6-ADV-ANAL-CHEM-sampling-Part 3.ppsx

Download as ppsx, pdf, or txt
Download as ppsx, pdf, or txt
You are on page 1of 26

Prayer of St.

Thomas
Aquinas Before
Study
Ineffable Creator, Origin of all things
who didst most wisely set among angels to govern,
and Who ordered the elements of the world,
pour forth the rays of Thine infinite brilliance
upon the darkness of my understanding.
Remove far from me the twofold darkness
in which I was born: sin and ignorance.
Do Thou, who bestows speech
upon the tongues of little children,
train my tongue and enliven my lips
with the grace of thy benediction.
Grant to me keen insight, capacity for remembering,
method and ease in learning, subtlety in interpretation,
and bountiful eloquence in speech.
Guide my beginning, direct my progress,
and set Thy seal upon the finished work,
Thou, Who lives and reignest, world without end.
AMEN
1
https://www.dspt.edu/prayer-before-study
SAMPLING PLAN:
“A predetermined procedure for the selection, withdrawal,
preservation, transportation and preparation of the portions to be
removed from a population as samples.” - IUPAC

• Intuitive sampling plan - based on incidental judgement


or previous experience
• Statistical sampling plan- provides basis for probabilstic
conclusions
• Protocol sampling plan - specified for decision-making
purpose

F Sevilla III/JH Bergantin Jr Sampling 2020


2
SAMPLING PLAN
(provides answers to these questions)

What do we want
to know?

Why do we need
this information?

What happens
to the results?

What actions
may follow?

F Sevilla III/JH Bergantin Jr Sampling 2020


3
Sequence of factors to be considered
in the design of a sampling program

1. Definition of objectives of the measurement

2. Selection of analytical methods

3. Determination of sampling locations

4. Determination of type of samples to be collected

4. Determination of number of increments

5. Determination of sampling methods


6. Selection of methods for sample preservation
and pretreatment

7. Preparation and review of final plan


F Sevilla III/JH Bergantin Jr Sampling 2020
4
IDENTIFICATION OF LOTS TO BE SAMPLED

heaps sacks cans

bulk liquid system bulk solid system bulk gas system

F Sevilla III/JH Bergantin Jr Sampling 2020


5
SIMPLE RANDOM SAMPLING

Each item in the population (or laboratory sample) has the same
probability of being selected
1. The items or sections are numbered sequentially.

2. Selection of samples is based on random numbers


(or lottery of the numbers).

Taking 5 samples from a lot of 20 cans

F Sevilla III/JH Bergantin Jr Sampling 2020


6
SIMPLE RANDOM SAMPLING
Random number generators

Fishbowl method

Shift + Ran# Alpha + RanInt

Calculator Ran# Function

F Sevilla III/JH Bergantin Jr Sampling 2020


7
https://gwcstones.com/author/alx/ https://www.youtube.com/watch?v=mxHlCR-4nGI https://www.youtube.com/watch?v=AxaWRK5qKIs
SIMPLE RANDOM SAMPLING
Random number generators Excel method

F Sevilla III/JH Bergantin Jr Sampling 2020


8
STRATIFIED RANDOM SAMPLING

The population being sampled is divided into segments (or strata)


and a simple random sample is selected from each segment.
• equal number of items per stratum
• proportional sampling: depending on the fraction
of each stratum in the bulk

Taking samples from 5 batches of 4 cans each

F Sevilla III/JH Bergantin Jr Sampling 2020


9
SYSTEMATIC RANDOM SAMPLING

Increments are taken from the bulk material at predetermined


intervals, as defined by the sampling plan
• first sample is selected at random
• succeeding samples are taken according to a
previously arranged interval, e.g. every 4th

Taking 5 samples from a lot of 20 cans

F Sevilla III/JH Bergantin Jr Sampling 2020


10
NON-PROBABILITY SAMPLING

Non-probability sampling involves selection of the sample groups by


the researchers leading to biases during sampling.

Judgement sampling - involves using knowledge about the


material to be sampled, and the reason for sampling, to select
specific samples.

Quota sampling - requires the consignment to be subdivided into


groups (similar to stratified sampling); judgement sampling is
used to select samples from each group.

Convenience sampling - involves selecting samples on the


basis of availability and/or accessibility.

F Sevilla III/JH Bergantin Jr Sampling 2020


11
SAMPLING SIZE

 How many incremental samples are to be taken?

 How big will each incremental sample be?

 How will the samples be obtained in terms of equipment,

positions of sampling and time of sampling?

F Sevilla III/JH Bergantin Jr Sampling 2020


12
STATISTICAL CONSIDERATIONS

STATISTICAL SAMPLING:
1) All particles or portion of the material should have an
equal probability of being present in the sample taken.
2) The parameter being determined during the analysis has a
normal distribution in the population.

y = 1
F
r exp { (x – )2 / 22}
e
q   2
u
e
n
c
y

x Values

F Sevilla III/JH Bergantin Jr Sampling 2020


13
ORIGIN OF SAMPLING VARIANCE

What is the probability of the randomly drawn sample have the


same composition as the bulk sample?

