Professional Documents
Culture Documents
Sampling and Sample Preparation
Sampling and Sample Preparation
chapter
Sampling and Sample
Preparation
Rubén O. Morawicki
Department of Food Science,
University of Arkansas,
Fayetteville, AR 72704, USA
e-mail: [email protected]
5.1 INTRODUCTION ity, of the final result. This final result depends on the
cumulative errors at each stage that are usually
Quality attributes in food products, raw materials, or described by the variance [3, 4]. Variance is an esti-
ingredients are measurable characteristics that need mate of the uncertainty. The total variance of the whole
monitoring to ensure specifications are met. Some testing procedure is equal to the sum of the variances
quality attributes can be measured online by using associated with each step of the sampling procedure
specially designed sensors and results obtained in real and represents the precision of the process. Precision
time (e.g., color of vegetable oil in an oil extraction is a measure of the reproducibility of the data. In con-
plant). However, in most cases quality attributes are trast, accuracy is a measure of how close the data are to
measured on small portions of material that are taken the true value. The most efficient way to improve accu-
periodically from continuous processes or on a certain racy is to improve the reliability of the step with the
number of small portions taken from a lot. The small greatest variance, and that is frequently the initial sam-
portions taken for analysis are referred to as samples, pling step. Attention is often given to the precision and
and the entire lot or the entire production for a certain accuracy of analytical methods, with less attention
period of time, in the case of continuous processes, is given to the validity of sampling and sample prepara-
called a population. The process of taking samples tion. However, sampling can often be the greatest
from a population is called sampling. If the procedure source of error in chemical analysis in general, but
is done correctly, the measurable characteristics especially a problem in food analysis because of the
obtained for the samples become a very accurate esti- food matrix [5–7] (Fig. 5.1). Sample homogenization
mation of the population. and other aspects of sample preparation are additional
By sampling only a fraction of the population, a sources of potential error, even prior to the actual
quality estimate can be obtained accurately, quickly, analysis.
and with less expense and personnel time than if the As you read each section of the chapter, consider
total population were measured. The reliability of application of the information to some specific exam-
sampling is dependent more on the sample size than ples of sampling needs in the food industry: sampling
on the population size [1]. The larger the sample size, for nutrition labeling (see Study Question 7 in this
the more reliable the sampling. However, sample size chapter), pesticide analysis (see also Chap. 33, Sect.
is limited by time, cost, sampling methods, and the 33.3), mycotoxin analysis (see also Chap. 33, Sect. 33.4),
logistics of sample handling, analysis, and data pro-
cessing. Moreover, in the case of food products, ana-
lyzing a whole population would be practically
impossible because of the destructive nature of most
analytical methods. Paradoxically, estimated parame-
ters using representative samples (discussed in
Sects. 5.2 and 5.3) are normally more accurate than the
Relative standard
(census).
The sample actually analyzed in the laboratory can
be of any size or quantity [2]. This laboratory sample
1
has generally undergone preparation such as homoge-
nization or grinding to prepare it for analysis and is 2
5.2.2.2 Nature of Population and Product sampling plan, is collecting samples to measure fungal
One must clearly define the population and under- toxins, named mycotoxins, in food systems. Mycotoxins
stand the nature of the product that is going to be sam- are distributed broadly and randomly within a popula-
pled to select an appropriate sampling plan. tion and a normal distribution cannot be assumed [2].
Populations for sampling are often defined in terms of Such distribution requires a combination of many ran-
being homogeneous or heterogeneous and being domly selected portions to obtain a reasonable estimate
either discrete or continuous. The population and of mycotoxin levels. Methods of analysis that are
product can also vary greatly in size. extremely precise are not needed when determining
The ideal population and product would be uni- mycotoxin levels, when sampling error is many times
form throughout and identical at all locations. Such a greater than analytical error [2]. In this case, sampling
population would be homogeneous. Sampling from and good comminution and mixing prior to particle
such a population is simple, as a sample can be taken size reduction are more important than the chemical
from any location, and the analytical data obtained analysis itself. Additional information on sampling for
will be representative of the whole. However, this mycotoxin analysis is provided in Chap. 33, Sect. 33.4.
