Reconceptualizing the Mini-Mental State Examination Sentence-Writing Item: From Binary Scoring to a Written-Language Sample

The sentence-writing item of the Mini-Mental State Examination (MMSE) should not remain an unexamined byproduct of binary scoring.This item is conventionally scored as 1 or 0 according to whether a meaningful sentence is produced.Such scoring is central to the simplicity, standardization, and clinical utility of the MMSE [1].However, this approach does not evaluate how the written response is constructed in terms of vocabulary, orthography, or script choice.The item should therefore be reconsidered not merely as a pass/fail task, but as a written-language sample that may reflect cognitive and language functions.The limitation of binary scoring is that qualitatively different written responses may receive the same score.A 1point response may include limited vocabulary, sparse information, grammatical incompleteness, minor orthographic errors, or avoidance of kanji with greater reliance on kana.Conversely, even a 0-point response may contain single words, phrases, verb-only productions, partial lexical information, attempts at sentence construction, or residual writing strategies.Thus, although binary scoring is useful, it does not sufficiently capture differences in sentence or phrase formation, lexical-semantic content, orthographic accuracy, and script choice embedded in written responses.