-2
$\begingroup$

Is it okay to perform the mean and standard deviation for a triplicate extraction based on averaged values of content determination?

I performed an extraction in triplicate. Each one of the replicates was injected into the HPLC system three times and the values were averaged. Is It okay to express the results as the mean of the three averaged values (and standard deviation of them) to show the error of the replicates of extractions? My procedure is based on the literature (see the fragment of a paper attached below):

Procedure from a paper

But I don't know if it is okay to use the averaged HPLC determinations. To propagate the errors of replicates and injections could be an alternative but the literature says that it is not always feasible and the authors that make the procedure similar to mine do not look to measure the pooled variance. So, is It okay to use the averaged values as mentioned?

Improve this question
$\endgroup$

1 Answer 1

Reset to default
2
$\begingroup$

There are no hard and fast rules in HPLC determinations but rather good practices, until and unless you are doing legal analysis where the analytical requirements can be very stringent.

For routine academic work, some people make three or more injections from the same sample and express the standard deviation. This error indicates the precision of the instrument only, mainly the injection precision.

If you make three different extractions of the same plant sample and run HPLC of each in triplicate, this is also fine. You will have nine readings in total (say HPLC peak areas or heights). If you calculate the mean and the standard deviation of those nine readings, this time the standard deviation will reflect the error from the HPLC injection and your own handling of the three extractions. This approach is better and one must clearly write all the details of sample handling and calculations for the readers, though the first one is the bare minimum.

Error propagation formulae can become a nightmare for real experiments. Avoid them.

Improve this answer
$\endgroup$
1
  • $\begingroup$ Comments are not for extended discussion; this conversation has been moved to chat. $\endgroup$
    – andselisk ♦
    Commented Apr 23, 2022 at 2:34

Not the answer you're looking for? Browse other questions tagged or ask your own question.