Methodology Brief

Kitchen-scale calibration for weighed-food reference protocols: a checklist

A methodology brief

Sofia Rivera, MS, RD; Meera Patel, PhD
Published January 27, 2025

Background

In validation work, the weighed-food record (WFR) is frequently treated as the reference against which candidate methods are compared. Its apparent simplicity - a scale, a plate, a number - can mask a number of error sources: instrument drift, off-centre loading, ambient temperature effects on strain-gauge sensors, and inconsistent tare handling. A 2021 review (Salinas et al., Br J Nutr) reported that only 28% of WFR-based validation studies documented any scale calibration procedure, and fewer than 10% reported a verification step during data collection.

The Initiative’s convention is that any weighed-food reference used in an Initiative-branded validation must follow the checklist below, and the calibration log must be deposited with the dataset.

The Method

The procedure has three components: pre-study calibration, in-study verification, and post-study verification.

Pre-study calibration. Each scale is calibrated against a set of class M1 or better reference masses spanning the expected range (typically 10 g, 100 g, 500 g, 1000 g, 2000 g). A scale is accepted for study use if the absolute error at each reference point is within the manufacturer’s stated tolerance and within 0.5% of the nominal mass, whichever is tighter. Scales that fail at any reference point are retired from the study pool.

In-study verification. On every data-collection day, before the first measurement, a single-point verification with a 100 g and a 500 g reference mass is performed and logged. If either point deviates by more than 1.0 g (100 g point) or 2.0 g (500 g point), the scale is re-calibrated before use. A verification is also logged every four hours during a data-collection session, because drift with ambient temperature is most pronounced in the first 30 minutes of use.

Post-study verification. At study close, the full five-point calibration is repeated. Any scale whose post-study error exceeds the pre-study acceptance bounds is flagged, and all data collected after the last successful verification is subject to a sensitivity analysis.

Tare handling. Plates and containers are tared at ambient temperature. For hot foods, the tare plate is pre-warmed when practical, and the tare is recorded numerically (not cleared) so that any drift during the measurement can be reconstructed.

Worked example

A scale calibration log entry for one study day might look like this.

The close-of-day failure at the 2000 g point triggers a sensitivity analysis for all day-three measurements above 1500 g (three eating occasions in this example).

Common pitfalls

• Relying on the scale’s self-calibration mode. Auto-calibration against an internal standard does not detect span error across the full measurement range.
• Using a single reference mass. A scale can be accurate at 100 g and biased by 1% at 1000 g.
• Tareing hot pans or plates at ambient temperature without logging the thermal condition. Thermal expansion of the platter can shift the tare by 0.2 to 1.0 g.
• Placing the scale on a surface with vibration (near a refrigerator compressor, on a pedestrian walkway). Strain-gauge scales oscillate visibly and settle to biased values.
• Recording final weights rather than the full tare-gross-net triad. Without the triad, errors cannot be traced.

Recommended reporting

• Scale make and model (generic class description is acceptable), resolution, and stated manufacturer tolerance.
• Pre- and post-study five-point calibration results.
• In-study verification schedule and all logged readings.
• Number of measurements collected after the last successful verification, and sensitivity-analysis results for these.
• Tare handling procedure for hot foods.
• Ambient temperature range during data collection.

References

1. Salinas E, Okafor N. Calibration practices in weighed-food validation studies: a systematic audit. Br J Nutr. 2021;126(10):1478-1487.
2. Patel R, Rivera M. A minimal calibration protocol for kitchen scales in nutrition research. Public Health Nutr. 2023;26(7):1411-1419.
3. Kessler F, Donnan C. Thermal drift in strain-gauge scales under kitchen conditions. Nutrients. 2022;14(9):1862.
4. Rivera M. The weighed-food record as reference: assumptions and limits. Am J Clin Nutr. 2024;119(4):810-817.
5. Holm T, Alvarez P. Class M reference masses and their role in nutrition laboratory QC. J Nutr. 2020;150(5):1291-1296.
6. Patel R. Reconstructing tare errors from triad records: a small simulation. Nutrients. 2024;16(11):1723.
7. Kristoffersen O. Documenting instrument provenance in dietary reference datasets. Public Health Nutr. 2022;25(3):540-548.