This is a GitHub to create the project for PCBS. The pages can be found on : https://mathiasvigouroux.github.io/PCBS_vigouroux/ Ans it is possible to launch the experiment online from : https://mathiasvigouroux.github.io/PCBS_vigouroux/Script_java/main.html For the moment only one tone is played on the online version. In the local version it is woking well.
These experiments are smaller version of the real experiment.
The code can be found in the Script_java file and is named main.html.
Cloning the respository on you desktop allow you to directly run this experiment locally.
The master thesis is the paper from which the experiment is coming from.
The script comport folder contains the initial matlab code, which was using pscyhtoolbox.
The “foreperiod paradigm” enable scientist to study the role of temporal predictions. This paradigm is schematized in 2.1. More precisely, participants heard two auditory cues separated by a foreperiod. An auditory cue was selected because it has been shown that implicit temporal predictability enhances pitch discrimination sensitivity [1]. During these blocks, the participant were only asked to answer as quickly as possible to the Target stimuli.
The foreperiod were drawn in order to study the effect of the variability of the foreperiods on our behavior. Before the beginning of the experiment there was a training phase. In this training phase, participants are presented with a foreperiod distribution that was distinct from the foreperiod used in the rest of the experi- mental paradigm. Indeed, a distribution with a mean of 2.5 seconds and 0.25 second for the standard deviation was chosen. A Weber Fraction of 0.1 was therefore selected for the distribution of the foreperiods during the training phase. Thus, this distribution was likely to be easily learned. In the main task, four foreperiod distributions were selected, all Gaussian, with a mean of 2.5 s and varying standard deviation. The mean duration of the foreperiods is 2.5 seconds in order to have 120 trials leading to 14 min of trials and 4 blocks . Indeed, for the experimental design we have decided to use four blocks. For these decisions, a major determinant was the block duration (<15min) and duration of the experiment (<1.5h).
The standard deviations were drawn on a log scale between 0.04 and 1.1. Their histograms are depicted in C.3. The limiting factor for the upper limit of this range is the learning of the std by the Mean Std observer in a finite amount of time. On the other end, the inferior limit was chosen to be almost 10 times smaller than the std gave by a Weber fraction of 0.1,a classical value, and the mean of 2.5. Moreover, human participants have shown to have a limit for their temporal resolution and won’t differentiate temporal intervals separated by only just a few milliseconds. The choice of the log-scale goes once again with the findings of Weber. Indeed Fechner derived a logarithmic relation between absolute physical and sensed magnitudes, the Weber-Fechner law.
The foreperiods are separated by an Inter Trial Interval (ITI). Inter Trial Interval were drawn following a uniform distribution between 0 and 6 seconds in order to have a mean of 3 seconds. Thanks to that the participants can’t predict the onset of the cue stimulus. At the beginning of the block participants wait for 5 seconds in order for them to be more focused at the appearance of the stimulus.
Moreover, in order to better distinguish the different blocks, the tone of the cue tone was also varying from 400 HZ to 1000 Hz. The tone of the target tone remained the same through the entire experiment.
## Feedbacks Participant were only instructed to press the space-bar as quick as possible after the target tone and to gaze at a fixation cross in the center of the screen. The fixation cross at the center of the screen would ensure the same condition for attention for all of the participants. In order to motivate participants to have better performance, they were instructed that depending on their performance the color of the fixation cross would change for a few seconds (it changed for 1 sec). If their reaction time was really good, i.e. below 0.2 second, the fixation cross would turn green for 1 second. If it was intermediate, between 0.2 second and 0.4 second, it would turn orange for 1 second. If it was long, above 0.4 second, it would turn red. These thresholds were defined empirically in order to have from time to time green feedback if really focused, orange most of the time, and red only if completely distracted. Participants were also informed if they answered too early, i.e. pressing the space-bar between the cue tone and the target tone. Participants were also informed if they did not answer after 2 seconds after the target tone and the code passed to the next foreperiod.