Beyond the BeatSIYAA’S PULSELAB
Documentation library

Research / PUBLIC-SAFE SUMMARY

Questions mapped to graphs

Implemented with documented inference limits

Exercise comparisons show paired participant values, exercise means and descriptive Student-t intervals. Exact-age scatter plots and trends preserve information that age groups would hide.

The native random-intercept REML model includes age × exercise and categorical exercise position, with repeated observations grouped by participant. Adjusted curves, slope differences and exercise contrasts use approximate normal Wald intervals.

Observed-versus-fixed-predicted values, participant mean residuals and residual checks explore variation left by the model. In-sample descriptive R² is not causal explained variation.

The model rejects incompatible versions, duplicate outcomes and unidentifiable designs. Missing metrics are not imputed. Final assumptions, influence, missingness and uncertainty interpretation still require scientific review.

Sex, activity and height/weight-derived summaries are exploratory, not primary causal explanations. Suggestions remain disabled; reference accuracy and repeatability are not established.

Supporting workspace artifacts

Source locations identify the original artifacts in the private project. Raw files are not publicly served. Hashes identify the reviewed source snapshot; they are provenance, not external certification.

Download public-safe provenance