occurs rarely, as even in an apparently uniform
product, such as sugar syrup, suspended particles and 5.2.2.3 Nature of Test Method
sediments in a few places may render the population Procedures used to test samples collected vary in sev-
heterogeneous. In fact, most populations and products eral characteristics that help determine the choice of
that are sampled are heterogeneous. Therefore, the sampling plan, e.g., cost, speed, accuracy, precision,
location within a population where a sample is taken and destructive vs. nondestructive. Low-cost, rapid,
will affect the subsequent data obtained. However, nondestructive tests that are accurate and precise
sampling plans and sample preparation can make the make it more feasible to analyze many samples.
sample representative of the population or take het- However, limitations on any of these characteristics
erogeneity into account in some other way. If the sam- will make the nature of the test method a more impor-
ple is heterogeneous, some additional questions tant determinant of the sampling plan.
become important [6]: What is the nature of the varia-
tion? Should samples be pooled or replicated? Should
different portions of the sample be analyzed sepa- 5.3 TYPES OF SAMPLING PLANS
rately? Should the surface of the sample be tested?
5.3.1 ampling by Attributes and Sampling
S
Sampling from discrete (or compartmentalized)
by Variables
populations is relatively easy, since the population is
split into multiple separate subunits (e.g., cans in a pallet Sampling plans are designed for examination of either
of canned food, boxes of breakfast cereal in a truck, bottle attributes or variables [4]. In attribute sampling, sam-
of juice on a conveyer belt). The choice of the sampling pling is performed to decide on the acceptability of a
plan is determined in part by the number and size of the population based on whether the sample possesses a
individual subunits. Sampling is more difficult from a certain characteristic or not. The result has a binary
continuous population since different parts of the sam- outcome of either conforming or nonconforming.
ple are not physically separated (e.g., potato chips on a Sampling plans by attributes are based on the hyper-
conveyer belt, oranges in a semitruck) [6]. geometric, binomial, or Poisson statistical distribu-
The population may vary in size from a production tions. In the event of a binomial distribution (e.g.,
lot, a day’s production, to the contents of a warehouse. presence of Clostridium botulinum), the probability of a
Information obtained from a sample of a particular pro- single occurrence of the event is directly proportional
duction lot in a warehouse must be used strictly to make to the size of the sample, which should be at least ten
inferences about that particular lot, but conclusions can- times smaller than the population size. Computing
not be extended to other lots in the warehouse. binomial probabilities will allow the investigator to
To use food analysis to solve problems in the food make inferences on the whole lot.
industry, it may be inadequate to focus just on the In variable sampling, sampling is performed to
nature of the population and product to develop a estimate quantitatively the amount of a substance (e.g.,
sampling plan. For example, to collect samples to protein content, moisture content, etc.) or a characteris-
address a problem of variation of moisture content of tic (e.g., color) on a continuous scale. The estimate
packaged products coming off a production line, it obtained from the sample is compared with an accept-
would be important to examine each processing step. able value (normally specified by the label, regulatory
It would be necessary to understand where variation agencies, or the customer) and the deviation measured.
likely exists to then determine how to appropriately This type of sampling usually produces data that have
collect samples and data to analyze for variability. a normal distribution such as in the percent fill of a con-
One of the most challenging cases of accounting tainer and total solids of a food sample. In general,
for the nature of the population, as it influences the variable sampling requires smaller sample size than
66 R.O. Morawicki
The consumer’s risk describes the probability of mixture of product from small containers that is suffi-
accepting a poor-quality population. This should hap- cient for the testing of a single unit” [15]. Furthermore,
pen rarely (<5 % of the lots), but the actual acceptable a sampling plan is specified for containers of specific
probability (β) of a consumer risk depends on the con- net weights. The container size is determined by the
sequences associated with accepting an unacceptable size of the lot. A specific number of containers must be
lot. These may vary from major health hazards and filled for sampling of each lot size. The lot is rejected if
subsequent fatalities to a lot being of slightly lower the number of defective units exceeds the acceptable
quality than standard lots. Obviously, the former limit. For example, out of a lot containing 48001–84000
demands a low or no probability of occurring whereas units, each weighing 1 kg or less, 48 samples should be
the latter would be allowed to occur more frequently. selected. If six or more of these units fail to conform to
The producer risk is the probability of rejecting (α) an the attribute of interest, the lot will be rejected. Based
acceptable product. As with consumer’s risk, the con- on statistical confidence intervals, this sampling plan
sequences of an error determine the acceptable proba- will reject 95 % of the defective lots examined, that is,
bility of the risk. An acceptable probability of 5 % consumer risk [15].
