OpenSST: A Browser-Based, Bilingual Stop-Signal Task for Research on Response Inhibition

Abstract

SST Test is a browser-based implementation of the stop-signal task (SST), a widely used behavioral paradigm for measuring response inhibition. The software is designed for researchers in psychology, psychiatry, and cognitive neuroscience who need a simple, self-contained tool for collecting stop-signal data. The task design follows the recommendations of the international consensus guide on the stop-signal task (Verbruggen et al., 2019, eLife) and the independent race model originally proposed by Logan and Cowan (1984). Participants perform a two-choice go task (left/right arrows) and withhold their response on a minority of trials when a stop signal (visual change plus optional tone) appears after a variable stop-signal delay (SSD). The SSD is adjusted trial by trial using a standard 1-up/1-down tracking procedure (default step: 50 ms, within a configurable range), which converges the probability of responding on stop trials toward approximately 50%. The primary outcome measure, the stop-signal reaction time (SSRT), is estimated with the integration method, in which go omissions are replaced by the maximum response window as recommended in the consensus guide; the median-based estimate is also reported as a reference. Several data-quality indicators (e.g., number of stop trials, probability of responding, go omission rate) are displayed alongside the results to help researchers judge whether SSRT estimates are reliable under the guidelines. The software is a single self-contained HTML file with no external dependencies. It runs entirely in the browser, works offline, and requires no installation, no server, and no programming knowledge. The interface is fully bilingual (English and Chinese). The default parameters are set to values consistent with common practice in the stop-signal literature, while all task parameters (number of practice and main trials, number of blocks, stop-trial probability, SSD initial value/step/range, response window, fixation duration, inter-trial interval, stop-signal modality, and response keys) can be adjusted through the settings page, with a one-click option to restore the defaults. A practice phase with per-trial feedback is included, followed by the main task divided into blocks with rest screens between blocks. At the end of the session, the software computes a summary of key measures (SSRT by integration and by median, mean/median go RT, go RT variability, go accuracy, omission and choice-error rates, stop-success rate, probability of responding, mean SSD, and mean RT on failed-stop trials), shows the go-RT distribution with failed-stop RTs for inspection, and exports per-trial data in CSV and JSON formats for downstream analysis in R, Python, SPSS, or other statistical software. This tool is intended for research purposes. It is not a diagnostic instrument, and its results should not be used alone to diagnose or classify any clinical condition. As noted in the consensus guide, individual-level interpretation of SSRT requires a sufficient number of stop trials, and data quality should be evaluated before any statistical analysis. Users are encouraged to review the consensus guide (Verbruggen et al., 2019) for detailed recommendations on design, analysis, and reporting. For research use only; not intended for clinical diagnosis or treatment decisions. This software was developed by students from Hangzhou No. 4 High School with the assistance of artificial intelligence (AI); the code, interfaces, and parameters were generated with AI assistance and have undergone preliminary manual review. Please feel free to contact us via email (15858235995@163.com) regarding any issues or suggestions for improvement.