Kappa Agreement Analysis

The Kappa calculator measures agreement between two or more raters on categorical data, supporting both two-rater and multi-rater scenarios, nominal and ordinal categories.


Step 1 — Choose the Kappa variant

Type Description When enabled
Cohen's Kappa Standard kappa for two raters, nominal 2 raters, nominal
Weighted Kappa (Linear) Linear weights for ordinal 2 raters, ordinal
Weighted Kappa (Quadratic) Quadratic weights for ordinal 2 raters, ordinal
Fleiss Kappa Multi-rater kappa 3+ raters

A Data Type selector switches between Nominal and Ordinal.


Step 2 — Provide your data

  • Excel Import — auto-detects two-rater vs multi-rater format
  • Sample Data Generator — built-in datasets
  • Manual Data Input — build custom table

Step 3 — Calculate

Results panel shows:

  • Kappa value (kappa) with 95% CI
  • Observed agreement (p_o)
  • Expected agreement (p_e)
  • Standard error and z-statistic
  • p-value
  • Strength of agreement — Landis & Koch (1977):
Kappa Strength
< 0.00 Poor
0.00 - 0.20 Slight
0.21 - 0.40 Fair
0.41 - 0.60 Moderate
0.61 - 0.80 Substantial
0.81 - 1.00 Almost perfect

For Fleiss Kappa: category-specific kappas also shown.

Export Results button available. A UniversalChatBot is available below.


Statistical methods used

Cohen's Kappa

kappa = (p_o - p_e) / (1 - p_e)

where p_o = sum of diagonal / N, p_e = sum of (row_i * col_i) / N^2.

Weighted Kappa

  • Linear weights: w[i][j] = 1 - |i-j| / (max(R,C) - 1)
  • Quadratic weights: w[i][j] = 1 - (i-j)^2 / (max(R,C) - 1)^2

kappa_w = (weighted_p_o - weighted_p_e) / (1 - weighted_p_e)

Fleiss Kappa

For N subjects, n raters, k categories:

P_bar = mean subject agreement
P_e = sum(p_j^2) where p_j = category proportion
Fleiss kappa = (P_bar - P_e) / (1 - P_e)

Hypothesis: H0: kappa = 0, H1: kappa != 0, alpha = 0.05.