Chapter 6 142 1 and 2). We expected moderate, positive correlations between the SNR50, and the SRT of a speech reception test performed in continuous noise (hypothesis 3) and in fluctuating noise (hypothesis 4). In addition, we expected moderate, positive correlations between the SNR50 and the engineers’ subjective rating of concentration and effort it takes to detect auditory warning signals in the train cabin (hypothesis 5). A moderate correlation was expected rather than a strong correlation, since the subjective rating was provided for the purposes of an auditory fitness for job assessment. We hypothesized only a moderate correlation (and not a strong correlation), since there is a great risk of reporting bias. Locomotive engineers may under-report their subjective difficulties with detecting the auditory warning signals if they are afraid of not passing the fitness for job assessment (with the worst-case scenario of losing their job). Further, we expected to find a higher (poorer) mean SNR50 in the ambient noise of the Mat64 compared with V-IRM (hypothesis 6), since the analysis of Van der Hoek-Snieders et al. (2021) showed a less favourable acoustic environment in theMat64. Finally, we expected significantly higher (poorer) SNR50’s on the signal detection test in locomotive engineers wearing hearing aids, compared with engineers who do not wear hearing aids (hypothesis 7). Locomotive engineers who have decided to wear hearing aids are expected to experience more severe functional listening difficulties compared with engineers who do not wear hearing aids. Statistical analysis The a priori formulated hypotheses regarding the expected correlation between the signal detection test, and the conventional hearing tests were tested by calculating Pearson correlation coefficients. The assumptions with respect to normality and linearity were checked. Biserial correlation coefficients were calculated between the SNRs derived with the signal detection test and the subjective rating of effort and concentration it takes to detect auditory warning signals (Kraemer, 2014). Therefore, this subjective rating was dichotomized. The first category was reserved for engineers who answered that detecting auditory warning signals did either not take effort and concentration, and the second category for engineers who answered that detecting auditory warning signals did take extra effort and concentration to some extent. The hypotheses regarding group differences were tested by t-tests with a p cut-off value of 0.05, specifically a paired t-test for testing hypothesis 6 and an
RkJQdWJsaXNoZXIy MTk4NDMw