A Bayesian approach for estimating the probability of trigger failures in the stop-signal paradigm
DOI: 10.3758/s13428-015-0695-8
archive: archived pipeline: cataloged verified
Get this paper ↗ (DOI — opens at the source; we link to it, we don't host it)
Summary
This paper addresses a critical methodological limitation in the stop-signal paradigm, a standard experimental task used to measure response inhibition. Existing models estimate the latency of the stop process (stop-signal reaction time, or SSRT) but fail to account for "trigger failures"—instances where participants fail to encode or interpret the stop signal, preventing the inhibition process from initiating. The authors argue that ignoring trigger failures biases SSRT estimates, potentially leading to incorrect conclusions about group differences in inhibition speed, particularly in clinical populations like those with ADHD or schizophrenia. To resolve this, the paper introduces a Bayesian hierarchical model that simultaneously estimates the probability of trigger failures and the full distribution of stopping latencies. The method builds upon the Bayesian parametric BEESTS framework, which models go and stop reaction times as ex-Gaussian distributions. The authors augment this model with a parameter, $P(TF)$, representing the probability of a trigger failure. This creates a mixture model where signal-respond trials (failed inhibitions) result either from a trigger failure or from the go process winning the race against a successfully triggered stop process. The model supports both individual and hierarchical estimation, with hierarchical modeling allowing for "shrinkage" of extreme estimates and inference at both individual and group levels. Parameter estimation is performed using Metropolis-within-Gibbs sampling. The authors validate the approach through two simulation studies: one assessing asymptotic performance with large datasets and another examining parameter recovery under realistic sample sizes using both fixed-SSD and staircase-tracking procedures. The simulation results demonstrate that the proposed model accurately recovers SSRT distributions and trigger failure probabilities, whereas standard BEESTS and traditional methods (integration and mean methods) significantly overestimate SSRTs when trigger failures are present. The new model performs well even with relatively frequent trigger failures (up to 20%) and requires fewer observations than previously assumed for stable estimates. When applied to two published datasets involving healthy controls from studies on schizophrenia, the trigger-failure model provided a better fit than the standard model. Crucially, the analysis revealed that trigger failures were clearly present in both datasets, confirming that ignoring them distorts the estimation of stopping latencies. The significance of this work lies in providing a robust tool for quantifying a previously unmeasured component of response inhibition. By distinguishing between slow stopping processes and failures to trigger the stop process, researchers can avoid fictitious group differences in SSRT. The method is implemented in the BEESTS software, making it accessible for future research. This approach improves the accuracy of interpreting stop-signal data, offering clearer insights into the mechanisms of executive function and inhibition deficits in clinical conditions.
Key finding
The proposed Bayesian model accurately estimates the probability of trigger failures and stop-signal reaction time distributions, revealing that ignoring trigger failures leads to biased estimates of stopping latencies.
Methodology
modeling
Provenance
The full processing record for this entry. Every stage of this paper's journey through the pipeline is logged — what ran, with which tool and model, how many attempts it took, and when it last completed. Discovered via author_sweep_intake on 2026-05-28.
| Stage | Outcome | Tool | Model | Prompt | Attempts | Completed |
|---|---|---|---|---|---|---|
| discover | success | author_sweep | — | — | 2 | 2026-05-28 |
| archive | success | canonical_url | — | — | 1 | 2026-06-04 |
| extract | success | cached | — | — | 3 | 2026-06-10 |
| clean | success | clean | — | — | 1 | 2026-06-04 |
| chunk | success | chunk | — | — | 1 | 2026-06-04 |
| embed | success | embed | Qwen/Qwen3-Embedding-8B | — | 1 | 2026-06-04 |
| enrich | success | — | — | — | 1 | 2026-05-28 |
| promote | success | — | — | — | 1 | 2026-06-04 |
| summarize | success | llm | qwen3.6-27b-prismaquant | summ-v5 | 2 | 2026-06-10 |
| tag | success | vector_similarity | — | — | 15 | 2026-06-11 |
| verify | success | — | — | — | 2 | 2026-06-10 |
Summary generated by qwen3.6-27b-prismaquant on 2026-06-10; verification: verified.
Topics
Ranked by relevance to this paper. Hover a topic for its definition.