The Stroop Effect: Why Conflicting Information Slows Responses

Author: Angela Bogdanova · Published: October 16, 2026 · Editorial Policy

Research cluster: Attention, Perception & Cognitive Control · Written by Artificial Sapiens

What Is the Stroop Effect?

The Stroop effect is a well-established finding in cognitive psychology: people generally take longer to name the ink color of a printed word when the word names a different color than when the word and ink color agree. Imagine the word “BLUE” printed in red ink. The task is to say “red,” but the familiar word also activates information about “blue.” The resulting delay reveals how task-relevant and task-irrelevant information can compete during perception, selection, and action. The phenomenon is named for John Ridley Stroop, whose 1935 experiments provided a particularly influential demonstration of the asymmetry between reading words and naming colors. Stroop (1935); MacLeod (1991).

A modern definition describes the Stroop effect as a difference in behavioral performance between conditions that vary in the correspondence of competing stimulus dimensions. In the classic color–word task, the relevant dimension is ink color and the irrelevant dimension is word meaning. Researchers most often examine differences in reaction time, error frequency, or both. The reliably observed group-level effect makes the task a central experimental tool for studying selective attention and cognitive control. The precise psychological mechanism behind a particular difference score depends on the experimental design and remains an active research question. Parris et al. (2022); Viviani et al. (2024).

The word “interference” refers to a cost relative to a suitable comparison condition. It does not mean that the reader completely loses control of their response, and a delay of a few milliseconds has no diagnostic meaning by itself. Much of the scientific value of the Stroop paradigm comes from comparing tightly controlled conditions across many trials and participants, rather than interpreting a person's single attempt. Scarpina and Tagini (2017); Hedge et al. (2018).

How the Classic Color–Word Stroop Task Works

A participant sees a color word and is instructed to report the color of its letters while disregarding the word's meaning. The instruction matters. If the person is asked to read the word instead, reading and color naming exchange their roles as relevant and irrelevant tasks; the classic asymmetry is that irrelevant word meaning interferes much more with color naming than irrelevant ink color does with word reading. This asymmetry was already evident in Stroop's original work and has shaped theories of practiced and automatic processing. Stroop (1935); MacLeod (1991).

Congruent, incongruent, and neutral trials

In a congruent trial, a color word matches its ink: “RED” in red ink. In an incongruent trial, the word and ink conflict: “RED” in green ink. A neutral comparison may involve a string of noncolor letters or symbols printed in a target color, a noncolor word, or another baseline selected for the research question. Participants respond either aloud or using assigned keys. The experiment records latency, errors, and sometimes additional behavioral or physiological measures. Neutral conditions require deliberate construction because pronounceability, word frequency, semantic content, visual form, and response associations may differ. Littman et al. (2019); Parris et al. (2022).

Three comparisons answer related but distinct questions. The incongruent-minus-congruent difference is commonly called the overall congruency or Stroop effect. The incongruent-minus-neutral difference estimates an interference cost relative to that particular neutral baseline. The neutral-minus-congruent difference estimates facilitation for matching information. These quantities are sometimes conflated, yet an overall congruency difference combines effects that can occur in both conditions. For example, a faster congruent response can enlarge the overall contrast even without a corresponding increase in incongruent slowing. MacLeod (1991); Parris et al. (2022).

A worked numerical example

Suppose a hypothetical experiment produces mean correct-response times of 610 milliseconds for incongruent trials, 530 milliseconds for neutral trials, and 495 milliseconds for congruent trials. The overall congruency effect is 115 milliseconds (610 − 495). Relative to neutral trials, interference is 80 milliseconds (610 − 530), while facilitation is 35 milliseconds (530 − 495). These are invented numbers used only to illustrate the computations; they are not pooled estimates or normative clinical thresholds. A sound analysis would also inspect the number of valid trials, errors, participant-level variability, response-time distributions, and the uncertainty around each estimate.

Different published studies report different magnitudes because they differ in stimulus sets, response modality, timing, practice, sample composition, trial proportions, and how conditions are balanced. The scientific claim supported across many experiments is the recurring pattern of conflict-dependent processing costs, rather than one universal number that applies to every individual or laboratory. MacLeod (1991); Viviani et al. (2024).

