Similar to reference periods, the way in which response formats are described or labelled sets a reference frame for individuals as they construct their answers. Scale labels can thus potentially influence mean values by communicating information about the expected range within which responses can fall. Variations in labelling the response scale can also affect the accuracy and reliability of scores, with response options that lack meaning or clarity being more likely to introduce random responding or satisficing.
A key decision for question design is whether to provide a verbal label for every response option (e.g. would you say that you are very happy, pretty happy, or not too happy?) or whether to simply label the scale end-points or anchors (e.g. 0 = completely dissatisfied; 10 = completely satisfied) and rely on numerical labels for scale intervals. As highlighted in the previous section, this choice will need to be considered in combination with other factors, such as the number of response options to offer and the survey mode being used.