SAM Standard Segmental Test

Status Completely developed test with software (SOAP) for item construction, test administration, and data analysis [Chapter 4]Howard-Jones92a.

Goal Diagnostic and comparative evaluation of the intelligibility of single initial, medial, and final consonants.

Languages In principle any language. The only thing needed are lists of permissible consonants in initial, medial, and final positions.

Items Open vocabulary, mostly meaningless but also (by chance) meaningful, of the structure CV, VC, and VCV, comprising all phonotactically permissible combinations of initial, medial, and final consonants and three point vowels, e.g.\ /i/, /u/, and /a/. Examples: pa, ap, apa.

Procedure Open response identification , preferably using unambiguous notation.

Time With 20 consonants in a particular word position and 3 vowel contexts, a stimulus list contains 60 items. With an interstimulus interval of 4 sec this amounts to about 15 min for three lists. With two randomly ordered lists for each structure (recommended) time is doubled to 30 min per synthesiser.

Analysis SOAP contains a pre-processing module to ``clean'' the responses. In addition it offers five scoring methods: (1) consonant confusion matrix , (2) percentage correct per consonant, (3) percentage correct per consonant per stimulus, (4) percentage correct per self-defined group of consonants, (5) group confusion matrix.

Status	Completely developed test with software (SOAP) for item construction, test administration, and data analysis [Chapter 4]Howard-Jones92a.

Goal	Diagnostic and comparative evaluation of the intelligibility of single initial, medial, and final consonants.

Languages	In principle any language. The only thing needed are lists of permissible consonants in initial, medial, and final positions.

Items	Open vocabulary, mostly meaningless but also (by chance) meaningful, of the structure CV, VC, and VCV, comprising all phonotactically permissible combinations of initial, medial, and final consonants and three point vowels, e.g.\ /i/, /u/, and /a/. Examples: pa, ap, apa.

Procedure	Open response identification , preferably using unambiguous notation.
Time	With 20 consonants in a particular word position and 3 vowel contexts, a stimulus list contains 60 items. With an interstimulus interval of 4 sec this amounts to about 15 min for three lists. With two randomly ordered lists for each structure (recommended) time is doubled to 30 min per synthesiser.

Analysis	SOAP contains a pre-processing module to ``clean'' the responses. In addition it offers five scoring methods: (1) consonant confusion matrix , (2) percentage correct per consonant, (3) percentage correct per consonant per stimulus, (4) percentage correct per self-defined group of consonants, (5) group confusion matrix.