docs: fix typo

README.md

@@ -75,7 +75,7 @@ Recommended practices
 - Should filter input tables on partition and clustering columns
 - Should use `_` prefix in generated column names not meant for output
 - Should use `_bits` suffix for any integer column that represents a bit pattern
-- Should not use `DATETIME` type, due to incompatiblity with
+- Should not use `DATETIME` type, due to incompatibility with
   [spark-bigquery-connector]
 - Should read from `*_stable` tables instead of including custom deduplication
   - Should use the earliest row for each `document_id` by `submission_timestamp`