Chandler, C., Diaz–Asper, C., Turner, R.S., Reynolds, B., & Elvevåg, B. (2023). An explainable machine learning model of cognitive decline derived from speech. Alzheimer's & Dementia: Diagnosis, Assessment & Disease Monitoring. 15. https://doi.org/10.1002/dad2.12516.
During the preparation of this manuscript, two errors were introduced:
ERROR 1:
In Section 2.2, Paragraph 1, there are multiple typographical errors in the formulas.
Specifically, there should be a set of parentheses around the formula for content density, and a space between “word” and “tokens”:
“(count(verbs + nouns + adjectives + adverbs)/count(word tokens))”.
Next, there should be a semicolon after the formula for type token ratio, and a space between “word” and “tokens”:
“type token ratio (count(word types)/count(word tokens));”.
There should be no open parentheses here: “(moving average type token ratio”, it should be “moving average type token ratio”. The italics on the font for “(averaged moving windows of” should be removed to match the style of the rest of the formulas. Finally, “word” and “tokens” should be separated with a space in that formula.
The corrected sentence is shown here:
Lexeme‐level features were first extracted from free speech responses. Token and type counts measured poverty of speech. Part of speech counts and frequencies; type token ratio (); moving average type token ratio (averaged moving windows of
); Brunét's index 
; and content density () measured poverty of content.
ERROR 2:
In Section 2.2, Paragraph 1, the phrase “size ∈ 2, 8”, where 2, 8 are written as citations, should be “size ∈ [2,8]” where this is a range of numbers (not citations).
We apologize for these errors.