Is taking into account the subjects degree of knowledge and expertise enough when rating quality?

This paper provides the results of a speech quality listening test on Wideband (WB, 500-7000 Hz) and narrowband (NB; 300-3400 Hz) codecs for new wideband-based, so-called “high-definition” (HD Voice) telephony services. Aside of error-free coding, the experiment also evaluated the effects of packet loss and external noise. Unlike “expert vs non-expert differentiation” typical of speech quality research, users are classified into six groups according to demographic characteristics, their attitude towards adopting new technologies and socio-economic information. With this experiment, we want to present how factors beyond the level of prior knowledge of the users affect their perception of quality.