Statistical Criteria to Manage Non-respondents’ Intensive Follow Up in Surveys Repeated along Time

When carrying out statistical surveys, detecting non-respondent units to undergo Intensive Follow Up (IFU) to disseminate quality estimates represents an important though under estimated aspect that is often not deeply studied. We suggest to use a generalised score function to quantify the risk related to the failed use of all the sample units for calculations together with techniques aiming at detecting IFU units through significance tests or methods – either parametric or not – based on acceptance thresholds. We also present an exercise related to a panel of industrial enterprises.