View article

Beginning with the use of “split ballots” in the early days of opinion polling in the United States, researchers have made advances in social science through experiments in national opinion surveys. Such experiments are important means to study cause-and-effect relationships in public opinion research, with evidence for the external validity provided by the use of representative national samples. The ability to do such experiments, however, has been hindered by the costs of fielding experiments in national surveys and by the long lead-time needed to apply for and receive large-scale grants for individual projects. For this reason, most survey experiments have been restricted to small-scale laboratory settings or to convenience samples.

To overcome these barriers, Arthur Lupia and Diana Mutz designed the Time-Sharing Experiments in the Social Sciences Project (TESS). Funded by the National Science Foundation …