The puzzle at the heart of the study concerns what psychologists call the alerting-congruency interaction. In everyday laboratory tasks, people respond to targets—say, identifying the color of a word or the direction of a central arrow—while ignoring distracting information flanking or embedded in the display. When the distractor suggests the wrong response, performance slows and errors increase; this is the congruency effect, a cornerstone measure of cognitive control. Since the early 2000s, researchers led by studies such as Callejas and colleagues’ work on the three attentional networks had shown that presenting a sudden alerting cue, like an auditory warning tone shortly before the display, reliably enlarges this congruency effect. The alert makes you faster overall, but paradoxically it also makes distractors more potent saboteurs. One influential interpretation holds that alertness broadens the attentional spotlight or boosts global processing, letting irrelevant information flood in alongside the relevant signal.

Yet a persistent anomaly complicated this tidy picture. The alerting-congruency interaction shows up robustly in the arrow version of the Eriksen flanker task, where participants report the direction of a central arrow flanked by misleading arrows, but it stubbornly fails to appear in Stroop tasks, where people must report the color of a color word whose meaning conflicts with the correct answer. Because arrows carry pre-existing directional associations with left and right responses, while color words in the versions used do not map directly onto motor responses, several theorists proposed that the interaction requires stimulus-response directional associations. In this view, alerting signals amplify only those distractors that can automatically activate a motor response through learned directional links. The claim mattered theoretically because it touched on foundational debates about automaticity, dimensional overlap, and the architecture of stimulus-response translation—a taxonomy famously laid out by Kornblum and colleagues in 1990.

Golden and her team, including Thomas G. Hutcheon of Bard College, Katherine M. Mathis of Bates College, Emily R. Cohen-Shikora of Washington University in St. Louis, and Todd A. Kahan of Bates College as senior author, saw a different possible culprit: the timing of distractor processing. In standard implementations of both the flanker and Stroop tasks, the distractor and target appear simultaneously and remain visible until response. But the two tasks differ subtly in how quickly their distractors activate competing information. Arrows are potent, rapidly processed directional signals whose interference typically peaks early in the reaction-time distribution and then declines as cognitive control suppresses the wrong response. Stroop-type distractors in keypress versions, by contrast, produce interference that builds more gradually across the distribution. If alerting signals accelerate processing overall, they reasoned, then the interaction might simply depend on where in time the distractor’s activation lands relative to the target’s—rather than on whether the distractor has directional meaning.

To test this, the researchers manipulated distractor preview. In the preview condition, the distracting information appeared on the screen alone for a brief interval before the target was added, giving the distractor a head start. In the no-preview condition, distractor and target appeared together, replicating the standard arrangement. Each display was either preceded by an alerting cue or not, and each trial was either congruent or incongruent. Experiment 1 used Stroop stimuli—color words printed in colors—with 158 participants responding via keypress. Experiment 2 used the arrow flanker task with 157 participants. The factorial design yielded 32 trials of each of eight conditions across 256 total trials, with the first six treated as practice and the rest split into five blocks of 50. Crucially, both experiments were conducted online, with subject-level data made publicly available on the Open Science Framework.

The results in the standard, no-preview conditions replicated the literature precisely. Stroop performers showed no significant alerting-congruency interaction—the alert sped people up but did not reliably change how much the conflicting word hurt them. Flanker performers, meanwhile, showed the classic robust interaction: following an alerting cue, the cost of incongruent flanking arrows was magnified. If the field had stopped here, the directional-association hypothesis would have stood unchallenged for another round. But the preview conditions rewrote the story entirely. When Stroop distractors were previewed before the target, a significant alerting-congruency interaction emerged for the first time in this paradigm—alerting now amplified the congruency effect. And when flanker distractors were previewed, the interaction not only vanished but reversed in tendency, with the congruency effect becoming numerically smaller after an alerting cue. Timing, in other words, did what a decade of stimulus-selection arguments could not: it flipped the pattern in both tasks.

