Before-and-after survey / preprint · Source language: English
Revisiting expectations before adoption against real work
Staff at a US transportation department who tried Copilot for eight weeks revised their expectations through real work and wanted role-specific training and places to consult.
- Who:
- Administrative, technical, planning and management staff at the North Carolina Department of Transportation, US
- Scope:
- About 175 participants / 133 responded at both points / 124 analyzed / United States
- Source published:
- 2026-07-15
- Period observed:
- Eight weeks in 2025 (month unconfirmed)
- Editorial check:
- 2026-10-23
Only the source ID is saved in this browser. Nothing syncs across devices.
What was confirmed
- Perceived usefulness on a five-point scale averaged 3.85 before introduction and 3.62 after. The authors interpret this as a revision of expectations.
- Responses included requests for role-specific training and a common point of contact for consulting about problems.
Who, and in what situation
Introduced Copilot with training and tried it for drafting documents, summarizing email, meeting notes and similar tasks.
What they needed
Finding uses that work for each role's tasks, and support when in trouble.
What they tried
Matched the same staff's responses before and after introduction to examine perceptions after training and use at work.
What this source lets you read
It seems good to record separately the expectations before adoption and the uses and needed support that remain after trying.
What to check next
In your own role, how do the uses you expected before trying differ from those you actually want to keep using?
Limits of this source
- A before-and-after survey self-reported at a single organization, with no comparison group. It is not a result measuring productivity or accuracy.
- This is a preprint, and we have not confirmed that it is peer-reviewed. Long-term retention is also unconfirmed.
Read the primary source
Links use the original title or a HAKOMIRO source description. Japanese originals retain their published wording.
Persona Migration and Expectation Recalibration in Generative AI Adoption (opens in new tab) ↗
O. Shoghli et al. / arXiv /
The authors are affiliated with UNC Charlotte and NCDOT. After matching and quality checks, 124 were analyzed.