The datasets

    Terms used on this page

    What ran

    Calibration quality

    By grain

    Against the enhanced FRS

    National rows, by source

    Constituency rows, by target family

    Local authority rows, by target family

    The same comparison as a table

    HBAI basis

    Support and weights

    Constituencies by Kish effective sample size. The release floor is 50.
    Local authorities by Kish effective sample size.

    Weights

    The two selections against each other

    Six gates on the compact frame

    Downstream

    Reforms, admin anchors and distributions

    T5 · eight reforms

    T4 · admin anchors

    T3 · variable distributions

    T6 · dashboard replays against the enhanced FRS

    Footprint

    Why the totals move

    Not measured

    What these runs cannot say

      Signed exclusions

      Local authorities excluded from the support floor

      Coverage against the incumbent

      Incumbent target rows this candidate does not bind, by status

      National targets signed out of our reference before binding

      What Chronicle's source checklist says about the same families

      Sources and reliability

      Where every expected value, anchor and benchmark comes from

      Where the numbers live