GIFSQ Tools · Module 3 — Food Microbiology & Enumeration · Tool 045 of 500
Two plates from the same dilution. Enter both counts to get the mean result and check whether the plates agree within normal counting variation.
Try sample values — fills the form with the worked example below.
–
Mean count
–
CFU/g
–
Difference between plates (%)
Equation: mean count = (C₁ + C₂) ÷ 2 · CFU/g = mean ÷ (V × d) · difference = |C₁ − C₂| · difference % = difference ÷ mean × 100 · agreement limit = 1.96 × √(C₁ + C₂).
Method: Duplicate-plate comparison against Poisson counting variation, consistent with ISO 7218 and FDA BAM plate-count practice. The 1.96 factor is the 95% normal-approximation limit for the difference of two Poisson counts.
Assumptions: the two plates are true duplicates from the same dilution; colonies are distributed at random (Poisson).
Limitations: the check covers counting variation only — it does not catch a shared dilution or pipetting error; a large difference can also come from uneven mixing.
Rounding: mean count to 1 decimal place; CFU/g as a whole number; difference % and the agreement limit to 2 decimal places; intermediates unrounded.
This is a calculation aid, not a laboratory report — confirm against your accredited method before citing a value.
Method: Duplicate-plate comparison against Poisson counting variation, consistent with ISO 7218 and FDA BAM plate-count practice. The 1.96 factor is the 95% normal-approximation limit for the difference of two Poisson counts.
Assumptions: the two plates are true duplicates from the same dilution; colonies are distributed at random (Poisson).
Limitations: the check covers counting variation only — it does not catch a shared dilution or pipetting error; a large difference can also come from uneven mixing.
Rounding: mean count to 1 decimal place; CFU/g as a whole number; difference % and the agreement limit to 2 decimal places; intermediates unrounded.
This is a calculation aid, not a laboratory report — confirm against your accredited method before citing a value.
Worked example: Duplicate plates at 10⁻³ with 1 mL plated give 142 and 158 colonies. Mean = 150; CFU/g = 150 ÷ (1 × 0.001) = 150,000 CFU/g; difference = 16 (10.67%); agreement limit = 1.96 × √300 = 33.95. 16 is within the limit, so the plates agree.
Enter values above to see results.
Duplicate plates from the same dilution should agree within Poisson counting variation. A gap larger than the agreement limit points to uneven mixing or a pipetting slip — repeat the plating. Tool 045 of 500 — GIFSQ