producer’s risk is usually 5–10 %. Further discussion of
sampling plans can be found in the following section. 5.4.3 Manual Versus Continuous Sampling
To obtain a manual sample, the person taking the sample
5.4 SAMPLING PROCEDURES must attempt to take a “random sample” to avoid human
bias in the sampling method. Thus, the sample must be
5.4.1 Introduction taken from a number of locations within the population
to ensure it is representative of the whole population. For
The reliability of analytical data is compromised if liquids in small containers, this can be done by shaking
sampling is not done properly. As shown in Table 5.1, prior to sampling. When sampling from a large volume
the use of the data to be obtained is one major factor of liquid, such as that stored in silos, aeration ensures a
determining the sampling procedure. Details for the homogeneous unit. Liquids may be sampled by pipet-
sampling of specific food products are described in the ting, pumping, or dipping. However, when sampling
Official Methods of Analysis of AOAC International [14] grain from a rail car, mixing is impossible, and samples
and in the Code of Federal Regulations (CFR) [15]. Two are obtained by probing from several points at random
such examples for specific foods follow. within the rail car. Such manual sampling of granular or
powdered material is usually achieved with triers or
5.4.2 Examples probes that are inserted into the population at several
The AOAC Method 925.08 [14] describes the method locations. Errors may occur in sampling [10], as rounded
for sampling flour from sacks. The number of sacks to particles may flow into the sampling compartments
be sampled is determined by the square root of the more easily than angular ones. Similarly, hygroscopic
number of sacks in the lot. The sacks to be sampled are materials flow more readily into the sampling devices
chosen according to their exposure. The samples that than does nonhygroscopic material. Horizontal core
are more frequently exposed are sampled more often samples have been found to contain a larger proportion
than samples that are exposed less. Sampling is done of smaller-sized particles than vertical ones [16].
by drawing a core from a corner at the top of the sack Continuous sampling is performed mechanically.
diagonally to the center. The sampling instrument is a Figure 5.4 shows an automatic sampling device that is
cylindrical, polished trier with a pointed end. It is used to take liquid samples from a continuous produc-
13 mm in diameter with a slit at least one third of the tion line. Continuous sampling should be less prone to
circumference of the trier. A second sample is taken human bias than manual sampling.
from the opposite corner in a similar manner. The
cores are stored for analysis in a clean, dry, airtight 5.4.4 Statistical Considerations
container that has been opened near the lot to be sam- 5.4.4.1 Probability Sampling
pled. The container should be sealed immediately Probability sampling plans prescribe the selection of
after the sample is added. A separate container is used a sample from a population based on chance. It pro-
for each sack. Additional details regarding the con- vides a statistically sound basis for obtaining represen-
tainer and the procedure also are described below. tative samples with elimination of human bias [2]. The
Title 21 CFR specifies the sampling procedures probability of including any item in the sample is
required to ensure that specific foods conform to the known and sampling error can be calculated. Several
standard of identity. In the case of canned fruits, 21 probability sampling methods are available to the
CFR 145.3 defines a sample unit as “container, a por- researcher, and the most common ones are described
tion of the contents of the container, or a composite in the next few paragraphs.
