Calibrating Skinfold Calipers
How to test the spring tension and jaw accuracy of your Harpenden or Slimguide calipers.
The Necessity of Calibration
If you use a skinfold caliper to track your body fat, the entire process relies on the mechanical accuracy of the device. If the spring weakens or the jaw alignment shifts, your data is compromised.
Professional Harpenden calipers (£200+) come with calibration blocks. Cheap plastic calipers do not, which is why they are effectively useless after a few months of heavy use.
The 10g/mm² Standard
For the Jackson-Pollock and Durnin-Womersley equations to be valid, the caliper jaw must exert a constant pressure of 10 grams per square millimeter (10g/mm²) across the entire range of jaw openings.
If the pressure is too high, it compresses the fat too much, resulting in an artificially low reading. If the pressure is too low, it fails to compress the fluid in the fat tissue, resulting in an artificially high reading.
How to Test Jaw Alignment
Before every testing session, visually inspect the jaws.
- Close the caliper completely.
- Hold it up to a light source.
- If you can see light passing between the closed jaws, the pivot point is warped or the metal is bent. The caliper must be repaired or discarded.
How to Test Measurement Accuracy
You must verify that the dial/scale is reading the correct millimeter distance.
Using a Calibration Block: If you own a Harpenden, use the provided steel calibration block (which typically has steps cut precisely at 10mm, 20mm, 30mm, and 40mm).
- Open the jaw and close it onto the 20mm step.
- The dial should read exactly 20.0mm.
- If it does not, use the adjustment screw on the dial face to zero it correctly.
The DIY Method: If you own a Slimguide and do not have a calibration block, use a precision object with a known, fixed width. A high-quality steel drill bit or a stack of standard coins (measured with a digital micrometer first) can serve as a makeshift calibration block.
If your caliper fails the alignment or accuracy tests, throw it away. Bad data is worse than no data.