M22-A3 distinguishes between simple linear regression (y = mx + b) and more complex weighted regression (e.g., 1/x or 1/x²). It warns that heteroscedasticity (increasing variance at high concentrations) is common in LC. The guideline provides statistical methods (e.g., lack-of-fit F-test) to verify the calibration model.
Even for exempt media, users must still check for obvious defects like: Cracked plates or agar detachment. Unequal filling or frozen/melted agar. Hemolysis in blood-containing media. Archived but Valid: clsi m22a3 pdf
Excessive moisture, dehydration, or frozen/melted agar. Contamination: Any visible growth or excessive bubbles. Purchasing and Access M22-A3 distinguishes between simple linear regression (y =