The authors supported these conclusions with delta plots, a distributional analysis technique that plots the congruency effect across quantiles of the reaction-time distribution, from the fastest responses to the slowest. Derived from the activation-suppression framework developed by Ridderinkhof and formalized in models by Ulrich, Schröter, Leuthold, and Birngruber, delta plots reveal the time course of automatic distractor activation and its suppression. In the flanker task with simultaneous presentation, congruency effects are typically largest in fast responses and shrink with slower ones—the signature of a distractor whose activation arrives early and is then inhibited. Stroop-type keypress tasks often show the opposite, with interference growing across the distribution. The delta plots in the new experiments confirmed that previewing the distractor shifted this temporal signature in exactly the way the timing account predicts, supporting the conclusion that when distractor activation peaks relative to target processing is the critical variable governing whether alertness helps or hinders conflict resolution.

Several additional details strengthen the interpretation. In the flanker experiment, participants were significantly faster on alerted trials overall, F(1, 116) = 38.18, p < .001, ηp² = .25, a substantial main effect of alerting that counters any argument the interaction should only be recognized when accompanied by such an effect. The team also checked robustness: rather than excluding reaction times by arbitrary cutoffs, they used geometric means to preserve the full distribution, and a supplementary trimming analysis removing values beyond two standard deviations from each participant’s condition mean reproduced the identical pattern of significance in every experiment, including replication sub-experiments labeled 1b, 2a, and 2b. The research formed part of Golden’s undergraduate honors thesis at Bates College, supported by a Bates College Student Research Fund grant, and was approved by the Bates College Institutional Review Board in accordance with APA ethical standards and the Declaration of Helsinki.

What does this mean for theories of attention? First, the results undermine the claim that alerting-congruency interactions require pre-existing stimulus-response directional associations. Stroop stimuli lacking such associations produced the interaction once distractors were given a temporal head start, while arrow stimuli possessing those associations lost the interaction under the same manipulation. Directionality, whatever its other roles, cannot be the deciding factor. Second, the findings breathe new life into temporal-overlap accounts of conflict, echoing classic work by Hommel on the Simon task and more recent electrophysiological and behavioral studies by Mackenzie, Mittelstädt, Ulrich, and Leuthold on the temporal order of relevant and irrelevant dimensions. Alerting signals appear to accelerate the engine of processing; whether this acceleration inflates or deflates measured conflict depends on whether the distractor’s activation curve is ahead of or behind the target’s at the moment responses are selected. A distractor that has already accumulated activation when alertness surges gets amplified; a distractor still ramping up may be caught by the target’s head start and suppressed more effectively.

The practical and methodological implications ripple outward. Task comparison studies that attribute differences between paradigms to stimulus properties may instead be capturing differences in processing dynamics—something researchers such as Pratte and Mittelstädt and colleagues have emphasized in recent distributional work on flanker and Stroop tasks. Experimenters who choose stimulus durations and preview intervals are implicitly choosing a point on the distractor’s activation curve, and the new results suggest this choice can determine whether alertness and control appear coupled or independent. Beyond the laboratory, the work speaks to the broader question of how phasic alertness—the brief surge of arousal produced by warnings, alarms, and sudden events—interacts with selective attention in real-world settings, from cockpit warnings to medical monitor alarms. Whether an alerting signal helps you ignore the noise or makes the noise louder may depend less on what the noise means than on when it started talking.

The authors caution that their studies were not preregistered, and they frame the discussion as a challenge for formal models of the interaction rather than a definitive verdict. Activation-suppression race models, diffusion-based dual-process accounts, and conflict-monitoring theories will each need to accommodate the preview reversal. Still, with subject-level data openly available, the findings offer a concrete empirical anchor. For now, the message is strikingly simple: to understand why being alert sometimes makes distraction worse, watch the clock, not the arrow.

At the heart of the study, the authors investigate the neural processes that underpin feature binding. They propose that our brains do not simply store individual visual features in isolation, but rather, they have the capability to integrate these features into unified perceptual wholes. This process is thought to occur in various stages, suggesting a dynamic interaction between visual input and cognitive processing. By employing advanced experimental designs, the researchers gained insights into the efficiency of feature binding and the potential pitfalls that can arise, leading to errors in perception.

Errors in feature binding, as identified in the study, can manifest in several ways. One common form is the misbinding of features, where elements from different objects become erroneously combined into a single percept. This misattribution raises important questions about the reliability of our cognitive faculties when processing visual stimuli. The researchers demonstrate through a series of controlled experiments that even subtle discrepancies in visual features can lead to significant perceptual errors, offering valuable insights into how the human brain prioritizes and organizes information.

