Analysis · Technology
What an AI Adoption Score Actually Measures
By AWEI · AI-compiled · Published · Analysis prepared · 1 source · doi.org
A healthcare worker willing to recommend AI can help raise an adoption score without any recorded increase in use. That distinction is built into a 2022 study by Mengting Cheng, Xianmiao Li and Jicheng Xu: its adoption measure combined willingness to learn and use AI, intention to use it in future, and willingness to recommend it to others.
The researchers analyzed 343 questionnaires from dental doctors, nurses and medical technicians in secondary and tertiary hospitals in Anhui Province, China. Eligibility included using AI-assisted diagnosis and treatment in their department. This was therefore a survey of attitudes among workers with relevant experience, rather than a test of whether previously unexposed staff would start using a system.
Participants rated expected benefits, ease of use, social influence, trust and adoption intention on five-point scales. Trust correlated with adoption intention at 0.604, compared with 0.441 for expected benefits and 0.261 for ease of use. These were relationships between questionnaire scores; the study did not measure improvements in diagnostic accuracy or working time.
The authors also tested a proposed chain linking expectations to social influence, then trust, then intention. Their mediation analysis used 5,000 bootstrap resamples to estimate uncertainty and reported support for those indirect relationships. But measuring all the variables in one survey cannot establish that this sequence happened over time or caused subsequent behavior.
The questionnaire wording makes the finding especially useful to interpret. Social influence included other people’s use and approval, while adoption intention included recommending AI to others. The result therefore describes a social dimension of acceptance alongside personal willingness. It does not isolate routine use, and its dental setting in one province limits generalization.
For deployment teams, the practical lesson is to inspect the questions behind an adoption score. Learning, recommending and intending to use a tool are distinct signals. This study offers a way to examine acceptance; establishing operational value requires separate evidence of actual use and work outcomes.
Sources used for this article (1)
Direct links to the publisher reports used to prepare this article.
- Source 1
- Promoting Healthcare Workers’ Adoption Intention of Artificial-Intelligence-Assisted Diagnosis and Treatment: The Chain Mediation of Social Influence and Human–Computer Trust Mengting Cheng, Xianmiao Li, Jicheng Xu