68 R.O. Morawicki
in AOAC Method 920.175 [14]. The method prescribes equipment with an integrated cooling system that per-
that the sugar should be ground, if necessary, and form the cryogenic freezing and grinding automatically.
mixed to uniformity. Raw sugars should be mixed
thoroughly and rapidly with a spatula. Lumps are to 5.5.2.2 Applications for Grinding Equipment
be broken by a mortar and pestle or by crushing with Mills differ according to their mode of action, being
a glass or iron rolling pin on a glass plate. classified as a burr, hammer, impeller, cyclone,
impact, centrifugal, or roller mill [19]. Methods for
5.5.2 Grinding grinding dry materials range from a simple pestle and
5.5.2.1 Introduction mortar to power-driven hammer mills. Hammer mills
Grinding is important both for sample preparation wear well and they reliably and effectively grind cere-
prior to analysis and for food ingredient processing. als and dry foods, while small samples can be finely
Various mills are available for reducing particle size to ground by ball mills. A ball mill grinds by rotating the
achieve sample homogenization [17]. To homogenize sample in a container that is half filled with ceramic
moist samples, bowl cutters, meat mincers, tissue balls. This impact grinding can take hours or days to
grinders, mortars and pestles, or blenders are used. complete. A chilled ball mill can be used to grind fro-
However, mortars and pestles and mills are best for zen foods without predrying and also reduces the like-
dry samples. Some foods are more easily ground after lihood of undesirable heat-initiated chemical reactions
drying in a desiccator or vacuum oven. Grinding wet occurring during milling [19]. Alternatively, dry mate-
samples may cause significant losses of moisture and rials can be ground using an ultracentrifugal mill by
chemical changes. In contrast, grinding frozen sam- beating, impacting, and shearing. The food is fed from
ples reduces undesirable changes. The grinding pro- an inlet to a grinding chamber and is reduced in size
cess should not heat the sample, and therefore the by rotors. When the desired particle size is obtained,
grinder should not be overloaded because heat will be the particles are delivered by centrifugal force into a
produced through friction. For especially heat-sensi- collection pan [19]. Large quantities can be ground
tive sample, grinders can be cooled with liquid nitro- continuously with a cyclone mill.
gen and then ground samples are stored at
−80 °C. Contact of food with bare metal surfaces 5.5.2.3 Determination of Particle Size
should be avoided if trace metal analysis is to be per- Particle size is controlled in certain mills by adjusting
formed [18]. the distance between burrs or blades or by screen
To break up moist tissues, a number of slicing mesh size/number. The mesh number is the number
devices are available: bowl cutters can be used for fleshy of square screen openings per linear inch of mesh. The
tubers and leafy vegetables, while meat mincers may be final particles of dried foods should be 20 mesh for
better suited for fruit, root, and meat [19]. Addition of moisture, total protein, or mineral determinations.
sand as an abrasive can provide further subdivision of Particles of 40 mesh size are used for extraction assays
moist foods. Blenders are effective in grinding soft and such as lipid and carbohydrate estimation.
flexible foods and suspensions. Rotating knives In addition to reducing particle size for analysis of
(25000 rpm) will disintegrate a sample in suspension. In samples, it also is important to reduce the particle size
colloidal mills, a dilute suspension is flowed under of many food ingredients for use in specific food prod-
pressure through a gap between slightly serrated or ucts. For example, rolled oats for a grain-based snack
smooth-surfaced blades until they are disintegrated by bar may have a specified granulation size described as
shear. Sonic and supersonic vibrations disperse foods in 15 % of the oats maximum passes through a #7 US
suspension and in aqueous and pressurized gas solu- standard sieve. A higher granulation (i.e., more smaller
tion. The Mickle disintegrator sonically shakes suspen- particles) would mean more fines and less whole oats
sions with glass particles, and the sample is in the finished bar. This would result in higher inci-
homogenized and centrifuged at the same time [19]. dences of snack bar breakage.
