Meta-analysis of variables that affect accuracy of threshold estimation via measurement of the auditory steady-state response (ASSR).

Reported are the results of meta-analyses of data derived collectively from a sample of 56 published research studies on electric response audiometry (ERA) using auditory steady-state responses (ASSRs). Several specific methodological issues were examined and hypotheses were posited to rigorously test common conclusions drawn from the ASSR literature on the accuracy of ASSR-ERA. Explanatory variables for analyses were type of population (normally hearing and hearing-impaired), type of modulation, number of sweeps acquired during response analysis, electrode montage, and modulation rate (80 vs. 40 Hz). No explanatory variables were found to be significantly related to the degree of disparity between thresholds obtained by ASSR-ERA versus behavioral audiometry in the normally hearing population. Conversely, all but one explanatory variable (i.e. electrode montage) was found to be significantly related to mean threshold differences in the hearing-impaired and combined populations. Results both substantiate some of common conclusions drawn from the literature but call others into question, helping to identify those methodological issues which appear to, or not to, significantly affect the accuracy of estimating threshold using ASSR measurement. In addition to these findings, another practical outcome of this study was the development of various summary tables of the data analysed from the literature sampled.