Standardization of a test of speech perception in noise (SPIN)

All ten forms of the SPIN test [D. N. Kalikow, K. N. Stevens, and L. L. Elliott, J. Acoust. Soc. Am. 61, 1337–1351 (1977)] were presented to 128 listeners who had some sensorineural hearing loss. Speech materials were presented at a level calculated to correspond, approximately, to the loudness equivalent to 60 dB SPL for a normal-hearing listener, and the babble noise was set 8 dB lower (S/B = +8 dB). Half of the subjects heard the test through earphones and half via loudspeaker; half were tested in a single session and half in two sessions. Statistical analysis of these results indicated that (a) transducer, (b) visits, and (c) order of testing had no influence on test scores. The average reliability coefficient was 0.906 for the high-probability items in a form and 0.848 for the low-probability items. The ten forms tested, however, do not constitute a set of equivalent forms; and it is difficult to identify a subset of equivalent forms, because of the large differences in difficulty among forms with respect to the low-probability sentences. [Work done under contract to NINCDS.]