Why Conflicting Information Slows Responses

The classic intuitive explanation begins with extensive reading experience. For a literate adult, written words strongly elicit linguistic processing even when the immediate goal concerns the ink color. Naming an arbitrary color usually needs a more specific selection of the task-relevant visual feature and its response. When the printed word names another color, multiple representations or responses become active and the system must settle on the response appropriate to the instruction. This account captures the central everyday experience, but “reading is automatic” is an initial theoretical explanation, not a complete specification of where every source of the effect originates. Cohen et al. (1990); MacLeod (1991).

A useful approach separates several processing questions. Is attention drawn to the word instead of the ink? Does the word activate semantic knowledge about a color that competes with the correct answer? Are incompatible motor or verbal responses activated? Does recognizing a word trigger the alternative task of reading it? Does the person adapt task settings before or after conflict appears? Research indicates that these sources can interact; the same behavioral delay can arise through more than one processing pathway. Banich (2019); Parris et al. (2022).

Automaticity as a matter of degree

Learned behaviors can become efficient and readily triggered by familiar stimuli. Reading is an especially powerful example for experienced readers, whereas a competing instruction asks the participant to give priority to ink color. Yet attention, expectations, strategy, and task context can alter the impact of word meaning. Automaticity therefore describes aspects of learned processing under particular conditions rather than an all-or-nothing claim that words must always be processed to an identical depth. Parallel distributed-processing models explain this graded competition using learned differences in processing strength and task-dependent inputs. Cohen et al. (1990).

The distinction matters when interpreting successful performance. An individual who eventually names the ink correctly may still show interference in response time. Conversely, a fast response does not prove that the irrelevant word was entirely unprocessed. Reaction time reflects the combined consequence of multiple operations and the threshold for making a response. Modeling, carefully matched controls, and experiments that manipulate one source of competition at a time help clarify what the observable result means. Parris et al. (2022).

Information conflict and task conflict

Information conflict concerns incompatibility between stimulus meanings, representations, or responses: the word says “green,” the correct ink-color response is “red.” Task conflict concerns a different competition: the stimulus evokes word reading while the instruction requires color naming. Importantly, a congruent color word can still encourage the irrelevant task of reading, even though its meaning is compatible with the ink color. Reviews argue that task conflict is distinguishable from broader information conflict, while attempts to isolate semantic, phonological, and response components of information conflict remain methodologically difficult. Littman et al. (2019); Parris et al. (2022).

This explains why the choice of “neutral” stimuli changes the interpretation. A row of colored symbols may remove the tendency to read a meaningful word; a meaningful noncolor word may still evoke reading, alter perceptual processing, or create response competition. Even conditions labeled neutral can recruit different psychological processes. A researcher who reports a single interference score must therefore specify the comparison condition and the inferential claim it supports. Parris et al. (2022).

From John Ridley Stroop to Modern Cognitive Psychology

John Ridley Stroop published “Studies of Interference in Serial Verbal Reactions” in 1935. His experiments compared reading and color naming under competing stimulus conditions. In the classic finding, incompatible printed words substantially slowed naming the colors, whereas incompatible ink colors produced relatively little interference with reading. The results established a striking experimental asymmetry between two familiar forms of processing and demonstrated a tractable method for probing selective performance under conflict. Stroop (1935).

Subsequent decades transformed the paradigm into a family of experiments on attention, reading, automaticity, response competition, and executive control. Colin MacLeod's integrative review organized the considerable pre-1991 literature and argued for theories that accommodate parallel processing of relevant and irrelevant dimensions rather than attributing all interference to a single bottleneck. The literature later developed computational accounts, trial-by-trial control theories, and approaches emphasizing learning, stimulus contingencies, and multiple loci of attentional selection. MacLeod (1991); Cohen et al. (1990); Parris et al. (2022).

The name “Stroop effect” now appears in research on other kinds of interference, including spatial correspondence and affectively salient material. These variants share a family resemblance with the color–word experiment but differ in stimuli, tasks, and theoretical assumptions. Their outcomes should be described as variant-specific unless evidence supports a common explanatory mechanism. A recent methodological analysis recommends greater precision about dimensional overlap and even considers spatial versions for isolating some processes. Viviani et al. (2024).

Major Scientific Models of Stroop Interference

Parallel processing and graded pathways

In a parallel distributed-processing account, color and word information travel through partially distinct pathways whose outputs can influence response selection. Extensive experience strengthens word-reading pathways, while the experimental instruction biases processing toward color naming. Incongruent stimuli create competing activation patterns, delaying selection of the task-relevant response. This model treats practice, attention, and control as interacting features of a learning system. It also accounts for why merely instructing a person to ignore a word may leave measurable interference. Cohen et al. (1990).

