Electronic Data Collection and Analysis System
Yin Foo, Des; Foley, Margaret
E-Publications; PublicationNT; E-Books; Report ; 39/1992
Made available via the Publications (Legal Deposit) Act 2004 (NT).
Power and Water Authority
Report ; 39/1992
Check within Publication or with content Publisher.
Technical Report WRD92039 Viewed at 15:07:17 on 29/07/2010 Page 153 of 192. test may period, and be suspect. the manual reading giving the maximum error The statistical nature cf measurement accur<9.cy lS to describe this 'work. well established, and future work should attempt the particular characteristics associated with If the errors found could be adequately described by a Normal distribution, then the l!1ean (11) and standard Deviation (SD) could be used to compare results and to gain a measure of the accuracy of a particular system. While there is always a possibility that the maximum error taken during subsequent system use may exceed that found in the test, approximately 95% of subsequent readings are expected to be within t1 +2*SD. For example, the mean and SD found in a test are 10mm and 18mm respectively, then 95% of that system's future data would be expected to be within -26 and +46mm. In reality, the distribution is expected to be "skewed" by individual operators. The Hean error over all readings (within range) should be zero, if the calibration is as good as possible. Therefore the analysis is best done in two stages - firstly finding the calibration parameters which give a zero mean error, and minimum SD and then using the recalibrated data to find the system accuracy. The recalibration must be applied to the syStE~IllS for all subsequent use (as per C2. 4) . Note that even with the best possible calibration some syst~m error is inherent, and that any particular later use of the 8ystem may not have a zero mean error. This is because the check readings available may not cover the full transducer operating range. the standard test The calibration for zero mean is, however, the best available error from over the full range, with use when the ,vater level is both increasing and decreasing. The folle,wing methods are based on systems which have a linear response to changing pressure - In future systems with non-l:Lnear ca:.ibration parameters may be available. In that
Aboriginal and Torres Strait Islander people are advised that this website may contain the names, voices and images of people who have died, as well as other culturally sensitive content. Please be aware that some collection items may use outdated phrases or words which reflect the attitude of the creator at the time, and are now considered offensive.
We use temporary cookies on this site to provide functionality.
You are welcome to provide further information or feedback about this item by emailing TerritoryStories@nt.gov.au