Part III: Measuring hearing-critical job tasks 149 adaptive two-interval, forced-choice procedure used, the signal detection test differs from locomotive engineer’s listening task to detect acoustic warning signals in their working environment. Since the listening intervals are marked by this procedure, and the test subject is forced to choose, the signal detection test results in the most critical SNR50 in which the warning signals can be detected. In a real working environment, signals at this SNR will likely not be reliably detected by the engineer. Therefore, it is advised to present signals at least 12 dB higher than the signal level of the measured SNR50 (Giguère et al., 2008; Hung & Hétu, 1996). Although the signal detection test would have even higher face validity when it would replicate all job characteristics, such a test would be unfeasible to administer, especially because the job task under study has a long duration (Beck et al., 2016; Payne & Harvey, 2010). When all job characteristics would have been replicated, this would have resulted in a test setting in which warning signals need to be recognized at the level at which they are normally presented in the train cabin, while the engineers perform a broader driving test. The duration of this test would have been 2.5 hours since this is reported to be the longest consecutive amount of time that a locomotive engineer spends in the train cabin (Zoer et al., 2014). The method used in this study has resulted in a test that is job and task specific as well as easy to administer (Payne & Harvey, 2010). The fact that the signal detection test does only replicate the acoustic environment of a work situation provides opportunities for broader use of the test. By expanding the test with other workplace signals and noises, it could also be used to evaluate the detectability of warning signals in other professional settings that require sufficient signal detectability. Study limitations Some study limitations should be noted. First, the sample size of experiment 1 was relatively low, and the study sample of experiment 1 did not include locomotive engineers. Adequate signal detection might be easier for engineers who are familiar with the presented noise environments and warning signals. This might implicate that normal-hearing engineers would achieve even a lower SNR50. However, all participants performed a practice round to get familiar with the noise environment and the signals. All normally-hearing subjects performed the test two times in twelve driving conditions. It can be concluded that there was no learning effect since there was no systematic difference between the first and the second time. Second, only two
RkJQdWJsaXNoZXIy MTk4NDMw