Such models are valuable because they generate quantitative predictions rather than relying only on metaphors such as “two thoughts fighting.” Their limitations depend on how learning histories, task settings, stimulus frequency, and performance measures are represented. Successful simulation of a behavioral pattern supports the model's adequacy for that pattern; it does not establish that the brain implements precisely the same internal computation.

Conflict monitoring and adaptive control

Conflict-monitoring theory proposes that the cognitive system tracks competition among processing or response alternatives and uses that information to adjust control. In experiments, adaptation can be studied by testing whether recent conflict predicts subsequent behavior. The framework has influenced research on cognitive control and associated neural activity, particularly research involving medial frontal and lateral prefrontal networks. Botvinick et al. (2001); Banich (2019).

The critical inference is more demanding than noticing that performance improves after an incongruent trial. An apparent adjustment may reflect learned stimulus–response contingencies, feature repetitions, expectancy, or changes in stimulus composition. Consequently, researchers must design conditions that distinguish control adaptation from item-specific learning before concluding that a difference is caused by conflict monitoring. Schmidt and Besner (2008); Viviani et al. (2024).

Multiple loci and competing interpretations

A multiple-loci perspective considers interference or facilitation at stages involving the stimulus, task representation, lexical or semantic activation, response selection, and execution. Neurocognitive reviews support the possibility of several interacting levels of control. At the same time, a critical analysis of Stroop methods cautions that many manipulations claimed to separate semantic from response conflict fail to isolate one component cleanly. This is a substantive ongoing theoretical dispute, not a reason to doubt the basic experimental effect. Banich (2019); Parris et al. (2022).

An alternative perspective gives stronger weight to input-driven attention and perceptual properties of the stimuli, questioning whether every finding routinely assigned to top-down control requires that explanation. This debate illustrates the value of designs that manipulate both stimulus characteristics and control demands and of keeping theoretical mechanisms separate from the directly observed difference between conditions. Parris et al. (2022); Viviani et al. (2024).

How Researchers Measure the Stroop Effect

A typical computerized study presents dozens or hundreds of randomized trials, records each correct response time in milliseconds, and logs errors. Some studies use vocal naming, recorded with a microphone or voice key; others assign color labels to keyboard keys. The mapping changes the task: a vocal response recruits speech production while a keypress recruits learned stimulus-to-key associations. The two formats are related experimental paradigms, not perfectly interchangeable measures of one underlying faculty. Scarpina and Tagini (2017); Parris et al. (2022).

Investigators generally plan response mappings and stimulus probabilities in advance, exclude technical failures and invalid trials using defensible criteria, and analyze error rates alongside speed. A reaction-time difference without an accompanying accuracy analysis is potentially misleading: participants can trade speed for correctness. Researchers may compare average or median response times, model the full distribution, or fit hierarchical models to trial-level performance. Pre-registered analytic choices and sufficiently many trials improve interpretability, particularly when interested in small differences between conditions. Hedge et al. (2018); Viviani et al. (2024).

What a Stroop interference score actually tells us

The usual difference score measures how response performance changes between designated conditions. It does not directly measure the quantity of attention in a person's brain, uniquely locate a mental “inhibition center,” or separate every stage of processing. Several mechanisms may change the same score. Researchers distinguish an experimental effect that is robust across a group from a measurement that reliably ranks individuals across different days. Those are separate psychometric questions. Hedge et al. (2018).

A particularly important finding is the reliability paradox. Tasks can produce strong and replicable average experimental effects while offering relatively weak test–retest reliability for individual-difference scores. Hedge and colleagues examined several classic conflict tasks, including Stroop, and showed why a robust experimental contrast need not provide a stable personal ranking. When a study seeks to relate interference scores to personality, symptoms, or neural variables, measurement reliability and the structure of the sample therefore become central. Hedge et al. (2018).

Validity depends on the research question

Construct validity concerns whether the measure supports the interpretation attached to it. A well-controlled color–word experiment has strong experimental value for demonstrating interference under particular conditions. Using the same latency difference to assert a broad, stable ability called “self-control” requires additional validation. Norms and calculation methods also vary among clinical-style Stroop Color and Word Test versions, making scoring details central to comparisons. The systematic review by Scarpina and Tagini specifically examined these scoring and normative issues. Scarpina and Tagini (2017).

What Changes Stroop Performance?

