Henkan counts too: How to build IME‑conversion‑aware typing tests for Japanese (MS‑IME/Mozc/ATOK)
Most Japanese text entry isn’t just “type and done”—it involves IME conversion and candidate selection. Here’s how to build typing tests that log henkan/muhenkan usage, candidate‑window dwell, and confirmed‑conversion rate so you can report a fair, IME‑aware WPM for Japanese typists.
Read more →