Survey respondents are routinely asked to answer batteries of items employing a single rating scale. Satisficing theory raises the possibility that respondents might shortcut the cognitive processes required to provide optimal responses and respond identically or nearly identically to all items in the battery so as to provide answers that appear to be reasonable while investing minimal effort, a behavior termed "nondifferentiation." Nondifferentiation is posited to be most likely to occur among respondents who have limited cognitive skills and who have minimal motivation to answer carefully, and when the response task is cognitively demanding. Some existing evidence suggests that respondents lower in cognitive ability are most likely to nondifferentiate, but little prior work has explored the effects of respondent motivation and task difficulty on nondifferentiation. We sought to test thoroughly whether satisficing implicated factors (low cognitive ability, low levels of respondent motivation, and high task difficulty) in fact increase nondifferentiation. To do so, we analyzed data from several sources, including telephone surveys conducted for the National Election Studies and by the Center for Survey Research at The Ohio State University, and Internet surveys conducted by Harris Interactive, Knowledge Networks, Greenfield Online, GoZing, and similar companies. This research is unique in that it also tested the mediational hypotheses posited by satisficing theory. Specifically, we tested whether cognitive ability, need for cognition, and need to evaluate affect respondent effort and the difficulty of the response task, which in turn affect the extent of nondifferentiation. Separate structural equation models estimated using each dataset provided converging evidence supporting these predictions. Thus, it seems that nondifferentiation in particular, and perhaps other satisficing-related compromising response strategies, may be reduced by increasing participant motivation and decreasing task difficulty.