Performance with linear and a non-linear signal processing in analog and digital hearing aids was evaluated by 25 severe-to-profound hearing impaired subjects. The subjects were randomly selected from the clinic's files and constituted a representative sample of the clientele. Two types of evaluation were performed: a single blinded field study designed according to "Minimum Nordic requirement for clinical testing of HA" and a double blinded paired comparison experiment in which different modes of amplification were compared in various sound environments. The subjects compared a digital power hearing aid (SENSO P38) programmed in 2 ways, with their recently fitted analog hearing aid. The digital power hearing aid was programmed
1) according to the manufacturer's specifications and 2) as a linear hearing aid according to NAL-RP. In the field test, each type of amplification was worn for at least 1 month and the test included speech recognition tests in quiet and noisy backgrounds and interviews. After the field trial, the paired comparison tests were made. Speech recognition was significantly better with either type of digital HA than with the subjects' analog HA. The subjects preferred the non-linear HA over the linear and own HA when rating sound quality of speech and music. The preference of digital non-linear and linear HA was dependent on presentation level of background noise when rating speech recognition.