Significant for what?? Comparison with standards or non polluted samples seems essential to me.

Agree with C.A. (Kees) Kan...

Are you asking about sampling or analysis uncertainty?

Paul Milham

One part of my research is based on the analysis of heavy metals of sewage sludge samples.

So to make the study representative (statistically representative and significant). I need to analyze how many sludge samples ??

Representative depends heavily on how much variability there is in heavy metal content within the sludge mass. So taking representative samples is of upmost importance.

If the total mass is below 100 kg, that will not be a major problem, but above that amount you have to consult experts how to perform sampling. It will also depend on the moisture content of the sludge.

Signifance is only applicable in a comparison, not in just analysing one type of sample (in my opinion).

If the sludge sample is from a single place, you can get a composite site. Then homogenise the sample, divide into 3-4 subsamples and analyse doe heavy metal.

But if the sludge are to be collected from different sites, you must put geogenic variations into consideration

Thank you Yacine, now you are getting useful responses, Paul.

I feel your question is about sample size and confidence intervals: The more samples you take, the greater your chance the upper and lower limits of the samples will encompass the "true" mean.

There are any number of statistics lectures and worksheet out there on this topic- this was first on my google search: https://online.stat.psu.edu/stat506/book/export/html/636

In essence it depends on the variability of the sludge you are sampling and how good your laboratory processes are. If well mixed, you won't need as many samples to encompass the full range (as they will all be uniform after the mixing process). If the contamination in the bulk sediment is stratified or otherwise patchy and the bulk sample not mixed adequately, the variability will be so great you might even see a bi-modal distribution and the "mean" won't actually be meaningful. (that's a statistics joke :)

i Agree with C.A. (Kees) Kan and Matthew Wheal

it is not necessary your sludge sample point has to be quantitative or a large sample point.

it is more significant if from a particular sample point, the representation and analysis of varying toxic metal ions is portrayed as per my opinion {{with very few sludge point for sample collection (less than 25 sample point is fair enough enough)}}

i would like to hear opinion of others as well

Minimum 10 with a SD or SE