Hello everybody,
Is there, by any chance, a statistical method that might help me determine whether two thermometers' data average differ significantly from one another? Since our laboratory does not have the right equipment at the moment to compare our data with we thought that maybe performing such an analysis would be of use, however, I'm lost.
I tried performing an ANOVA analysis, but that doesn't seem to solve my question since it'll only tell if the variances of each thermometer's registered data differ significantly, but what we're trying to verify is whether their averages are significantly differing, not their variances.
I've tried to solve this problem on my own, but don't seem to find a solution to it. Could anybody suggest any statistical analysis that might help here?
We usually register 5 to 10 data for each thermometer and we analyze at least 3 thermometers each time.
This is one of our data sets for contact thermometers A, B, C, and D (In °C)
A 36.6 36.8 36.7 36.6 36.7 36.7 36.7 36.7 36.9 36.7
B 37.2 37.1 37.0 37.0 37.0 37.0 37.0 37.0 37.0 36.9
C 37.2 37.2 37.2 37.2 37.2 37.3 37.3 37.3 37.2 37.3
D 37.3 37.2 37.2 37.2 37.3 37.2 37.3 37.3 37.3 37.3
We actually have a standard thermometer and for the thermometers, it works well with, we accept a maximum of +-1°C difference. However, we can't use any of them for a while.
Dear Francisco Cid,
there are detailed procedures for analysing comparison data, a good document is
M.G. Cox: The evaluation of key comparison data Metrologia 39 589 (2002 ).
In principle you have to calculate the uncertainty of the measurements according, e.g. the following documents
https://www.bipm.org/utils/common/documents/jcgm/JCGM_100_2008_E.pdf
https://emtoolbox.nist.gov/Publications/NISTTechnicalNote1297s.pdf
and, as a first hint, the results agree, if they agree within their respective standard uncertainties.
Details of - slight - systematic deviations can be found in the document from M.G. Cox. or in
https://www.ptb.de/emrp/fileadmin/documents/nmasatue/NEW04/Papers/BPG-WP2-NEW04.pdf
Hope, these informatoins are of help
Sincerely Yours
Jürgen Hartmann
Dear Francisco
ANOVA is a compare means method, so is testing the significant difference between groups.
Regards
- Badges
- Science topic