Skip to content
PaperFren

Do flanker and Simon tasks really measure self-control?

Open paper intelligence

Scores on popular conflict tasks were barely predicted by skills thought to train inhibition and did not relate at all to people's self-rated self-control or impulsivity.

Source

Interference scores have inadequate concurrent and convergent validity: Should we stop using the flanker, Simon, and spatial Stroop tasks?

Paap KR, Anders-Jefferson R, Zimiga B, et al. · Cognitive research: principles and implications · 2020

doi.org/10.1186/s41235-020-0207-yRead the full paper ↗41 citationscc by

Study at a glance

Design
Cross-sectional — Single-session individual-differences study: four arrow-based interference tasks, Raven's matrices and questionnaires, analysed with stepwise, residual and LASSO regressions and Bayes factors
N
N=201 · 201 San Francisco State University undergraduates retained from 213 tested (9 incomplete, 3 excluded for extreme accuracy or speed); a matched sub-analysis paired all 52 men with 52 women of identical Raven's score
Population
US university undergraduates (149 women, 52 men)
Outcome
Reaction-time interference scores (incongruent minus congruent) on flanker, Simon, spatial Stroop and vertical Stroop tasks, and a composite of the three tasks that clustered together

Structured fields used in claim comparison tables when every cited study has a complete layer.

What they did

Students completed four quick arrow tasks in which the correct response sometimes conflicted with irrelevant information such as where the arrow appeared or which way neighbouring arrows pointed. The slowdown on conflicting trials (the interference score) is widely treated as a measure of inhibitory control. Participants also did a short Raven's reasoning test and reported their bilingualism, music training, video gaming, meditation, exercise, sports ability, income and other background details, plus standard impulsivity and self-control questionnaires. The researchers tested which of these predicted interference scores using several regression methods and Bayes factors.

What they found

Only two predictors held up consistently: higher Raven's scores and being male went with smaller interference, together explaining about 18.4% of the variance in the composite score. The male advantage stayed the same size when men were matched to women with identical Raven's scores. Bilingualism, music training, meditation and exercise showed no meaningful link, with Bayes factors favouring no effect; video gaming correlated weakly but dropped out once sex and reasoning were included. None of the correlations between interference scores and self-reported impulsivity or self-control was significant.

The limits

What it doesn't show

All participants were students from one university, so the lack of effects for things like income or exercise may not hold in children or broader populations. Experiences such as meditation and gaming were measured with single self-report items, and self-control questionnaires rely on memory and self-image. The tasks shared very similar stimuli and responses, and only 25% of trials were incongruent, which could shape what the scores reflect. Because the design is correlational, it cannot show that training in any activity fails to change inhibition; it only shows that natural variation in these activities does not predict interference scores here.

Key terms

Interference score
Mean reaction time on incongruent (conflicting) trials minus mean reaction time on congruent trials; smaller scores are often taken to mean better conflict control.
Flanker task
A task where people respond to the direction of a central arrow while ignoring flanking arrows that may point the same or opposite way.
Simon task
A task where the side on which a stimulus appears conflicts with the response its identity requires, creating stimulus-response conflict.
Inhibitory control
The ability to suppress irrelevant information or unwanted responses; one proposed component of executive function.
Convergent validity
The extent to which a measure correlates with other measures that are supposed to tap the same construct.
Fluid intelligence
The ability to reason and solve novel problems without relying on learned knowledge, here measured with Raven's matrices.

Flashcards

1 / 11

0 of 11 answers reviewed

Research intelligence for this paper

See its role on concept claims, tensions it is part of, placement history, and related discoveries.

Open paper intelligence

Quiz yourself

1 / 6

What is an interference score in the flanker or Simon task?

Common questions

If interference scores don't track self-control, what might they measure?

The author suggests they may reflect task-specific ways of resolving a particular kind of conflict in fast reaction-time tasks, partly linked to general reasoning ability, rather than a broad trait that governs everyday self-control.

Does this prove bilingualism or meditation have no cognitive benefit?

No. It shows that in this sample natural variation in these activities did not predict interference scores; training studies and other outcomes could still show effects.

Why were the reliabilities of the tasks important?

A measure with low reliability cannot correlate strongly with anything. The flanker score was highly reliable, but the vertical Stroop and Simon scores were less so, which limits how strongly they could relate to other measures.

More on Executive function