Precision and accuracy

BactoBox® is a precise instrument. Repeat measurements of the same sample land close together.

Accuracy is a separate question. A cell count needs a target to be accurate against, and ours is the plate count for actively growing cultures.

All figures on this page come from Jordal et al. 2025, a peer-reviewed performance qualification of BactoBox® against plate counts. The study covers six species that differ in size, shape, and cell envelope type: Staphylococcus epidermidis, Escherichia coli, Klebsiella aerogenes, Acinetobacter baumannii, Listeria innocua, and Pseudomonas fluorescens.

These figures only apply to BacTotal_v2024-10

BacTotal_v2024-10 is the default gating configuration. Any other configuration counts a different set of objects, so the precision and accuracy figures on this page no longer hold.

The difference between precision and accuracy

Precision is how close repeat measurements are to each other. It says nothing about whether they are right.

Accuracy is how close the average of those measurements is to the target. It says nothing about how much they scatter.

A dartboard makes this easy to see.

Precise, not accurate. The darts group tightly, so the throw repeats well. The average sits away from the bullseye, so the throw is off target.

Accurate, not precise. The darts scatter widely, so the throw repeats poorly. The average sits on the bullseye, so the throw is on target.

Precision comes from the instrument. Accuracy comes from the calibration.

Precision

The average coefficient of variation (CV) of the cell count is 6.6 %, measured across the six species. Per species, the CV ranges from 3.2 % to 9.6 %.

Plate counts of the same samples have an average CV of 15.3 %.

Each CV covers triplicate measurements of the same culture, so it includes your sample preparation and dilution as well as the instrument itself. The CV covers the full growth curve, including lag, acceleration, and decline phase.

These figures depend on careful sample preparation

We prepare and dilute every sample behind these figures with great care. Work with the same care to reach the same repeatability for BactoBox® as well as the plate counts.

Accuracy

BactoBox® ships with a linear calibration factor on the cell count. The calibration target is the plate count of a culture that contains culturable cells only. These are cultures in exponential phase and early stationary phase, where essentially every cell still forms a colony.

With that calibration in place, 37 out of 42 paired sampling points fall within 0.1 log₁₀ units of the plate count. That is 88 %.

A band of 0.1 log₁₀ units means the cell count sits within 26 % of the plate count.