jsQuestPlus: A JavaScript implementation of the QUEST+ method for estimating psychometric function parameters in...

Area of Science:

• Psychophysics
• Computational Psychology
• Human-Computer Interaction

Background:

• Bayesian adaptive psychometric methods like QUEST and QUEST+ are standard for laboratory experiments.
• The rise of online psychophysical experiments necessitates web-based implementations of these methods.

Purpose of the Study:

• To introduce jsQuestPlus, a JavaScript library for the QUEST+ psychometric method.
• To evaluate the performance and accuracy of jsQuestPlus for online psychophysical studies.

Main Methods:

• Development of a JavaScript library, jsQuestPlus, for the QUEST+ algorithm.
• Integration of jsQuestPlus with online experimental platforms (jsPsych, PsychoPy/JS, lab.js).
• Performance evaluation through computation time measurements and accuracy comparisons across various browser-computer combinations.

Main Results:

• jsQuestPlus demonstrated efficient performance with mean initialization time of 461.08 ms and update time of 79.39 ms.
• Stimulus parameter computation time was consistently under 1 ms.
• Estimated psychometric parameters showed accuracy comparable to the original QUEST+ method.

Conclusions:

• jsQuestPlus is a computationally efficient and accurate tool for conducting online psychophysical experiments.
• The library overcomes the complexities of matrix calculations for web-based applications.
• jsQuestPlus is freely available under the MIT license for broader research use.