AI Workshop · Duke University DC
This site is a collection of skills you can drop into your AI to level up your experimental research — not just write more papers. We’re focused on field experiments, because that’s where this library was built.
Audit study designs against qualitative pre-experiment interviews to identify missing constructs, causal tensions, and measurement gaps before launch. Use this skill when preparing to field an experiment or observational study to ensure pre-analysis plans match fieldwork.
Map a research claim against the existing literature through a staged review process that audits terminology, scores papers on relevance and liveness, assesses home-faculty overlap, and identifies framing strategies for paper introductions.
Critique experimental draft strategy papers and abstracts against target journal standards using literature scans and multi-persona reviewer rubrics. Use this skill before running an experiment or submitting a manuscript to identify design flaws, literature gaps, and publication potential.
Audit an experimental design to check whether a proposed treatment uniquely maps to its target theoretical construct. Use before data collection to identify confounding constructs, construct invalidity, and required design fixes or control arms.
Scan and inventory a research project folder to reconstruct file version lineages, diff code, extract document contents, and generate a chronological status report. Use it to quickly audit or get an overview of a project folder's full history.
Generate, critique, and refine field experiment concepts into publishable empirical paper plans across various social-science domains. Activate this skill when brainstorming RCT topics, repairing weak experimental designs, or developing structured research plans for empirical research teams.
Evaluate an experimental design prior to data collection to identify fatal methodological flaws, power constraints, or unaddressed threats to validity. Use this skill when planning a field or lab experiment to assess its publishability before committing resources.
Screens and narrows a list of candidate hypotheses down to three distinct, testable rivals using six joint criteria. Use this skill before data collection to select mutually exclusive mechanisms and prevent ad hoc or redundant shortlist selection.
Refine rough social-science and management research ideas into structured, publication-ready empirical study proposals through a staged mentor-style gate process. Use this skill when developing early-stage study designs, dissertation proposals, or stress-testing research ideas before field implementation.
Interviews empirical social-science researchers to build, critique, and refine a Strategy Experiment Canvas and field experiment proposal without generating substantive research content. Use it to design interventions, define control groups, clarify mechanisms, or prepare a study pitch.
Analyze data-construction code to trace observation counts across loads, merges, and filters. Use this skill to generate reproducible sample-selection flow diagrams, audit merge logic, and document data provenance for empirical papers.
Find cross-disciplinary research by translating domain-specific jargon into shared underlying mechanisms. Use this skill during literature reviews, theory building, or paper positioning when searching within a single field yields redundant results or misses relevant work from other disciplines.
Audit study designs against pre-experiment qualitative interviews to identify gaps, misalignments, and unsupported causal assumptions before launching a trial. Use this skill before fielding an experiment or observational study to confirm interviews support planned measures, treatments, and outcomes.
Transform field-experiment ideas from a Strategy Experiment Canvas into a structured three-paragraph journal introduction using Jay Barney's AMR guidelines. Use this skill when framing field experiments for top management journals or compiling introduction drafts into manuscript PDFs.
Structure and refine experiment ideas into the 8-box Strategy Experiment Canvas (Hasan, Kim & Koning). Use this interview-driven skill to clarify setup, subjects, friction, insight, intervention, mechanism, comparison group, and business impact for field tests or partner presentations.
Audit research manuscripts against live submission guidelines for top management journals like AMJ, ASQ, and Management Science. Use this skill before manuscript submission to generate a formatting compliance report and reformat document styling without altering author prose.
Guide researchers in analyzing open-ended text outcomes in randomized experiments using a structured detection-description-causation workflow. Use this when analyzing open-ended survey responses, chat logs, or free-text deliverables from field experiments or A/B tests.
Guide researchers through statistical power calculations and experiment design using David McKenzie's practical framework. Use it to determine sample size requirements, adjust for imperfect take-up, evaluate cluster designs, optimize measurement strategies, and calculate minimum detectable effects.