More independent readings are added under stable conditions using a sensor with a fixed offset. What can change without a calibration correction?

• Both the mean's statistical uncertainty and the fixed offset must decrease.
• The mean's statistical uncertainty can fall; the fixed offset remains.
• The mean's statistical uncertainty stays fixed; the offset averages away.

A scale reads 55 g three times for a 50 g reference mass. Which description separates precision from accuracy?

• Precise repeats, but inaccurate relative to the reference
• Accurate relative to the reference, but imprecise across repeats
• Accurate relative to the reference and precise across repeats

An estimate is 14.276 s with uncertainty 0.36 s. Round uncertainty to one significant digit, then the estimate to the same decimal place. Which report follows this convention?

• 14.276 ± 0.4 s
• 14.3 ± 0.4 s
• 14.28 ± 0.36 s

A method records instruments, settings, calibration and analysis code. It says 20 specimens were tested and two excluded, but gives no exclusion rule. What should be added?

• State the exclusion rule and when it was chosen.
• List retained specimens and leave the excluded measurements unexplained.
• Repeat the final estimate without describing specimen selection.

Two sensors' mean readings are plotted without axis units or marker meanings. Error bars are shown but not defined. Which revision makes the figure interpretable?

• Add decimal places and let readers infer the units and sensor labels.
• Add a larger title and remove the bars without explaining uncertainty.
• Add units, a sensor legend, and an error-bar definition.

An observational figure shows exposure and response rising together in a sample. No design or analysis isolates a causal effect. Which caption stays within this evidence?

• Exposure and response were associated in the observed sample.
• Increasing exposure caused the response change throughout the population.
• The association rules out every other cause of response variation.

Lee counts exactly 12 intact blocks and measures their total mass as 96.437 g. He reports no uncertainty for either. Which statement distinguishes the two results?

• Give uncertainty only for the count; the scale's displayed digits make the mass exact.
• Omit both uncertainties because an exact block count makes their measured mass exact.
• Give uncertainty for the measured mass; the exact count needs none.

Another group follows a documented method and obtains a different estimate. What should be checked before deciding that the results conflict?

• Treat the difference alone as proof that the original procedure cannot be repeated.
• Compare uncertainties, conditions, calibration and analysis.
• Discard the new estimate because repetition must return an identical number.