Alternatively, a low shear continuous tissue homoge- There are a variety of methods for measuring par-
nizer is fast and handles large volumes of sample. ticle size, each suited for different materials. The sim-
Another alternative is cryogenic grinding or plest way to measure particle sizes of dry materials of
cryogrinding. This method is ideal for biological sam- less than 50 μm in diameter is by passing the sample
ples and materials that are sensitive to oxygen or tem- through a series of vertically stacked sieves with
perature. However, most materials are suitable for this increasing mesh number. As the mesh number
technique. Cryogrinding can be performed manually increases, the apertures between the mesh are smaller
with a mortar and pestle after freezing the sample with and only finer and finer particles pass through subse-
liquid nitrogen. The mortar and pestle have to be pre- quent sieves (see Table 5.2). Sieve sizes have been spec-
chilled with liquid nitrogen before adding the material. ified for salt, sugar, wheat flour, cornmeal, semolina,
Also, there are several brands of specialized grinding and cocoa. The sieve method is inexpensive and fast,
72 R.O. Morawicki
5 . 2 US standard mesh with equivalents in
table inches and millimeters
Sieve opening
US standard mesh Inches Millimeters
4 0.1870 4.760
6 0.1320 3.360
7 0.1110 2.830
8 0.0937 2.380
10 0.0787 2.000
12 0.0661 1.680
14 0.0555 1.410
16 0.0469 1.190
18 0.0394 1.000
20 0.0331 0.841
30 0.0232 0.595
40 0.0165 0.400
50 0.0117 0.297
60 0.0098 0.250
70 0.0083 0.210
80 0.0070 0.177 5.6
W.S. Tyler ®
Ro-Tap® Sieve Shaker (Courtesy of
figure
W.S. Tyler®
, Mentor, OH)
100 0.0059 0.149
120 0.0049 0.125
140 0.0041 0.105 with an unchanged wavelength, at a different angle,
170 0.0035 0.088 and generally with a lower intensity. Some of the light
200 0.0029 0.074
can be absorbed by the particle and re-irradiated at a
230 0.0024 0.063
270 0.0021 0.053
different frequency, which can display as fluorescence
325 0.0017 0.044 or heat. The last interaction, diffraction, is the result of
400 0.0015 0.037 light interacting with the edges of the particle. This
interaction makes the light spread out, or scatter, and
produce wave fronts around the particle that act as sec-
but it is not suitable for emulsions or very fine pow- ondary spherical waves. This phenomenon is similar
ders [20]. A standard in the industry is the W.S. Tyler® to Young’s double-slit experiment, which readers can
Ro-Tap® Sieve Shaker (Fig. 5.6). find in a simple Google search.
To obtain more accurate size data for smaller par- The secondary waves generated at the edge of the
ticles (<50 μm), characteristics that correlate to size particles interact with each other, in some areas con-
are measured, and thus size is measured indirectly structively and others destructively, thus producing
[21]. Surface area and zeta potential (electrical charge interference patterns that correlate with the particle
on a particle) are characteristics that are commonly size. In general, larger particles scatter light at smaller
used. Zeta potential is measured by an electroacous- angles and with a higher intensity. Smaller particles,
tic method whereby particles are oscillated in a high- on the other side, scatter light at wider angles and with
frequency electrical field and generate a sound wave less intensity. These scatter patterns can be correlated
whose amplitude is proportional to the zeta poten- with particle size by using mathematical algorithms.
tial. Optical and electron microscopes are routinely Besides size, shape and optical properties affect
used to measure particle size. Optical microscopes the angle and intensity of the scattered light. When
are interfaced with video outputs and video-imaging particles are capable of transmitting light, the refrac-
software to estimate size and shape. The advantage tive index plays a role in particle size determination.
of the visual approach is that a three-dimensional Therefore, light scattering instruments take advantage
size and detailed particle structure can be observed. of the refractive index to improve accuracy, especially
A widespread technique to measure particle size for small particles.
