spatial resolution, and speed vith vhich refractive indices can be determined are thus superb. 

Another factor which influences the speed in performing an analysis is calibration of the instrument. Calibration is especially time-consuming in cases where different elements are run on every analysis but even in cases where the same elements are determined time after time, the frequency of instrument calibration required to maintain a desired level of accuracy is an important consideration. Since manual data collection is not feasible in multielement determinations, the ideal system would undoubtedly be computerized. The computer would handle all data collection steps, the construction of calibration curves by mathematical curve-fitting methods, and the calculation of concentrations from these curves. 

Examples of quality assurance protocols that are considered standard practice in any data collection scheme include the use of both internal control samples (e.g. use of field blanks and spikes6) and external quality assurance samples (e.g. duplicate samples of known concentrations sent to different laboratories) to determine the extent of intra- and interlaboratory variation. Ensuring that the data have not been compromised or corrupted may also require setting up accessible data archives of original paper or electronic records so that the accuracy of summaries of the data published in documents and articles can be verified. 

An ED study requires a complete set of intensity data, measured to a much higher Bragg angle (sin 9IX > 1.1 A , sometimes up to 1.77A ) than is required for an lAM study, and with much more accuracy and precision. It is necessary to measure all symmetrical equivalents of each reflection, carefully address all possible sources of systematic errors, especially extinction and TDS. With a point detector, the data collection required many weeks or even months, but with an area detector (especially using SR) this time has been reduced to a few days, without any loss in precision. There is an intrinsic difficulty to distinguish between the ED distribution of a stationary atom and an apparent smearing of its ED from thermal motion, therefore the data must be collected at as low a temperature as possible. ... 

The Data Handling and Record Retention section of the protocol will address the requirement to maintain data (whether on a paper CRF or using an electronic data collection tool (DCT)) of each trial subject. It will address expectations of ownership of the completed CRF data, the investigator s responsibility to ensure accuracy and completeness of data recording. This section will also address the requirements for retention of records at the trial site in accordance with relevant guidelines and regulatory requirements. 

Investigator site audits are probably the most frequent type of audits conducted by clinical QA departments and, therefore, deserve particular attention. The purpose of investigator site audits is to assess compliance with the GCP regulations (with a focus on the country-specific regulatory requirements) and the protocol. Further, thesafety of the trial participants, the ethical conduct of the trial and the validity, completeness and accuracy of the data collected and recorded are verified during the audit. 

Currently, the scope of verification varies widely. Some verification statements may simply confirm the accuracy of data collection or provide an overall general endorsement of the report. Others may be far more comprehensive and challenging, involving discussions with management, site visits, a review of documentation and management systems, and finally pinpointing areas requiring improvement. It is felt that some statement on any activities of this type carried out should appear in the CER itself to consolidate the authenticity. 

In routine data collection for macromolecular crystallography we are interested in knowing the wavelength to =0.001 A, whereas in current anomalous scattering experiments an accuracy of 0.0001 A is required. Hence, the values of 2d for the (111) and (220) planes of silicon and germanium are given in table A3.4 to four decimal places calculated... 

The measurements of performance capacities associated with human functions should conform to certain criteria. Details can be found in Brand and Crownshield [ 1981 ] and Chaffin [ 1982]. A general set of psychometric criteria is also discussed by Sanders and McCormick [1993]. It includes measurement accuracy, reliability, validity, sensitivity, and freedom from contamination. Meeting these criteria depends not only on the instruments used but also on the methods of analysis and the expertise of the analyst. It is almost impossible to satisfy these criteria perfectly, especially when the task is being performed in its natural environment (as opposed to a laboratory simulation). However, the analyst must always be aware of them and must be pragmatic in measuring task variables. Meister [1985] lists the following practical requirements for measurements (1) objective, (2) quantitative, (3) unobtrusive, (4) easy to collect, (5) requiring no special data-coUection techniques or instrumentation, and (6) of relatively low cost in terms of money and effort by the experimenter. However, these are not necessarily mutually exclusive. 

All data collected by the TRACS survey vehicle are referenced to the network sections to a longitudinal accuracy of 1 m, processed by specially designed software and fed, if required, to HAPMS. 

Tests for resin and film property values can be performed in-house, by an outside testing facility, or by a supplier. There are advantages and disadvantages to each. The most important consideration is the accuracy of the data collected. After that, consideration must be given to costs associated with the amount of data needed (equipment, consumables, training, labor, etc.) and the frequency of measurement required. 

There is a growing awareness of the limitations of current approaches to the collection and assessment of dietary data. There is a search for new and better methodology. Alternative approaches to both procedures for data collection and to interpretation of collected data have been presented. These appear to represent a considerable improvement on previous strategies. Careful application of these approaches may enhance the accuracy of estimates of nutrient requirements and intakes. By comparing the situation predicted from dietary analyses with that suggested by biochemical studies, an epidemiologic test of the validity of requirement estimates can be undertaken. The approaches described have both immediate operational application and future research implication. 

---