ABSTRACT
Survey research is in the midst of an era of extraordinary new developments and challenges. People have increasingly moved away from having landline telephones to having mobile phones or no phone at all, which poses new challenges for all aspects of telephone surveying. Survey response rates have continued their now decades-long pattern of decline. Alarm over declining response rates has been tempered by research showing that the extent of bias in survey estimates is at best weakly related to the survey response rate. Survey researchers have long sought to design surveys so as to optimize the response rate conditional on a budget constraint, thinking that as the response rate increased, so too would the representativeness of the sample. Studies distributing the same questionnaire to respondents from different online panels demonstrate substantial variation across the panels in how well the results match population benchmarks.
