Engineering / research
Conduct a credible systems investigation
Construct falsifiable systems claims, controlled comparisons, causal performance explanations, and reproducible artifacts that survive competing interpretations.
Your route through engineering
Follow the path or begin with a question. Every lesson is open.
Your place, at your pace. Progress is optional and saved only in this browser when available.
Read / explore / reason
Inside this module
Module 01 research
Conduct a credible systems investigation
Construct falsifiable systems claims, controlled comparisons, causal performance explanations, and reproducible artifacts that survive competing interpretations.
- 01
Systems claims and falsifiable hypotheses
Turn a promising observation into a scoped hypothesis with measurable outcomes, a practical threshold, and a result that could contradict it.
30 min - 02
Workload representativeness and trace provenance
Connect a performance claim to its target workload through explicit sampling, trace transformations, and population weights.
30 min - 03
Baseline equivalence and tuning fairness
Compare complete systems under equivalent correctness and resource contracts while accounting for search effort and tuning selection.
30 min - 04
Factor isolation and ablation design
Use controlled factor combinations to distinguish main effects, interactions, and the narrower conditional meaning of an ablation.
30 min - 05
Repeated runs and variance decomposition
Allocate repetition to independent experimental levels instead of treating every nested timing as a new system execution.
30 min - 06
Performance counterfactuals and mechanism tests
Test why an improvement occurs with explicit interventions, predicted mediators, and alternatives to a favorable profile.
30 min - 07
Simulation calibration and external validity
Separate implementation verification, fitting parameters, held-out validation, and transfer beyond a simulator’s tested application domain.
30 min - 08
Negative results and failure-envelope mapping
Turn regressions, timeouts, and unsupported conditions into a precise map of where a mechanism succeeds or fails.
30 min - 09
Artifact packaging and independent reproduction
Package the complete evidence path from immutable inputs to claim-linked results, and report exactly what independent execution established.
30 min - 10
Research claim scope and limitation writing
Write conclusions that preserve the measured domain, uncertainty, comparison contract, and unresolved alternatives.
30 min