Odd-Even Oddball Task: Evaluating Event-Related Potentials (ERPs) during Word Discrimination Compared to Speech-Token and Tone Discrimination.

Abstract

The use of tonal and speech token auditory oddball tasks has been used abundantly in past literature; however, it has been argued that tasks using non-word sounds fail to capture the higher-level ability to interpret and discriminate stimuli based on meaning which are critical to language comprehension. As such this study aims to examine how neural signals associated with discrimination and evaluation-processes (P3b) from semantic stimuli compare with those elicited by pure tones and speech tokens. This study comprises of two experiments, both containing thirteen adults (Exp 1: (Mean(SD)age = 25.20(3.89) years and Exp 2: Mean(SD)age = 25.3(3.79) years) with normal hearing in both ears (PTA [≤] 20dB HL). Scalp electroencephalography and auditory event related potentials was recorded in free field whilst they completed three different oddball tasks: 1) Tones, 2) Speech tokens and 3) Odd/Even numbers. In experiment two, the duration of each stimulus was the same. P3b peak latency was significantly different between all three tasks, in both experiments. P3b amplitude was identified to be sensitive to reaction time, with tasks that have a large reaction time variability resulting in the P3b amplitude to be smeared out, thereby reducing the amplitude size. The findings from this study highlight the need to take into consideration all factors of the task before attributing any effects to any additional process such as semantic processing and mental effort. Furthermore, it highlights the need for more cautious interpretation of P3b results in auditory oddball tasks.