REF2 should divide refusals and break-offs (R) by the total of complete interviews (I), partial interviews (P), refusals and break-offs (R), non-contacts (NC), other non-interviews (O), and the estimated eligible share (e) of cases whose eligibility is unknown because housing status is unknown (UH), because the sampled unit's eligibility is unknown (UR), or for other reasons (UO). Like RR3 and RR4, it includes an estimate of eligible cases among the unknowns rather than assuming none.
The graph holds this control, the 0 it maps to, and the evidence behind each claim, over MCP and REST.