The recommended curve-fitting model (typically Four-Parameter Logistic) does not adequately fit the standard data points, resulting in poor correlation even with properly prepared standards.
Common Causes
1Assay dynamic range characteristics do not match Four-Parameter Logistic (4PL) model assumptions
2Standard curve exhibits linear behavior in the working range rather than sigmoidal response
3Five-parameter logistic (5PL) model may be needed for asymmetric sigmoid curves
4Residual pipetting errors or replicate variation masking appropriate model selection
Solutions
1Always follow manufacturer's instructions for curve-fitting model in the first instance
2If recommended model fails, systematically test alternative models: linear, 4PL, or 5PL
3Verify that poor fit is not due to replicate variation by examining coefficient of variation (CV%) between duplicates/triplicates
4Confirm pipetting accuracy by re-running standards with meticulous technique before changing curve model
5Consult ELISA kit datasheet for acceptable R² values and curve shape examples specific to the assay