Measuring Statistical Learning Across Modalities and Domains in School-Aged Children Via an Online Platform and Neuroimaging Techniques

The ability to extract patterns from inputs is foundational for language and cognitive development. The technique can address whether the remarkable heterogeneity of statistical learning in children is explained by domain general or domain specific mechanisms. Our protocol allows for the measurement of statistical learning across domains and modalities within individuals, is child-friendly and combines web-based behavioral with lab-based neuroimaging techniques.

This method will provide insight into the learning frameworks of language development. If applied in special populations, our protocol may also advance our understanding of language learning difficulties. We are in the process of making our tasks accessible to the research community on Zenodo and GitHub.

We encourage other researchers to reproduce our tasks once they become available. To perform a web-based protocol, first navigate to the web-based statistical learning paradigm. For the syllable task, introduce the participant to an alien and its favorite word in its alien language.Doe.

Inform the participant that they will listen to the alien's language and to remember to press the space bar whenever they hear the favorite word. Have the participant complete a practice trial before the familiarization phase, in which they must press the space bar as soon as the alien's favorite word is heard. During the test phase, instruct the participant to select one of the two sequences that sound more familiar.

The target sequence will be presented in the familiarization phase. Klaptoo the foil sequence will not have been previously presented. Klaptoo In the tone task, inform the participant that they will have to keep track of the alien favorite note in its folk song.

The test phase will follow the same design as the syllable task. In the image task, informed the participant that they will have to keep track of a special alien as a group of aliens lineup to enter a space ship. During the test phase present both targets in the familiarization phase and foil triplets not seen before to the participant in pairs.

In the letter task informed the participant that they will have to keep track of the aliens favorite sign. As the alien holds up signs for a parade. To help participants especially children become comfortable in scanner practice the MRI scanning session with a mock scanner.

Introduce the participant to the mock scanner or brain camera and to their scan buddy. Explain that the purpose of the scan buddy is to keep them company and to help them if they need anything. The scan buddy will also gently remind the participant to keep still if too much motion is detected by the camera.

Introduce the participant to the statistical learning paradigm and have them complete a brief portion of the task on a computer. After completing the task, help the participant into the mock scanner and play child-friendly videos to help them acclimate to the sound and video. When the participant is ready, play a few prerecorded scanner or sound clips to prepare them for the noises produced by the real MRI.

During this time, have the participant practice staying still and working with the scan buddy. When the participant is comfortable with the mock scanner confirmed that the FMRI data collection protocol is appropriately set up on the MRI acquisition computer. Have the participant lie comfortably on the bed of the MRI scanner with headphones that protect their ears from the scanner noise and a response pad in their hand.

Place additional padding around the participant's head to ensure limited head motion during the data collection and give the participant the option of a scan buddy. Give the participant a squeeze ball to use to notify if they are distressed or need to stop and place the head coil over the participant's head. After aligning the participant in the scanning bed register a new participant in the acquisition computer and choose the correct scanning protocol.

Insert the scanning bed and participant into the bore of the MRI and show a movie to the participant while acquiring a structural MRI scan. To have the participant perform an auditory task use an Intercom system connected to their headphones to inform the participant. Now we are going to play a button pressing game.

You will hear the aliens, say words and play music remember to press the button on the response pad whenever you hear the sound you're listening for. After giving the instructions, start the paradigm on the presentation computer to allow the acquisition of the auditory statistical learning data. For a visual task inform the participant.

Now, you're going to see the pictures of the aliens and the letters. Whenever you see the pictures that you're looking for press the button on your response pads. After giving the instructions start the paradigm on the presentation computer to acquire the visual statistical learning data.

Once the participant has completed the paradigm stop the MRI, safely remove the participant from inside the scanner and remove the head coil. In the web based statistical learning task. Children performed significantly better than a 0.5 chance level on all conditions indicating successful statistical learning at the group level.

The mean reaction time slope was negative and significantly below zero in the syllable condition and marginally significant in the letter condition suggesting a faster acceleration of the target detection during the familiarization phase in the linguistic tasks. The mean reaction time slope was not significantly different from zero in the image or the tone condition despite evidence of learning in the offline measures of accuracy. Here preliminary FMRI results based on data from nine developing school aged children who completed auditory and visual statistical learning tasks are shown.

When comparing structured blocks to random blocks significant clusters were observed across a variety of brain regions, depending on the stimulus type. These results suggest that the current task design is sensitive to detecting learning induced hemodynamic changes in children's brains. Representative findings in adults indicate that the structured sequences are learnable supported by a significantly quicker response time in the structured condition, compared to the random condition in all of the in scanner tasks with the exception of the tone task.

We encourage researchers, to include other individual difference measures to better understand how variability and statistical learning across domains and modalities relates to higher level processes. This protocol is a first step towards delineating how individual differences in statistical learning may account for variation in language outcomes in both typical and atypical development.