Enrichment · Passion Projects · Lab science
Paste the calibration standards and a set of blank readings. The calculator fits the line, measures the noise, and turns it into the smallest concentration you can detect and the smallest you can report.
The worked example is made up: a blue dye measured by absorbance, six standards and ten blanks. Nothing you type leaves this browser.
Loading the calculator. If nothing appears, turn on JavaScript.
Type the standards’ concentrations and their responses in the same order.
Enrichment · Passion ProjectsFree
Noise, turned into concentration.
The ICH guideline on validating analytical procedures gives LOD = 3.3σ ÷ S and LOQ = 10σ ÷ S, where S is the calibration slope and σ the standard deviation of the response at zero. σ can come from repeated blanks or from how far the standards scatter around the line. At the LOD a signal is unlikely to be noise; at the LOQ the noise is a tenth of the signal, so the number is worth reporting.
Use at least seven blank readings, prepared and measured like real samples. Keep standards close to the expected limit, since a line fitted only to high standards can misjudge the low end. Check the residuals: a pattern means the response curves. Then confirm by measuring a sample spiked near the LOQ.
The worked example is made up. Follow your lab’s safety rules and its own validation procedure. Your data stay in this browser. Support: team@caveatacademic.com.