Self-test 10.2 – Estimation of measurement uncertainty in chemical analysis

Here you can test your understanding how the uncertainty component accounting for the random effects can be estimated in practice.

A laboratory wants to estimate the within-lab reproducibility component (component accounting for the random effects) of measurement uncertainty of protein content determination in different foods. For this protein content in a control sample of smoked sausage was determined repeatedly during a time period of almost one year. Every day when this smoked sausage was analysed only one parallel measurement was made. The following results were obtained:

Please calculate the within-lab reproducibility component of uncertainty (please give 2 or 3 decimals). Absolute uncertainties can be used in this case.

Clue 1:

The reproducibility component is found as the standard deviation of the results obtained on different days.

Which of the following is true?

It does not matter, whether the weeks on which the determination of u(R_w) was done are spread out over the entire year or are made on ten consecutive weeks, e.g. on wwke 3-12.

It would be better to do more measurements than 10 for u(R_w) determination.

The true protein content in the control sample is with 95% probability within ±2·u(R_w) from the mean value of the ten determinations.

In order for this u(R_w) estimate to be valid it is very important that the control sample is stable in time.

In order for this u(R_w) estimate to be valid it is very important that the control sample is homogenous.

Which of the following issues are important in determining s_RW?

Sufficient number of replicate measurements

Sufficiently long timeframe

Homogeneous sample

Matrix match

Concentration range match

Reliable reference value

Including the sample preparation