Skip to content
Omnikalc

Lin concordance correlation coefficient calculator

Enter the means, the standard deviations and the Pearson correlation of two measurement methods to get Lin’s CCC, its precision and accuracy parts and the strength of agreement.

Updated on
Your inputs

Value out of range

Value out of range

Value out of range

Value out of range

Value out of range

Results

Lin’s concordance correlation coefficient ρc

0.9404

Accuracy (bias correction factor Cb)

0.9796

Precision (Pearson r)
0.9600
Location shift u
-0.194
Scale shift v = σx/σy
0.938
Strength of agreement (0 poor, 1 moderate, 2 substantial, 3 almost perfect)
1

The Lin concordance correlation coefficient calculator gives Lin’s CCC, its precision and accuracy components, the location and scale shifts and the strength of agreement, from the means, the standard deviations and the correlation of two methods.

How the calculation works

  • Enter the mean and standard deviation of the reference method X.
  • Enter the mean and standard deviation of the new method Y.
  • Enter the Pearson correlation between the paired measurements.

Example

With a correlation of 0.96, standard deviations of 15 and 16 and means of 120 and 123, the concordance coefficient is 0.9404. The precision is 0.9600 and the accuracy 0.9796, with a location shift of -0.194 and a scale shift of 0.938. The strength-of-agreement class is 1.

Why it matters

The calculator helps researchers, clinicians, laboratory scientists and students who compare a new instrument, an assay or a questionnaire with a reference, and who want to know whether the two agree, and not only whether they are correlated, before moving on to a full method comparison with confidence intervals.

Good to know

Use the statistics of the paired measurements, with the same units for both methods, and the sample versions of the standard deviations. The coefficient is sensitive to the range of the subjects, so a narrow range lowers it even when the errors are small. The scale of agreement is only a rule of thumb, and the confidence interval, for which the raw data are needed, matters more than the point value. Check the plot of the data for outliers and for a trend of the difference with the size of the measurement.

The Bland–Altman calculator gives the limits of agreement, the intraclass correlation calculator measures reliability, Cohen’s kappa calculator covers agreement on categories and the correlation significance calculator tests a Pearson coefficient.

Frequently asked questions

What does the concordance correlation coefficient measure?

Lin’s coefficient measures how well pairs of measurements from two methods fall on the 45-degree line of perfect agreement. Unlike the Pearson correlation, which only checks that the points follow any straight line, it also penalizes a shift in the mean and a difference in the spread. It ranges from −1 to 1, and 1 means perfect agreement.

How is it computed?

The coefficient is ρc = 2 r σx σy / (σx² + σy² + (μx − μy)²). It equals the product of the precision, which is the Pearson correlation r, and the accuracy Cb, which is a correction factor between 0 and 1 that falls when the means differ or when the spreads differ. This calculator shows the three numbers.

What values are acceptable?

McBride proposed a descriptive scale for the strength of agreement: poor under 0.90, moderate from 0.90 to 0.95, substantial from 0.95 to 0.99 and almost perfect above 0.99. The right threshold depends on the clinical or technical use of the measurement, and a confidence interval from the raw data is needed to judge it.

How does it differ from the Bland–Altman analysis?

The Bland–Altman plot shows the average difference and the limits of agreement in the units of the measurement, which is easier to interpret clinically. The concordance coefficient gives a single index and is sensitive to the range of the data. The two methods are complementary, and many method comparison studies report both.

Was this calculator helpful?

Thanks for your feedback!

Results are estimates provided for information purposes. They do not replace professional advice.