Research project record
Deep Reasoning in China’s banking industry
Author-maintained record of a reported survey of 4,897 Chinese credit officers, including reported associations with usefulness, trust and perceived risk.
- Status
- completed
- Role
- Researcher
Research question
The study examined what influences credit officers’ intention to adopt DeepSeek R1 reasoning for credit-risk assessment in China’s banking industry.
Author’s contribution
Dr. Kushan Liyana Arachchige conducted the study as the researcher. The work connects technology-acceptance research with the governance demands of consequential financial decisions.
Evidence status
The public profile linked on this page records the study title, sample, method and headline findings. A repository copy, dataset and full measurement instrument have not been linked here; that absence is stated explicitly rather than inferred away.
How
Methods and evidence
Methodology
The author reports quantitative survey research using an extended Technology Acceptance Model and structural equation modelling to examine adoption of DeepSeek R1 reasoning in credit-risk work.
Sample
The author reports 4,897 credit officers working in China’s banking industry.
Data sources
- The author reports survey responses from 4,897 Chinese credit officers.
What the record supports
Key findings
- The author reports that perceived usefulness, ease of use and trust were associated with stronger adoption intention.
- The author reports perceived risk as a material barrier to adoption in high-stakes credit decisions.
Trust
Limitations and disclosures
Limitations
This public record summarises the research design and reported findings. The full dissertation, instrument, model diagnostics and dataset are not published on this site, so readers should not treat the summary as a substitute for the complete academic work.
Funding
No funding information is stated in the public record.
Disclosures
This page documents research undertaken by the site author. No competing interests are stated in the public record.