distribution uses a principle known as light scattering A typical laser diffraction particle analyzer has a
or laser diffraction. When a coherent light source, sample port, a medium for the sample transport, a
such as laser beam, is pointed to a particle, four inter- flow cell, a laser beam, a detector, electronics, and
actions of the beam with the particle can take place: software. The sample is introduced into the port,
reflection, refraction, absorption, and diffraction. which contains some dispersion system to avoid par-
Reflection is the portion of light that is rejected by the ticle agglomeration. The sample is then carried by a
particle. Refraction is the light that goes through trans- liquid, for wet samples, or compressed air, for dry
parent or translucent materials and exits the particle samples, to a flow cell. As the sample goes through
Chapter 5 • Sampling and Sample Preparation 73
5 . 7 Basics of an
Scattered
image
figure
instrument to
measure particle
size distribution
by light Particles
scattering (laser
diffraction)
(Adapted from
Shimadzu
Corporation
[22])
Detecting
plane
the flow cell, a laser beam irradiates the sample. This For particles in the nanoscale range, a widely used
results in diffraction patterns that after passing technique for particle analysis is dynamic light scatter-
through a condensing lens are projected on a detector ing. These instruments determine particle size and even
where the diffraction angles and intensities are mea- the molecular weight of large molecules in solution.
sured. This information is then transformed into a This is achieved by measurement of frequency shifts of
particle size distribution by the instrument electron- light scattered by particles due to Brownian motion.
ics and a computer software algorithm (Fig. 5.7). Understanding of the principles of the instrument
Results are displayed as frequency distributions with used to obtain size data is vital to appreciate the limita-
particle size on the x-axis and frequency on the y-axis. tions of each method. For example, data obtained from
Determination of particle size by light scattering the same sample using sieves and light scattering will
has many advantages, such as: differ [21]. Sieves separate particles using square holes,
and therefore they distinguish size in the smallest
• Wide dynamic range—from the nanometer to the dimension, independent of shape. However, light scat-
several millimeters range tering techniques assume that the particle is spherical,
• Suitable for wet and dry materials—such as solid in and data are derived from the average of all dimen-
liquid suspensions, dry powders, liquid-liquid sions. Particle size measurement is useful to maintain
emulsions, and pastes sample quality, but care must be taken in choosing an
• Fast measurement—in the order of seconds appropriate method and interpreting the data.
• No calibration needed
• Easy to use and interpret results 5.5.3 Enzymatic Inactivation
• Proficient at detecting mixes of large and small
Food materials often contain enzymes that may
particles
degrade the food components being analyzed. Enzyme
• Technique covered by ISO standard 13320:2009
activity therefore must be eliminated or controlled
using methods that depend on the nature of the food.
The most important disadvantages are the
Heat denaturation to inactivate enzymes and freezer
following:
storage (−20 to −30 C) for limiting enzyme activity are
common methods. However, some enzymes are more
• Assumption that particles are spherical, which
effectively controlled by changing the pH or by salting
introduced error in particles that deviates a lot from
out [19]. Oxidative enzymes may be controlled by
sphericity.
adding reducing agents.
• Particles agglomeration.
• Results are expressed in volume basis, which gives
5.5.4 Lipid Oxidation Protection
more weight to large particles.
• Results cannot be compared with other size deter- Lipids present particular problems in sample prepara-
mination methods. tion. High-fat foods are difficult to grind and may
74 R.O. Morawicki
need to be ground while frozen. Unsaturated lipids are quality product, while the vendor risk is the probabil-
sensitive to oxidative degradation and should be pro- ity of rejecting an acceptable product. An acceptable
tected by storing under nitrogen or vacuum. probability of risk depends on the seriousness of a
Antioxidants may stabilize lipids and may be used if negative consequence.
they do not interfere with the analysis. Light-initiated Sampling plans are determined by whether the
photooxidation of unsaturated lipids can be avoided population is homogeneous or heterogeneous.
by controlling storage conditions. In practice, lipids Although sampling from a homogeneous population
are more stable when frozen in intact tissues rather is simple, it rarely is found in practical industrial situ-
than as extracts [19]. Therefore, ideally, unsaturated ations. Sampling from heterogeneous populations is
lipids should be extracted just prior to analysis. Low- most common, and suitable sampling plans must be
temperature storage is generally recommended to pro- used to obtain a representative sample. Sampling
tect most foods. methods may be manual or continuous. Ideally, the
sampling method should be statistically sound.
