WEBVTT

00:00.000 --> 00:04.748
This simulated method walkthrough uses
synthetic writing and answers, not

00:04.748 --> 00:06.296
participant data or findings.

00:06.296 --> 00:12.000
Actual-app footage is recorded locally;
title and procedure cards provide context.

00:12.000 --> 00:16.926
The procedure card outlines an unaided
writing sample, followed by three revision

00:16.926 --> 00:20.342
blocks, each with a different assistance
mode and document.

00:20.342 --> 00:25.210
A questionnaire follows every block;
reflection and closing questions come

00:25.210 --> 00:26.317
afterwards.

00:26.317 --> 00:31.000
We show one example counterbalanced order:
Guide, then Suggest, then Rewrite.

00:31.000 --> 00:36.127
The setup sequence covers sound,
orientation, consent, recording, and

00:36.127 --> 00:37.309
screen sharing.

00:37.309 --> 00:42.163
We advance through these checks before
opening the writing sample.

00:42.163 --> 00:45.860
Identifying details are masked for
anonymous review.

00:48.000 --> 00:51.309
The unaided writing sample lasts eight
minutes.

00:51.309 --> 00:54.189
We enter text without AI assistance.

00:54.189 --> 00:59.683
Waiting periods are omitted in this
walkthrough, and labels identify any

00:59.683 --> 01:01.860
demonstration timer changes.

01:05.000 --> 01:09.296
Before revision, the app presents
thinking-aloud instructions.

01:09.296 --> 01:12.910
Writers are asked to explain their
thinking while revising.

01:12.910 --> 01:17.000
We move from these instructions into the
first block.

01:17.000 --> 01:20.324
Each revision block allows thirteen
minutes.

01:20.324 --> 01:25.361
Assistance uses prepared responses,
without a participant prompt or live model

01:25.361 --> 01:26.511
call.

01:26.511 --> 01:32.088
Guide identifies what it thinks is wrong
and explains why it matters; it supplies

01:32.088 --> 01:33.633
no replacement text.

01:33.633 --> 01:38.876
We accept the guidance, recording
acknowledgment without changing the editor

01:38.876 --> 01:40.177
text.

01:40.177 --> 01:45.180
We then type a revision ourselves: the
writer constructs the new wording.

01:57.000 --> 02:01.964
After each block, eight seven-point
ratings cover perceived quality, control,

02:01.964 --> 02:04.060
ownership, effort, and attention.

02:04.060 --> 02:09.000
A ninth, optional question asks for the
reasoning behind one change.

02:09.000 --> 02:14.239
Suggest proposes individual edits, each
with a reason, and shows what would

02:14.239 --> 02:15.495
change.

02:15.495 --> 02:20.734
Here, the proposal replaces a local
passage rather than the whole document.

02:20.734 --> 02:25.620
We accept the proposal and watch that
passage change in the editor.

02:25.620 --> 02:28.035
The local replacement is reversible.

02:28.035 --> 02:32.445
We then add further text ourselves;
accepting the proposal does not end

02:32.445 --> 02:34.180
editing.

02:49.000 --> 02:52.074
The same nine-item questionnaire follows
Suggest.

02:52.074 --> 02:56.683
We select ratings on the scale from
strongly disagree to strongly agree, then

02:56.683 --> 02:59.000
move to the optional written question.

02:59.000 --> 03:04.383
Rewrite supplies a complete revised
document with one overall explanation.

03:04.383 --> 03:09.954
The choice concerns the whole rewrite,
rather than individual proposed edits.

03:09.954 --> 03:13.748
We accept it, replacing the full text in
the editor.

03:13.748 --> 03:15.955
The replacement remains editable.

03:15.955 --> 03:18.354
We make a further manual edit.

03:18.354 --> 03:24.304
This shows the distinction between
accepting the complete version and making

03:24.304 --> 03:25.980
further revisions ourselves.

03:39.000 --> 03:43.143
The third block has its own questionnaire,
answered about Rewrite.

03:43.143 --> 03:47.794
Only after this stage does the session
move to reflection across the three

03:47.794 --> 03:49.000
blocks.

03:49.000 --> 03:54.267
The reflection asks for at least three
editorial decisions and their reasoning.

03:54.267 --> 03:59.439
Three further questions follow here, asked
once across the documents: what was not

03:59.439 --> 04:04.000
used, what became easier or harder, and
what was left unchanged.

04:04.000 --> 04:08.867
Eighteen closing items cover which mode to
keep, recommend, and judge best; trust and

04:08.867 --> 04:13.415
reliance; voice; and two background
questions, followed by nine written

04:13.415 --> 04:15.054
questions.

04:15.054 --> 04:19.514
The writing-background answer and these
nine responses are optional: their boxes

04:19.514 --> 04:21.000
can be left blank.

04:21.000 --> 04:24.434
Submitting the closing questionnaire
completes the session.

04:24.434 --> 04:29.395
Study information remains available to
read optionally; no further questions or

04:29.395 --> 04:31.000
required steps follow.