Understanding the Precision vs. Accuracy of Spectrofluorimeters

[Beer-monster]

I was reviewing a past paper question which asked me to describe a spectrofluorimeter, which is okay. Then it asked to explain why such an instrumental method may be precise but not accurate and I was lost.

If an instruments measurements are precise, shouldn't they be accurate?

 

[Hurkyl]

Suppose I use some device 7 times, and I get measurements of 3.01, 3.00, 2.99, 3.00, 3.01, 3.01, and 2.99. The precision of this device is pretty good right?

But the thing I'm trying to measure is really 2.54. My device is not very accurate, is it?

 

[Beer-monster]

So does that mean a spectrofluorimeter would give reasonably consistent results, but not neccessarily the right one? Is this because of the interferences associated with the method or some other problem.

Seems like an odd thing to happen in such a popular method

 

[Hurkyl]

I don't know anything about that particular device.

 

[GCT]

not always. "Accurate" is referring to how close the measurement is to the "actual" value. "Precise" is in reference to the reproducibility of the experiment, measurement after measurement values may not deviate from the average, having a low standard deviation, yet this average may be far off from the "actual" value. The analytical method employed is precise, but not accurate in this case. The spectrofluorimeter may give consistent values, and the s.d. may be relatively of low value. However, the average of the consistent measurements of the same parameter may actually be far off from the actual value, e.g. if you were to forget to subtract for the blank intensity.

 

[GCT]

Think about a spectrum obtained from a spectroscopic measurement...the linewidth and possible shifts in wavelength peaks. Linewidth may be due to random errors, but an analytical method may not be accurate if such systematic errors are not accounted for. For instance, shifts in wavelength due to solvent polarity. and as you mentioned some spectral interferences, such as emissions from another substance, may alter the accuracy of the measurements.

 

[Beer-monster]

I get it. So in the case of the spectrofluorimeter, the analytical process is affected by changes in intensity of the source and tempertaure and pH fluctuations. Therefore though the results given might be similar, these effects may make the reading different from the actual value.

Thanks guys

 

[GCT]

Fluctuations usually affect the precision, that is the standard deviation in error analysis. The things you mentioned pertain precision, although technically we wish to focus specifically on the variance in the analytical method. If I have time, I'll browse through a I. Analysis text for examples. The key here is the reproducibility of the measurements, precision, e.g. sources of noise thermal noise, electrical noise, source fluctuations etc...accuracy, for instance, has to do with the detection limit and the dynamic range (when you're ascertaining the concentration). e.g. one instrumental method may be more accurate over another in the determination of the concentration of a element/compound over a desired dynamic range.