Stimulus presentation and task

For each participant, the experiment included two sessions on two different days, with the same stimuli but a different task: the “active” session required attentive listening, and a behavioral report on the stimuli; the “passive” session involved passive listening, where the task on the stimuli was replaced by distracting tasks as well as tasks allowing to estimate mind wandering^67,68.

Each trial started with the presentation of a fixation point, which remained on screen until the appearance of a response screen. The target vowel could be played any time between 1 and 3 s after the beginning of the trial. It lasted 200 ms, and was followed by a random delay between 2 and 3 s before a response screen was presented. Participants were asked to keep their eyes on the fixation point and to avoid moving or blinking, except during the response periods. The TEN noise was present continuously throughout each block; it stopped only during the pauses between the blocks. Participants used the computer mouse to give their response. Written instructions were provided to the participants, and we checked with them for their understanding. Each experimental session started with a training block.

During the active session, participants were asked to perform two tasks on each trial:

Identify the vowel: the two response options, “A” and “E”, were shown to the left and right of the fixation point (initial position of the mouse at fixation), with the side assigned to each vowel changing between blocks, as announced at the beginning of each block. Positioning the cursor with the mouse over one of the letters highlighted a square around this choice. The choice was validated by a left click. Participants were asked to make a guess when they were not able to identify the target.

Report the subjective audibility level of the vowel: once they had responded to the first question, participants were asked to report how well they heard the target by placing a cursor on an analogous scale represented by a horizontal bar, labeled “0” on the left and “max” on the right. The initial position of the cursor was random. Participants could slide the cursor on the scale with the mouse, and validated their choice with a left click. They were instructed to use the entire scale, as follows: answer zero when they did not hear anything, use the left half of the scale for cases where they doubted a vowel was present, and the right side when they were certain a vowel was present; move the cursor to the right when the sound was more audible, independently of their ability to identify it. The highest answer was to be used when the vowel was perceived as the most audible stimulus in this experiment. The scale had 11 levels (0–10), “0” and “max” at both ends of the scale were the only landmarks presented to the participants.

Prior to the experiment, the participants were trained on both tasks during a training block where they received feedback on identification responses and where vowel at maximum volume were signaled, to allow participants to adjust their audibility judgment. During the experimental blocks, feedback on identification performance was only given at the end of each block.

During the passive sessions, the stimulation periods were identical to those in the active session, but at the end of each stimulation period, the task on the stimulus was replaced by one of the following four tasks, randomly intermixed across trials:

A speeded response task: participants were required to click on the mouse whenever a large green circle was displayed at fixation.

Mind wandering probe: participants were asked to answer the question “What is on your mind just now?”, by choosing one of the four options: “the sound “, “the task”, “my thoughts”, “nothing/I feel sleepy”.

Quiz: participants were asked to answer simple questions (arithmetic operations, questions of general culture…) with four answer choices.

Just a “Click to continue” message.

At the end of each block, participants answered a questionnaire on their experience (attention, mind-wandering…) during the previous block. The questionnaire is reproduced below:

In how many trials did you hear and recognize a vowel?

In how many trials did you hear something without recognizing it?

How often were you thinking of the sound? (0) never (1) once (2) sometimes (3) often (4) very often.

How often were you thinking of the various tasks you had to do? (0) never (1) once (2) sometimes (3) often (4) very often.

How often were you thinking about visual fixation, eye blinks, and how comfortable or uncomfortable you were? (0) never (1) once (2) sometimes (3) often (4) very often.

How often were you thinking about something else? (0) never (1) once (2) sometimes (3) often (4) very often.

How often was your attention captured by something in the environment that was not the experiment? (0) never (1) once (2) sometimes (3) often (4) very often.

How often were you falling asleep? (0) never (1) once (2) sometimes (3) often (4) very often.

Each session contained 960 trials (2 vowels × 6 sound levels × 80 repetitions), grouped into 20 blocks of 48 trials (about 7 min per block), for a total duration of 2.5–3 h with pauses. An optional additional block was provided in the event of a technical problem leading to the loss of part of the data. By combining the two vowels, we obtained 160 trials per sound level (before removing the artifacts). For the last ten subjects, we included an additional SNR level of −3dB in the passive session, in prediction of a shift in auditory threshold during this session relative to the active one.