Practice, expectancy, and context

Repeated experience with a particular stimulus set can alter speed and accuracy. Participants learn response mappings, become familiar with the procedure, and may discover that some words predict certain colors. An apparent improvement may therefore combine general task practice, particular stimulus–response learning, and shifts in control. Such mechanisms are of scientific interest, but improvement at this task is not automatically evidence for a transferable increase in everyday concentration. Cohen et al. (1990); Schmidt and Besner (2008).

The proportion of congruent trials matters because it can change expectations and the statistical relationships among words, colors, and correct responses. An item-specific proportion-congruent effect can sometimes be explained by learning what answer a given word usually predicts, rather than by voluntarily shifting attention each time that word appears. This result is a reminder that stimulus probabilities are part of the mechanism under study, not a minor detail of experiment administration. Schmidt and Besner (2008).

Language experience and individual differences

Reading proficiency, linguistic context, color naming, age, and familiarity with the response procedure can influence results. A task developed around printed color words also makes assumptions about literacy and language processing. Differences across languages or groups may arise from several interacting features of reading and experimental design; they should not be reduced to a universal hierarchy of cognitive ability. A qualitative systematic review published in 2026 addresses the heterogeneous research on individual differences among healthy adults, reinforcing the need to inspect methods before attributing variation to a single personal trait. Vasta et al. (2026); Scarpina and Tagini (2017).

Conditions that affect vision, color discrimination, speech output, motor speed, fatigue, or familiarity with computers may change performance through pathways outside the central conflict mechanism of interest. When populations differ on these characteristics, researchers should use appropriate controls and avoid interpreting raw response latency as a direct measure of inhibition. For clinical assessment, standardized procedures and suitable comparison data matter at least as much as the named task. Scarpina and Tagini (2017).

The Stroop Effect, the Brain, and Cognitive Control

The Stroop task is frequently used in cognitive neuroscience because conflict-dependent changes in behavior can be paired with brain measures. Research discusses the contributions of prefrontal and medial frontal areas, including systems involved in representing task goals, detecting competition, and selecting responses. Reviews suggest that control can operate at multiple stages, with different networks contributing under different task conditions. These studies provide convergent clues about mechanisms; they do not establish one exclusive “Stroop region” responsible for every outcome. Banich (2019); Botvinick et al. (2001).

The language of brain “activation” also needs care. An association between a difficult condition and increased activity in a region cannot, by itself, prove the psychological computation performed there. Brain imaging, direct electrophysiology, perturbation studies, computational models, and behavioral dissociations offer different kinds of evidence. The best-supported explanation combines these approaches while acknowledging that functional roles often overlap and may change with task design. Banich (2019).

Within the encyclopedia, the Stroop effect occupies an experimental research node beneath the broader science of inhibitory control. It informs theories of goal-directed attention and response competition; it does not exhaust the concept of executive functioning. The more extensive accounts of executive function, task switching, and response stopping involve partly distinct constructs and experimental paradigms.

Does the Stroop Effect Measure Intelligence, ADHD, or Mental Health?

A Stroop task can contribute information about performance under one kind of cognitive conflict. It does not function as a stand-alone test of intelligence, a diagnosis of ADHD, or a general clinical verdict about executive dysfunction. A psychological construct, a task score, a reported symptom, and a diagnosed disorder belong to different levels of evidence. Clinical interpretation requires a suitable assessment question, standardized procedures, appropriate norms, functional history, and professional evaluation. Scarpina and Tagini (2017); Hedge et al. (2018).

Differences in interference scores may coexist with differences in attention, language, processing speed, or other cognitive processes, but such associations do not isolate their cause. Researchers must distinguish mean differences between groups from the capacity of an individual score to classify a particular person. A result can be significant in an experimental sample while having limited discriminative value for real clinical decisions. It is particularly important to avoid turning a few slow responses or an online Stroop demonstration into a disease label. Hedge et al. (2018); Vasta et al. (2026).

Formal neuropsychological batteries sometimes include a standardized version of the Stroop Color and Word Test because performance can add information to a broader profile. Its meaning depends on the version, scoring algorithm, normative reference group, and clinical context. The experimental color–word effect and the use of a named clinical instrument must therefore be clearly distinguished. Scarpina and Tagini (2017).

Does Repeating the Stroop Task Improve Attention?

Practice can improve familiarity with task rules, response mappings, and stimulus sets. That is a credible and testable near-task outcome. A broader promise—that doing online color-word games improves general attention, professional performance, or mental health—requires separate randomized evidence that improvement transfers to meaningful activities outside the practiced task. The Stroop effect itself establishes neither that a particular training program works nor that transfer is durable. Its strongest demonstrated value remains as a research paradigm for studying interference, learned associations, and cognitive control. Hedge et al. (2018); Viviani et al. (2024).

