Drawing on responses from 63 faculty, the thorough 240 page study examines how generative AI is changing work with spreadsheets, statistical packages, and data-analysis programs, including Excel, Google Sheets, R, Python, SPSS, and other software.
The report provides benchmarks for academic leaders, research support teams, software providers, and faculty planning AI training and investment. It covers applications, integrations, analytical tasks, time savings, work quality, prompt libraries, autonomous agents, and plans for future use.
Selected findings
Experimentation exceeds intensive adoption. During the past year, 58.73% used generative AI with analytical software regularly, occasionally, or experimentally. Only 12.70% currently use it daily or several times per week.
ChatGPT leads the application field, used for spreadsheet or statistical work by 47.62% of the entire sample. Claude follows at 23.81%, Microsoft Copilot at 22.22%, and Google Gemini at 17.46%. Respondents could select multiple systems.
Manual workflows predominate: 47.62% copy and paste between AI and analytical software, and 31.75% upload data or spreadsheet files. Only 15.87% use a built-in AI assistant; 4.76% connect through an API or custom script.
Writing formulas or code is the leading reported task, at 38.10%, followed by creating tables, charts, or visualizations at 31.75% and cleaning or restructuring data at 30.16%. Interpreting results is less common, at 15.87%.
Across the full sample, 26.98% report saving at least 25% of the time spent on tasks for which they use AI. Overall, 46.03% say AI has improved the quality of their spreadsheet or statistical work, while 9.52% report some reduction in quality.
Written responses describe faster coding, help with unfamiliar software commands, and automation of recurring analyses. Faculty also report errors, verification costs, privacy concerns, and difficulty connecting AI to preferred packages. One-third are likely or very likely to increase AI use during the next year.