New items, pass 1 vs pass 2: n=677  four-category kappa=0.817 (agreement 88.8%)  related (Y/E) vs unrelated kappa=0.833 (agreement 92.8%)  Y vs not-Y kappa=0.740 (agreement 97.0%)
  of which labelled pairs: n=177  four-category kappa=0.681 (agreement 81.4%)  related (Y/E) vs unrelated kappa=0.707 (agreement 90.4%)  Y vs not-Y kappa=0.693 (agreement 89.8%)
  of which ordinary pairs: n=500  four-category kappa=0.824 (agreement 91.4%)  related (Y/E) vs unrelated kappa=0.749 (agreement 93.6%)  Y vs not-Y kappa=0.748 (agreement 99.6%)
35 anchor items, pass 1 vs first sample: n=35  four-category kappa=0.875 (agreement 91.4%)  related (Y/E) vs unrelated kappa=0.940 (agreement 97.1%)  Y vs not-Y kappa=0.800 (agreement 94.3%)
35 anchor items, pass 2 vs first sample: n=35  four-category kappa=0.918 (agreement 94.3%)  related (Y/E) vs unrelated kappa=0.881 (agreement 94.3%)  Y vs not-Y kappa=1.000 (agreement 100.0%)
35 anchor items, pass 1 vs pass 2: n=35  four-category kappa=0.835 (agreement 88.6%)  related (Y/E) vs unrelated kappa=0.940 (agreement 97.1%)  Y vs not-Y kappa=0.800 (agreement 94.3%)

Cross-tabulation of the two passes on the new items (rows = pass 1, columns = pass 2):
p2_code    E    N   X   Y
p1_code                  
E        143   22   0   7
N         23  341   7   0
X          1    3  86   0
Y         10    3   0  31

Anchor disagreement (pass 1) 壴/鼓: first sample E, pass 1 Y
Anchor disagreement (pass 1) 反/返: first sample E, pass 1 Y
Anchor disagreement (pass 1) 卑/稗: first sample E, pass 1 N
Anchor disagreement (pass 2) 弋/必: first sample N, pass 2 E
Anchor disagreement (pass 2) 卑/稗: first sample E, pass 2 X

Addendum (not used in any analysis here, see the plan, section 7): second answers returned by the same coder instance, compared with the first answer that counts
  Pass 1, batch 3: n=102, four-category agreement 76/102, kappa=0.554
  Pass 1, batch 6: n=101, four-category agreement 82/101, kappa=0.710
