The tools on this page run in your browser: nothing to install, and nothing you paste or load is uploaded. The practice data come from published human-AI studies, so you can check your analysis against the paper. If your project doesn’t have the data a method needs, start with Practice data.
Practice data
Openly licensed, each with a codebook on the practice-data page. Pick one quantitative set and stay with it across the statistics assignments.
| Set | What it is | Files |
|---|---|---|
| A. Generative AI and story writing quantitative |
Doshi & Hauser (2024). 293 stories written alone, with one AI idea, or with five; about 12 ratings per story on 1–9 scales; writers’ creative ability. | stories · ratings |
| B. Debating GPT-4 quantitative |
Salvi et al. (2025). 750 participants debated a person or GPT-4, personalized or not (2 × 2); agreement before and after. | debates |
| Q1. Creative professionals on AI qualitative |
Anthropic Interviewer (2025). Ten text interviews with creative professionals about AI in their work. | interviews (zip) |
| Q2. Cancer-care staff on AI qualitative |
Bearup (2026). Six interviews with staff at a specialist cancer centre about trust, responsibility, and AI. | interviews (zip) |
Qualitative analysis
| Resource | What it does | From |
|---|---|---|
| κ calculator | Paste two coders’ codes, one row per segment. You get Cohen’s κ, a table of who coded what, and the disagreements to discuss. | Oct 23 |
| Codebook template (CSV) | One row per coded segment: the text, its in-vivo code, the concept label, and who coded it. The first three rows are a worked example; delete them before use. | Oct 23 |
| Open-coding worksheet (Google Doc) | Open-code a transcript as a pod, then compare codes. File → Make a copy. | Oct 23 |
Quantitative analysis
| Resource | What it does | From |
|---|---|---|
| R lab | Every R example from the statistics sessions, ready to run on sets A and B: descriptives, plots, correlation, effect sizes with CIs, factor analysis and alpha, χ², t-tests, ANOVA, and regression (linear, logistic, mixed). Edit the code, or load your own CSV. | Oct 30, Nov 6 |
| Test picker | Answer a few questions about one hypothesis and get the test, the effect size to report, the assumptions to check, what to do if they fail, and the R lab example that runs it. | Nov 6 |
| Factor analysis script (R) | Set A from items to scales: reverse-keying, parallel analysis, rotation, loadings, scale scores, and Cronbach’s alpha, with comments. | Oct 30 |
| Describe before you test (Google Doc) | Worksheet: descriptives, a distribution plot, and one effect size with its CI. | Oct 30 |
| Pick the test, run one (Google Doc) | Worksheet: hypotheses to tests, assumptions and effect sizes, then one test run and written up. | Nov 6 |
Planning a study
| Resource | What it does | From |
|---|---|---|
| Experimental design guide | The seven steps of experimental design, one decision at a time; collects what you write into a draft plan. A one-page overview prints. | Oct 9 |
| G*Power | Free power analysis for Mac and Windows: the sample size you need for a given effect size, α, and power. In R, the pwr package does the same. |
Oct 16 |