How this calculation works
An absorbance standard curve converts a measured response into an estimated concentration using prepared standards. Its fitted intercept can capture a consistent background offset, so forcing the curve through zero is not automatically appropriate. This page fits absorbance against known concentration and then inverts that line for one unknown. It also restores a separately entered dilution multiplier to report an original-sample estimate.
Fit A = intercept+slope C by unweighted OLS; original C = (A_unknown−intercept)/slope ×dilutionA worked example
Standards at concentrations zero, one and two with absorbances 0.1, 0.3 and 0.5 define an exact slope of 0.2 and intercept 0.1. An unknown absorbance of 0.4 therefore corresponds to measured concentration 1.5. With an original-sample dilution multiplier of two, the original estimate is three in the standard concentration unit. Changing dilution does not alter the fitted line, residuals or measured aliquot estimate.
Good to know
Nonlinear response, saturation, matrix effects and mismatched blanks can invalidate the straight-line assumption. No instrument file or laboratory data is uploaded; the entered observations are processed locally in the browser. The page does not estimate uncertainty, a detection limit or analytical validity. An extrapolated or negative result should prompt review of the actual calibration and sample procedure rather than an automatic analytical conclusion.
Reference: NIST — calibration and linear least squares (opens in a new tab) ↗
What this calculation answers
An absorbance standard curve converts a measured response into an estimated concentration using prepared standards. Its fitted intercept can capture a consistent background offset, so forcing the curve through zero is not automatically appropriate.
This page fits absorbance against known concentration and then inverts that line for one unknown. It also restores a separately entered dilution multiplier to report an original-sample estimate.
The Simple linear regression calculator explains a direct least-squares line. This calibration adds response inversion, a dilution multiplier and an explicit unknown-range flag.
Use the right input conventions
Provide at least three standards with varying concentration and absorbance. Use one concentration unit for all standards and retain the same measurement protocol, path length and blank convention for the unknown.
The fit is unweighted ordinary least squares with an estimated intercept and a required positive slope. It is not an errors-in-both-axes calibration and does not give greater weight to selected standards or replicates.
Interpret the result
The table shows fitted responses and residuals for the standards, while R squared describes response variation explained by the fitted line. A high R squared does not establish assay specificity, low-end accuracy or an acceptable detection limit.
The unknown is flagged when its absorbance falls outside the entered standard-response range. Negative inverse estimates are retained as signed mathematical results rather than clipped to zero and presented as physical concentrations.
Work through the example
Standards at concentrations zero, one and two with absorbances 0.1, 0.3 and 0.5 define an exact slope of 0.2 and intercept 0.1. An unknown absorbance of 0.4 therefore corresponds to measured concentration 1.5.
With an original-sample dilution multiplier of two, the original estimate is three in the standard concentration unit. Changing dilution does not alter the fitted line, residuals or measured aliquot estimate.
Where the model stops
Nonlinear response, saturation, matrix effects and mismatched blanks can invalidate the straight-line assumption. No instrument file or laboratory data is uploaded; the entered observations are processed locally in the browser.
The page does not estimate uncertainty, a detection limit or analytical validity. An extrapolated or negative result should prompt review of the actual calibration and sample procedure rather than an automatic analytical conclusion.
A few questions, answered
Does R squared equal one prove the unknown is valid?
No. Exact standard alignment does not establish sample compatibility, specificity or accuracy beyond the calibrated range.
What changes in the comparison?
Move the response above the standard range and inspect the explicit extrapolation flag.