The most useful practical lesson is conceptual. An instruction to focus on one dimension of information does not guarantee that irrelevant but highly practiced information ceases to influence processing. That fact is relevant to understanding competing information in demanding environments, yet real workplaces, classrooms, digital interfaces, and emotionally meaningful decisions add variables absent from the classic laboratory setup. Applications should be evaluated independently instead of presented as direct consequences of a color-naming experiment.

Methodological Challenges and Open Questions

One major challenge is confounding: studies may change more than the intended variable when they manipulate stimulus congruency or trial proportions. Words and colors can differ in frequency, response mapping, readability, perceptual salience, or association history. The resulting patterns can support more than one explanation unless researchers build matched baselines and tests that discriminate among theories. Recent methodological analysis emphasizes the importance of separating the effects of stimulus dimensional overlap from broader claims about cognitive control. Viviani et al. (2024).

A second challenge is the exact locus of interference. The existence of information conflict is clear, and the distinction between task conflict and information conflict has substantial support. Whether a particular experiment has independently isolated semantic, phonological, response, and execution-related interference is more contentious. The critical review by Parris and colleagues explains why published manipulations often cannot uniquely identify the proposed level, even when the observed behavioral effect is reproducible. Parris et al. (2022).

A third challenge concerns the difference between experimental reliability and individual-difference measurement. Stable group-level congruency effects do not guarantee stable individual ranks, and high internal consistency in one condition does not automatically produce a precise subtraction score. Study design, trial quantity, individual variability, and the intended inference determine the relevant reliability requirements. Hedge et al. (2018).

Finally, ecological validity remains a substantive question. The standard task compresses a complex control problem into a carefully controlled response to isolated stimuli. This is its experimental strength. How well specific Stroop parameters predict sustained everyday performance, response to interruptions, or clinical function must be investigated through separately validated measures and longitudinal or applied studies. Scarpina and Tagini (2017); Vasta et al. (2026).

Frequently Asked Questions

Why is it harder to say the ink color than to read the word?

Experienced readers have strongly practiced word-processing pathways. When the task instead requires naming ink color, word information can intrude on the response process. Incompatible word and color information may slow selection, while the word may also activate the irrelevant task of reading. Several sources can contribute simultaneously, so the broad explanation is stronger than any one-mechanism story. Cohen et al. (1990); Littman et al. (2019).

Is the Stroop effect the same as distraction?

The effect is a particular experimentally measured pattern of interference between stimulus dimensions. It helps researchers understand aspects of distraction, selection, and competing responses, but everyday distraction includes broader processes such as interruptions, motivation, environmental salience, and fatigue. Calling everything distracting a “Stroop effect” loses the precision that makes the paradigm useful.

Does everyone show the same size Stroop effect?

No single difference score applies universally. Average interference is a robust finding under many standard conditions, but the magnitude varies with procedures and participants. Some individuals or trials show weak, absent, or atypical differences for methodological or substantive reasons. The interpretation of stable individual differences also requires specific reliability evidence. Hedge et al. (2018); Vasta et al. (2026).

Can an online Stroop test diagnose ADHD or another disorder?

A casual online exercise cannot establish a diagnosis. Even a standardized test contributes one piece of information within a wider evaluation. Performance can reflect language, sensory function, response speed, practice, situational state, and task-specific interference. Diagnosis depends on established clinical criteria and the person's functioning, not one game or one laboratory contrast. Scarpina and Tagini (2017).

What is the difference between Stroop interference and facilitation?

Interference compares incongruent performance with an appropriate neutral baseline; facilitation compares neutral with congruent performance. A simple incongruent-minus-congruent difference includes both contributions and may be affected by the type of neutral condition, response method, and stimulus composition. Researchers should report the actual contrast used. Parris et al. (2022).

What does the Stroop effect reveal about cognitive control?

It shows that relevant instructions can guide performance while competing stimulus information continues to influence response selection. By varying the task carefully, researchers can test hypotheses about learned automaticity, attention, conflicts between task sets, and adaptive control. The effect alone does not uniquely decide between every theory; its explanatory power comes from experimental contrasts designed to separate competing mechanisms. Banich (2019); Viviani et al. (2024).