Foerster et al. also emphasize the role of attention in feature binding. The study reveals that attentional focus can greatly influence the likelihood of errors occurring during the binding process. When individuals are distracted or their attention is divided, the risk of misbinding features increases considerably. This finding aligns with previous research suggesting that attention acts as a powerful filter, shaping the way visual information is integrated and ultimately perceived. The study provides empirical evidence that highlights the delicate balance between attentional resources and perceptual accuracy.

Moreover, the research illustrates how context can modulate feature binding. When presented with complex images that contain multiple objects with overlapping features, participants demonstrated variations in their ability to accurately bind those features based on the surrounding context. This phenomenon underscores the importance of the visual environment in shaping perceptual outcomes, suggesting that context not only aids in feature binding but can also lead to confusion and errors when stimuli are not clearly delineated.

The implications of these findings extend beyond the confines of academic interest. Misbinding errors can have practical consequences in various domains, including eyewitness testimony, where individuals may mistakenly combine features from different people or events. This raises critical considerations for legal practices and the reliability of human memory. The insights gained from this study can inform strategies aimed at mitigating errors in high-stakes situations where accurate feature binding is essential.

In addition to exploring misbinding errors, Foerster and colleagues also investigate the neural mechanisms linked to feature binding. Utilizing techniques such as neuroimaging, they highlight specific brain regions involved in integrating visual features. The study suggests that regions implicated in attentional processes play a crucial role in orchestrating feature binding, revealing an intricate relationship between attention and perception at the neural level.

One of the standout contributions of this research lies in its methodological rigor. By employing a diverse array of experimental strategies, including visual search tasks and psychophysical measurements, the authors crafted a robust investigation that lends credibility to their conclusions. This rigorous approach not only enhances the academic value of the study but also sets a benchmark for future research in the field of cognitive psychology.

While the findings of this study are compelling, it also opens up avenues for further inquiry. Future research could expand on the relationship between feature binding and various cognitive processes, such as memory and decision-making. Additionally, investigating individual differences in feature binding could provide deeper insights into why some individuals experience more frequent misbinding errors than others.

As this research makes its way into broader discussions within the field, the potential for interdisciplinary applications emerges. For instance, the principles underlying feature binding could be leveraged in the development of artificial intelligence and machine learning algorithms designed to enhance image recognition and processing capabilities. By emulating the cognitive strategies humans use to bind features, these technologies could achieve greater accuracy and reliability in their outputs.

The study ultimately underscores the remarkable complexity of the human visual system and its inherent vulnerabilities. Understanding the mechanisms that drive feature binding and the conditions that lead to errors not only enriches our knowledge of cognitive processes but also highlights the nuanced interplay between perception, attention, and context. As researchers continue to peel back the layers of cognitive psychology, studies like “Feature binding and error commission” provide essential insights that improve our understanding of how we interact with and interpret the world around us.

In conclusion, Foerster et al.’s study is a significant step forward in the exploration of feature binding and its implications for perception. The clear demonstration of how attentional resources and contextual cues can influence binding errors contributes to a deeper understanding of cognitive function. Moreover, the research serves as a call to action for ongoing investigations into the complexities of perception and the myriad factors that shape our experiences of reality.

As we look to the future, the findings presented in this research establish a foundation for advancing both theoretical and practical applications in cognitive science. By continuing to unravel the intricacies of feature binding and the potential for errors, researchers are poised to enhance not only our understanding of human perception but also the technology we develop to aid and augment our cognitive capacities.

Subject of Research: Feature binding and misbinding errors in perception.

Article Title: Feature binding and error commission.

Article References: Foerster, A., Linz, S., Moeller, B. et al. Feature binding and error commission. Atten Percept Psychophys 88, 50 (2026). https://doi.org/10.3758/s13414-025-03164-w

Image Credits: AI Generated

DOI: https://doi.org/10.3758/s13414-025-03164-w

Keywords: Feature binding, cognitive psychology, perception, attentional resources, visual stimuli, misbinding errors, contextual influence, neural mechanisms, cognitive function.