5.5.5 Microbial Growth and Contamination However, nonprobability sampling is sometimes
unavoidable, even though there is not an equal prob-
Microorganisms are present in almost all foods and
ability that each member of the population will be
can alter the sample composition. Likewise, microor-
selected due to the bias of the person sampling.
ganisms are present on all but sterilized surfaces, so
Probability sampling is preferred because it ensures
sample cross-contamination can occur if samples are
random sampling and is a statistically sound method
not handled carefully. The former is always a problem
that allows calculation of sampling error and the prob-
and the latter is particularly important in samples for
ability of any item of the population being included in
microbiological examination. Freezing, drying, and
the sample.
chemical preservatives are effective controls and often
Each sample must be clearly marked for identifi-
a combination of these is used. The preservation meth-
cation and preserved during storage until completion
ods used are determined by the probability of contam-
of the analysis. Official and legal samples must be
ination, the storage conditions, storage time, and the
sealed and a chain of custody maintained and identi-
analysis to be performed [19].
fied. Often, only a portion of the sample is used for
analysis and sample size reduction must ensure that
the portion analyzed is representative of both the sam-
5.6 SUMMARY
ple and population. Sample preparation and storage
should account for factors that may cause sample
Food quality is monitored at various processing stages
changes. Samples can be preserved by limiting enzyme
but 100 % inspection is rarely possible or even desir-
activity, preventing lipid oxidation, and inhibiting
able. To ensure a representative sample of the popula-
microbial growth/contamination.
tion is obtained for analysis, sampling and sample
reduction methods must be developed and imple-
mented. The selection of the sampling procedure is
5.7 STUDY QUESTIONS
determined by the purpose of the inspection, the
nature of the population and product, and the test
method. Increasing the sample size will generally 1. As part of your job as supervisor in a quality
increase the reliability of the analytical results, and assurance laboratory, you need to give a new
using t-test techniques will optimize the sample size employee instruction regarding choosing a
necessary to obtain reliable data. Multiple sampling sampling plan. Which general factors would
techniques also can be used to minimize the number of you discuss with the new employee? Distinguish
samples to be analyzed. Sampling is a vital process, as between sampling for attributes vs. sampling
it is often the most variable step in the entire analytical for variables. Differentiate the three basic sam-
procedure. pling plans and the risks associated with select-
Sampling may be for attributes or variables. ing a plan.
Attributes are monitored for their presence or absence, 2. Your supervisor wants you to develop and
whereas variables are quantified on a continuous implement a multiple sampling plan. What
scale. Sampling plans are developed for either attri- would you take into account to define the
butes or variables and may be single, double, or mul- acceptance and rejection lines? Why?
tiple. Multiple sampling plans reduce costs by rejecting 3. Distinguish probability sampling from non-
low-quality lots or accepting high-quality lots quickly, probability sampling. Which is preferable and
while intermediate-quality lots require further sam- why?
pling. There is no sampling plan that is risk-free. The 4. (a) Identify a piece of equipment that would be
consumer risk is the probability of accepting a poor- useful in collecting a representative sample for
Chapter 5 • Sampling and Sample Preparation 75
analysis. Describe precautions to be taken to 3. Harris DC (2015) Quantitative chemical analysis, 9th edn.
ensure that a representative sample is taken and W.H. Freeman and Co., New York
a suitable food product that could be sampled 4. Miller JC (1988) Basic statistical methods for analytical
chemistry. I. Statistics of repeated measurements. A
with this device. (b) Identify a piece of equip- review. Analyst 113: 1351–1355
ment that would be useful for preparing a sam- 5. Lichon, MJ (2000) Sample preparation for food analysis,
ple for analysis. What precautions should be general. In: Meyers RA (ed) Encyclopedia of analytical
taken to ensure that the sample composition is chemistry: applications, theory and instrumentation.
not changed during preparation? John Wiley, New York
5. For each of the problems identified below that 6. Lichon MJ (2004) Sample preparation, ch. 44. In: Nollet
LML (ed) Handbook of food analysis, 2nd edn. Vol 3.
