FOOD2000 Chap.2 Reliable Measurement and Laboratory Methods
Reliable Measurement and Laboratory Methods
Reliable food measurements require a representative sample, an appropriate method, controlled conditions and an honest treatment of variation. This chapter distinguishes accuracy from precision, random variation from systematic bias, and repeat readings from independent replication.
It also develops a disciplined serial-dilution and viable-count workflow in which every dilution, plated volume, unit and validity condition remains visible. Calibration, blanks and uncertainty are treated as parts of the inference rather than paperwork added after calculation.
What this chapter covers
- 01
Representative sampling and sample preparation
- 02
Accuracy, precision and measurement error
- 03
Calibration, standards and blanks
- 04
Repeatability, reproducibility and replication
- 05
Serial dilution and viable counts
- 06
Units, significant figures and conclusion boundaries
Worked example · free
Reverse dilution and plated volume
- +1Record 64 colonies, dilution 10^-4 and plated volume 0.1 mL.
- +1Reverse the dilution by multiplying by 10^4.
- +1Convert the 0.1 mL plate to a 1 mL basis by multiplying by 10.
- +1Report 6.4 x 10^6 CFU/mL and state assumptions about mixing and plate validity.
Key terms
- Accuracy
- Closeness to an accepted reference value.
- Precision
- Closeness among repeated measurements.
- Systematic bias
- A consistent shift in results caused by a method, instrument or design feature.
- CFU
- Colony-forming unit, an operational estimate of viable organisms able to form colonies under the stated method.
Reliable Measurement and Laboratory Methods FAQ
Can a measurement be precise but inaccurate?
Yes. Repeated values can cluster tightly while all remain displaced from the reference because of calibration error, extraction loss or another systematic bias. Precision describes agreement among repeats, not truth. Accuracy requires appropriate standards, method validation and checks that reveal consistent displacement.
Why is a technical repeat not independent replication?
Repeated readings of the same prepared sample mainly assess measurement repeatability. They do not reproduce sample selection, biological material or processing treatment. Independent replication begins at the experimental-unit level and therefore captures the variation needed to support a broader treatment comparison.
What must accompany a viable-count result?
Record the count, dilution, plated volume, final unit and assumptions behind the selected plate. The numerical back-calculation is not enough. Mixing, sampling, incubation conditions, colony overlap and the fact that one colony can arise from more than one cell all affect interpretation.
What makes a calibration range appropriate?
Standards should bracket the expected sample response and produce a relationship suitable for the method, with a blank and an independent check where appropriate. Extrapolating far beyond the highest standard turns the displayed number into an unsupported estimate.
Dilute or otherwise prepare an out-of-range sample and remeasure it within the validated region, while carrying the preparation factor into the final result and units.
Exam move
Draw a measurement chain from lot sampling through preparation, instrument or plating, calculation and decision. At each step label one source of random variation and one possible bias. Practise dilution calculations as unit-bearing factor chains. For experimental questions, identify the true experimental unit and separate independent replicates from repeat readings.
Finish every answer by stating what the method does and does not justify. Keep an error notebook organised by stage rather than instrument. Under sampling, record heterogeneity, selection and storage risks. Under preparation, record transfer loss, contamination and incomplete mixing. Under measurement, record calibration, blank response, resolution and range.
Under analysis, record unit conversion, dilution factors and premature rounding. This structure makes it easier to diagnose a scenario than memorising a generic list of mistakes. Rework every viable-count example from the physical actions. Write the transfer ratio beside each tube, multiply the ratios, write the plated volume and reverse both factors only after the count is fixed.
Say CFU per millilitre or gram at the end and explain that the estimate depends on viable organisms forming visible colonies under the chosen conditions. Create deliberate wrong solutions with one missing volume correction or reversed exponent, then explain how the units expose the error. For accuracy and precision, sketch four target diagrams and attach a plausible laboratory cause to each pattern.
Include a tightly clustered biased set, a widely scattered set centred on the reference and a drifting sequence that a mean would conceal. Practise reporting mean ± standard deviation only after naming what was repeated. A set of three instrument readings from one sample is technical repeatability, not three independent batches. For equipment questions, match capacity and resolution to the expected range.
More decimal places do not repair an unsuitable pipette or balance. Write a one-line calibration plan containing the blank, standards covering the sample range and an independent check. End each revision session with an inference audit: what population does the sample represent, which uncertainty is estimated, what systematic errors remain possible and what operational decision can the result support?
If any link is missing, revise the method before calculating again.
