Data & measurement
Each Kifuliiru variable v_i(t) is observed on its native scale, then norm_i maps it to [0,1]. Weights α_i reflect importance for Kifuliiru documentation and adoption and should sum to a documented total; only variables in 𝒞 ∪ 𝒜 with known values enter each run. Version the weight table whenever stakeholders rebalance Kifuliiru corpus vs adoption emphasis.
Solution & proof
LIS(t) is a weighted average of normalized Kifuliiru scores: numerator ∑ α_i norm_i(v_i) and denominator ∑ α_i over the active set. Hence LIS lies in [0,1] when every norm_i does. Missing variables are excluded from both sums so remaining weights renormalize—this is the score-with-missing-data policy applied to the full Kifuliiru index.
Worked examples and interactive charts now live on their own page.
Open the playground to test sample values, review worked scenarios, and explore how this formula behaves across multiple cases.
Source & tags
Source
Impact measurement framework — Core Architecture