Diagnostics › Method comparison

Comparison of multiple methods (≥ 3)

Multi-method agreement compares three or more continuous methods at once, reporting a two-way ICC alongside every pairwise Pearson r and Lin's CCC.

What is Comparison of multiple methods (≥ 3)?

When you have three or more methods on the same subjects (e.g. central lab, two point-of-care devices, manual reading), you want both a single global-agreement number (ICC) and pairwise insight into which two methods agree best (Pearson r + CCC for each pair).

ICC(2, 1) = absolute-agreement form, ICC(2, k) = average-of-k form. Use ICC(2, 1) when you'd be reporting a single method as the result; ICC(2, k) when you'd be averaging all k.

When should I use Comparison of multiple methods (≥ 3)?

≥ 3 methods on the same subjects.
Multi-reader / multi-instrument validation studies.

What data does it need?

≥ 3 numeric columns, each one method.

What does it report?

ICC (consistency + absolute agreement) + per-method means / SDs + pairwise Pearson r / CCC / mean diff / SD diff.

How do I interpret the result?

Use the pairwise table to spot a single method that disagrees with the others — low CCC against multiple peers points to a faulty / mis-calibrated instrument.