can be associated with collection and prepara-
Marcel Dekker, New York, pp. 1741–1755
tion of samples for analysis, state one solution 7. Mälkki Y (1986) Collaborative testing of methods for food
for how the problem can be overcome: analysis. J Assoc Off Anal Chem 69(3):403–404
8. FDA (2016) Sampling to protect the food supply (last
(a) Sample bias
updated 5/5/2016) http://www.fda/gov/Food/
(b) Change in composition during storage of ComplianceEnforcement/Sampling/ucm20041972.htm
sample prior to analysis 9. Pehrsson PR, Haytowitz DB, Holden JM, Perry, CR,
(c) Metal contamination in grinding Beckler DG (2000) USDA’s national food and nutrient
(d) Microbial growth during storage of product analysis program: food sampling. J Food Comp Anal.
prior to analysis 13:379–389
10. Pehrsson P, Perry C, Daniel M (2013) ARS, USDA updates
food sampling strategies to keep pace with demographic
6. The instructions you are following for cereal shifts. Procedia Food Sci 2:52–59
protein analysis specify grinding a cereal sam- 11. IUPAC (1997) Compendium of chemical terminology,
ple to 10 mesh before you remove protein by a 2nd edn. (the “Gold Book”). Compiled by McNaught
series of solvent extractions: AD Wilkinson A. Blackwell Scientific, Oxford. XML on-
line corrected version: http://goldbook.iupac.org
(a) What does 10 mesh mean? (2006-) created by Nic M, Jirat J, Kosata B; updates com-
(b) Would you question the use of a 10-mesh piled by Jenkins A. ISBN 0–9678550–9–8. doi:10.1351/
screen for this analysis? Provide reasons for goldbook
your answer. 12. Weiers RM (2007) Introduction to business statistics, 6th
edn. South-Western College, Cincinnati, OH
13. NIST/SEMATECH (2013) e-Handbook of statistical
7. You are to collect and prepare a sample of cereal methods, chapter 6: process or product monitoring and
produced by your company for the analyses control. http://www.itl.nist.gov/div898/handbook/
required to create a standard nutritional label. 14. AOAC International (2016) Official methods of analysis,
Your product is considered “low fat” and “high 20th edn., (On-line). AOAC International, Rockville, MD
fiber” (see information on nutrient claims and 15. Anonymous (2016) Code of federal regulations. Title 21.
FDA compliance procedures in Chap. 3). What US Government Printing Office, Washington, DC
16. Baker WL, Gehrke CW, Krause GF (1967) Mechanism of
kind of sampling plan will you use? Will you do
sampler bias. J Assoc Off Anal Chem 50: 407–413
attribute or variable sampling? What are the 17. Anonymous (2016) Code of federal regulations. 21 CFR
risks associated with s ampling in your specific 101.9 (g), 9 CFR 317.309 (h), 9 CFR 381.409 (h). US
case? Would you use probability or nonproba- Government Printing Office, Washington, DC
bility sampling, and which specific type would 18. Pomeranz Y, Meloan CE (1994) Food analysis: theory and
you choose? What specific problems would you practice, 3rd edn. Chapman & Hall, New York
anticipate in sample collection, storage, and 19. Cubadda F, Baldini M, Carcea M, Pasqui LA, Raggi A,
Stacchini P (2001) Influence of laboratory homogeniza-
preparation? How would you avoid or mini-
tion procedures on trace element content of food samples:
mize each of these problems? an ICP-MS study on soft and durum wheat. Food Addit
Contam 18: 778–787
20. Kenkel JV (2003) Analytical chemistry for technicians,
REFERENCES 3rd edn. CRC, Boca Raton, FL
21. Jordan JR (1999) Particle size analysis. Inside Lab Manage
1. Horwitz W (1988) Sampling and preparation of samples 3(7): 25–28
for chemical examination. J Assoc Off Anal Chem 71: 22. Shimadzu Corporation (2016) Particle Size Distribution
241–245 Calculation Method. Lecture on Partical Analysis –
2. Puri SC, Ennis D, Mullen K (1979) Statistical quality con- Hands-on Course http://www.shimadzu.com/an/pow-
trol for food and agricultural scientists. G.K. Hall and der/support/practice/p01/lesson22.html [Accessed on
Co., Boston, MA June 29, 2016]