Scientific Synthesis

The Stroop effect is one of cognitive psychology's clearest demonstrations that a goal-directed response unfolds in the presence of competing, learned information. Its defining observation is robust across an extensive experimental tradition: incompatible color-word information typically imposes a measurable processing cost. What produces that cost in any specific task can involve graded reading automaticity, information conflict, task conflict, response competition, adaptation, perceptual properties, and learned contingencies. Experimental measurements become most informative when the comparison condition, response modality, stimulus statistics, and intended construct are explicit. MacLeod (1991); Parris et al. (2022); Viviani et al. (2024).

For the science of attention and cognitive control, the lasting achievement of the Stroop paradigm is methodological: it converts a question about competing information into a testable pattern of behavior, then invites more precise theories and experiments to explain it. Its importance rests in that continuing program of measurement and explanation, with clear limits on clinical diagnosis, individual ranking, and claims of transferable training benefits.

Related Articles

For the role of interference in everyday digital environments, see Task Switching and Digital Multitasking: Why Constant Switching Drains Focus. For the separate question of whether online media change attention span, see Social Media and Attention Span: Does Scrolling Really Shorten Your Attention?. These articles address adjacent digital-life questions; the present chapter concerns the foundational color–word experimental effect.

Inhibitory Control: How the Mind Suppresses Competing Responses — Explains stopping responses, resisting interference, and the limits of inhibition measures.

Task Switching: What Switching Costs Reveal About Mental Control — Explains laboratory task switching, switch costs, and control of action rules.

References

Banich, M. T. (2019). The Stroop effect occurs at multiple points along a cascade of control: Evidence from cognitive neuroscience approaches. Frontiers in Psychology, 10, 2164. https://doi.org/10.3389/fpsyg.2019.02164.

Botvinick, M. M., Braver, T. S., Barch, D. M., Carter, C. S., & Cohen, J. D. (2001). Conflict monitoring and cognitive control. Psychological Review, 108(3), 624–652. https://doi.org/10.1037/0033-295X.108.3.624.

Cohen, J. D., Dunbar, K., & McClelland, J. L. (1990). On the control of automatic processes: A parallel distributed processing account of the Stroop effect. Psychological Review, 97(3), 332–361. https://doi.org/10.1037/0033-295X.97.3.332.

Hedge, C., Powell, G., & Sumner, P. (2018). The reliability paradox: Why robust cognitive tasks do not produce reliable individual differences. Behavior Research Methods, 50, 1166–1186. https://doi.org/10.3758/s13428-017-0935-1.

Littman, R., Keha, E., & Kalanthroff, E. (2019). Task conflict and task control: A mini-review. Frontiers in Psychology, 10, 1598. https://doi.org/10.3389/fpsyg.2019.01598.

MacLeod, C. M. (1991). Half a century of research on the Stroop effect: An integrative review. Psychological Bulletin, 109(2), 163–203. https://doi.org/10.1037/0033-2909.109.2.163.

Parris, B. A., Hasshim, N., Wadsley, M., Augustinova, M., & Ferrand, L. (2022). The loci of Stroop effects: A critical review of methods and evidence for levels of processing contributing to color-word Stroop effects and the implications for the loci of attentional selection. Psychological Research, 86, 1029–1053. https://doi.org/10.1007/s00426-021-01554-x.

Scarpina, F., & Tagini, S. (2017). The Stroop Color and Word Test. Frontiers in Psychology, 8, 557. https://doi.org/10.3389/fpsyg.2017.00557.

Schmidt, J. R., & Besner, D. (2008). The Stroop effect: Why proportion congruent has nothing to do with congruency and everything to do with contingency. Journal of Experimental Psychology: Learning, Memory, and Cognition, 34(3), 514–523. https://doi.org/10.1037/0278-7393.34.3.514.

Stroop, J. R. (1935). Studies of interference in serial verbal reactions. Journal of Experimental Psychology, 18(6), 643–662. https://doi.org/10.1037/h0054651.

Vasta, N., Mulatti, C., & Treccani, B. (2026). A qualitative systematic review of individual differences in Stroop task performance among healthy adults. Psychological Research, 90, Article 14. https://doi.org/10.1007/s00426-025-02224-y.

Viviani, G., Visalli, A., Montefinese, M., Vallesi, A., & Ambrosini, E. (2024). The Stroop legacy: A cautionary tale on methodological issues and a proposed spatial solution. Behavior Research Methods, 56, 4758–4785. https://doi.org/10.3758/s13428-023-02215-0.