Recorded Text Testing
Recorded Text Testing (RTT) can help surveyors understand how well people in certain areas understand particular speech varieties. The first form of the method was used by C.F. Voegelin and Zellig S. Harris to determine dialect intelligibility. They published their work as Determining Intelligibility among Dialects of Natural Languages in the Proceedings of the American Philosophical Society, vol. 95, no. 3 in June 1950.
Others have since refined the methodology, notably Eugene Casad in his 1974 book Dialect Intelligibility Testing. These early studies were carried out mostly on Native American languages and in Mexico. Provided that enough social distance remains between communities that the tool is employed in, material for the test is likely to be valid. In locations where community connections are strong. In 2003, an RTT was attempted in Western Province of Papua New Guinea. Due to the strong connections between even geographically distant communities however, the test was rendered invalid as test takers were able to identify who was speaking on the recording. RTTs are not always valid assessment tools therefore.
The pages linked in the table to the right can help walk you through the RTT process.
Some sample RTTs are here.