Class Notes (834,153)
Chemistry (266)
ENCH 213 (25)
Lecture

# 5sampling statistics.pdf

4 Pages
49 Views

School
Department
Chemistry
Course
ENCH 213
Professor
Diane Beauchemin
Semester
Fall

Description
Sampling statistics Sampling considerations • Lot: total material • Gross or bulk sample: representative of the lot • Laboratory sample: with exact same composition as the bulk sample • Test portions Select the one best answer: aAbulk sample is taken from a lot. b Abulk sample is taken from a laboratory sample. cAbulk sample is also called a gross sample. d a and c Statistics of sampling 2 2 2 • s o s +ss a ◦ s = overall variance ◦ s = variance of the sampling process 2 ◦ s = aariance of the analytical procedure • if sa< s s3 ◦ loss of time to try to reduce s (maan source of error is then the sampling error) ◦ s obaained from replicate measurements of a standard ◦ s oboained from replicate measurements of the samples Gross sample size depends on: • Uncertainty between the composition of the gross sample and that of the lot • Heterogeneity of the sample • Particle size distribution • Analyte concentration • Homogeneous solutions of liquids and gases: ◦ heterogeneity and particle size at the molecular level ◦ sample size can be relatively small ◦ depends on final use Particulate solids • Simplify by treating the sample as 2-component mixtures with a single particle size ◦ soil mixture= 367 Smartees + 554 Reeces pieces nA nB • p = probability of drawing a Smartee = n /(n +A ) =A0.3B8 • q = probability of drawing a Reeces piece = n /(n +nB) =A0.6B2 = 1-p Small sample size • small sample size ≈ 15 particles ◦ expected number of Smartees = np = 15 x 0.398 = 6.0 ◦ expected number of Reeces pieces = nq = 15 x 0.602 = 9.0 • standard deviation in sampling operation (i.e., of many drawings) = s = (npq)½ =n1.9 ◦ applies to both kinds of particles relative errors ◦ 32% of expected number of Smartees ◦ 21% of expected number of Reeces pieces Large sample size • large sample size ≈ 70 particles ◦ expected number of Smartees = np = 70 x 0.398 = 28 ◦ expected number of Reeces pieces = nq = 70 x 0.602 = 42 • standard deviation in sampling operation (i.e., f many drawings) = s = (npq)½ n 4.1 ◦ applies to both kinds of particles ◦ 15% of expected number of Smartees ◦ 10% of expected number of Reeces pieces How many particles for representative sampling with σ ≤ 4%? r • relative standard deviation σ = sr/n =n(npq) /n = (pq/n) ½ ½ ◦ small sample: (0.398 x 0.602/15) = 0.13 or 13% ◦ large sample: (0.398 x 0.602/70) = 0.059 or 5.9% • to decrease it to 4% ◦ σ = r /n n (npq) /n = (pq/n) ½ 1/2 ◦ 0.04 = (0.398 x 0.602 /n) → solve for n ◦ n = 0.398 x 0.602 / (0.04) = 1502 Establishing and using a sampling constant • because sample mass (m) α n, and n σ = pq: r 2 ◦ mσ = K r sampsing constant = mass of sample producing a relative sampling standard
More Less

Related notes for ENCH 213
Me

OR

Join OneClass

Access over 10 million pages of study
documents for 1.3 million courses.

Join to view

OR

By registering, I agree to the Terms and Privacy Policies
Just a few more details

So we can recommend you notes for your school.