Listening-test statistics must inform experimental design from the outset

Evidence

An important part of the research on psychoacoustics is the statistical analysis of the results from listening tests. Actually, in some cases, the statistical considerations should be taken into account in the design of the experiments. Tests should be conducted with the proper number of subjects and a meaningful selection of stimuli and tasks, listening conditions, and repetitions. In many cases, such designs should be conducted carefully so that the results prove or disprove the existence of the phenomenon that is hypothesized based on informal listening before the test. — Ville Pulkki, Matti Karjalainen, Communication Acoustics- An…, 948

Categorization

• XREF: Echoes standard research-design guidance across experimental psychology; connects to general knowledge about power analysis and stimulus selection. — connects to: The indirect measurement method required a long questionnaire to yield statistical soundness — Rolf Wiggershaus, The Frankfurt School: Its History,…, 1579