For a mixture composed of nA particles of type A and


nB particles of type B:

nA
probability of drawing A = p
nA + nB
=
nB
probability of drawing B = q =
nA + nB

q=1 - p

F Sevilla III/JH Bergantin Jr Sampling 2020


14
If n particles are drawn at random:

 Expected number of particles of type A = n p

 From the binomial distribution:

Standard deviation in many drawings =  n p q

Standard deviation
sn =  n p q
for sampling operation

mean μ = np

F Sevilla III/JH Bergantin Jr Sampling 2020


15
http://www.stat.yale.edu/Courses/1997-98/101/binom.htm
SAMPLE SIZE

SAMPLE SIZE - the mass, volume or dimension of the sample


- can be estimated for a homogeneous sample

Standard deviation in many drawings: sn =  n p q

npq
Relative Standard deviation: s =
np


q
s =
np

Number of particles q
n=
that must be sampled p s2

F Sevilla III/JH Bergantin Jr Sampling 2020


16
Number of particles q
= n=
that must be sampled p s2

 s is the tolerable standard deviation of the sampling.

• As s increases n decreases.

• As p increases n decreases.

soil samples

F Sevilla III/JH Bergantin Jr Sampling 2020


17
SAMPLE SIZE - the mass, volume or dimension of the
sample
Number of particles of the mixture that must be sampled:
q n = (1 - f) / (f E2)
n=
p s2
where
f is the probability of randomly choosing a discrete
particle of substance to be measured
f = n1 / (n1 + n2)
E is the tolerable standard deviation of the sampling.

• As E increases n decreases.
• As f increases n decreases.

F Sevilla III/JH Bergantin Jr Sampling 2020


18
EXERCISE 1

In a brewery, 10,000 cases of beer (24 bottles/case) were accidentally stored


longer than the recommended time with the consequence of some spoilage.
If an estimated 25% of the bottles contained spoiled beer and the sampling
error of 10% could be tolerated, what is the number of bottles which must be
randomly selected for analysis to estimate the actual number of bottles
containing spoiled beer.

F Sevilla III/JH Bergantin Jr Sampling 2020


19
Number of particles q
= n=
that must be sampled p s2

Real Samples

where, dA = density of particle A


dB = density of particle B
d = mean density of particles
PA = percentage of analyte in particle A
PB = percentage of analyte in particle A
P = average percent of analyte
σr = sampling relative standard deviation
F Sevilla III/JH Bergantin Jr Sampling 2020
20
EXERCISE 2
A column-packing material for chromatography consists of a mixture of two types of
particles. Assume that the average particle in the batch being sampled is approximately
spherical with a radius of about 0.5 mm. Roughly 20% of the particles appear to be pink
in color and are known to have about 30% by mass of a polymeric stationary phase
attached (analyte). The pink particles have a density of 0.48 g/cm3. The remaining
particles have a density of about 0.24 g/cm3 and contain little or no polymeric stationary
phase. What mass of the material should the gross sample contain if the sampling
uncertainty is to be kept below 0.5% RSD?

F Sevilla III/JH Bergantin Jr Sampling 2020


21
MINIMUM SAMPLE SIZE

Standard deviation in
sn =  n p q
many drawings:
sn npq
Relative standard deviation R=
in many drawings: np = np

R2 n = q / p

Since the number of particles n is directly proportional to the


mass m:
Sampling constant: R2 m = K s

F Sevilla III/JH Bergantin Jr Sampling 2020


22
MINIMUM SAMPLE SIZE

As the size of the sample is increased, the variability of the


samples decreases.

Sampling diagram of sodium-23 in liver homgenates.

F Sevilla III/JH Bergantin Jr Sampling 2020


23
MINIMUM NUMBER OF INCREMENTS

If the bulk material is not homogeneous, a number of replicate


sample increments must be analyzed.

Extent of uncertainty: =x  ts/n

Error: -x = ts/n=E

Number of replicates: n = t2 s2 / E2

F Sevilla III/JH Bergantin Jr Sampling 2020


24
Sampling uncertainty: sn =  n p q

Number of particles q
n=
that must be sampled p E2

Sample size: Ks = m R2

Number of replicates: n = t2 s2 / E2

F Sevilla III/JH Bergantin Jr Sampling 2020


25
EXERCISE 3 Number of replicates: n = t2 s2 / E2

A 0.7g sample of liver homogenate gives a sampling error of 7%.


How many 0.7-g samples must be analyzed to give 95%
confidence that the mean is known to ± 4%?

n = 14 samples

An iterative method is used to get the correct value of t


Initially, take df = , so that
t = 1.96 and n = 12
Then, take n = 12, so that df = 11
t = 2.209 and n = 15
Then, use n = 15, so that df = 14
t = 2.150 and n = 14
Then, use n = 14, so that df = 13
t = 2.170 and n = 14

F Sevilla III/JH Bergantin Jr Sampling 2020 26

You might also like