Surveys are frequently employed in needs assessment to collect information about gaps (the needs) in "what is" and "what should be" conditions. Double-scale Likert-type instruments are routinely used for this purpose. Although in accord with the discrepancy definition of need, the quality of such measures is being questioned to the point of suggesting that the results are not to be trusted. Eight factors supporting that proposition are described with explanations of how they operate. Literature-based examples are provided for improving surveys with double scales especially as they relate to attenuating the effects of the factors. Lastly, lessons learned are offered with a call for more research into this issue in assessing needs.