Can the link between numbers and space be split between two people?
People judged numbers slightly faster when a human partner's arrow rule matched the small-left, large-right number line, but a computer partner produced no such effect.
Source
Number to me, space to you: Joint representation of spatial-numerical associations
Study at a glance
- Design
- Human experiment — Social go/no-go task: the participant responded only to digits smaller or larger than 5 on a single central key, while a partner (a confederate or a computer) was assigned left- or right-facing arrows; rule pairings were congruent or incongruent with the mental number line; partner type varied between subjects
- N
- N=48 · 48 University of Bologna students, 24 paired with a human confederate and 24 paired with a computer
- Population
- Italian university students with normal or corrected vision, naive to the purpose of the study
- Outcome
- Reaction time to go trials for digits under rule pairings that were congruent (small/left, large/right) versus incongruent (small/right, large/left) with typical spatial-numerical associations
Structured fields used in claim comparison tables when every cited study has a complete layer.
What they did
Each participant sat facing a partner across a table and pressed one central key only for digits smaller (or larger) than 5, while the partner was supposedly responding to left- or right-pointing arrows on another laptop. Across four blocks the pairing of the participant's number rule with the partner's arrow rule was either congruent with the mental number line (small with left, large with right) or incongruent. For half the participants the partner was a same-sex confederate; for the other half the partner was simply a computer. Because the participant's own task had no spatial component at all, any congruency effect had to come from representing the partner's rule.
What they found
With a human partner, participants responded about 11 ms faster in congruent than incongruent blocks, a small but reliable effect supported moderately by a Bayesian analysis. With a computer partner there was no congruency benefit (a non-significant difference of about -5 ms in the opposite direction). The difference between the two partner conditions was significant, while overall speed did not differ between conditions or between small and large numbers.
The limits
What it doesn't show
The human-partner effect is small (Cohen's d of 0.27) and comes from only 24 people per condition, so it needs replication; the authors note it was smaller than in earlier single-person versions. Partner type was manipulated between subjects rather than within, and there was no manipulation check of whether participants believed the partner was acting or saw the computer as an agent. The confederate never actually pressed a key, so the study cannot say which features of a real co-actor (visibility, humanness, believed agency) are critical. Error rates were too low to analyse, so conclusions rest on reaction times alone.
Key terms
- Spatial-numerical association (SNA)
- The tendency to link small numbers with left space and large numbers with right space, often described as a mental number line.
- Go/no-go task
- A task in which people respond to some stimuli and withhold responses to others, here using one central key so responses have no left-right component.
- Co-representation
- Mentally representing a partner's task rule alongside one's own, even when the partner's task does not require coordination.
- Social Simon effect
- A spatial compatibility effect that disappears when one person does half a task alone but reappears when a partner does the other half, suggesting the partner's task is represented.
- Confederate
- A member of the research team who poses as a fellow participant and follows a script.
- Congruency effect
- The speed advantage for trials or blocks whose features match an expected association compared with those that conflict with it.
Flashcards
0 of 10 answers reviewed
Research intelligence for this paper
See its role on concept claims, tensions it is part of, placement history, and related discoveries.
Quiz yourself
In this study, what did the real participant respond to?
Common questions
Why use a single central key instead of left and right keys?
With lateralised keys, any number-space effect could come from the response hands themselves. A single central key removes all spatial features from the participant's own task, so any effect must come from a conceptual link.
Why include a computer partner?
It tests whether the effect depends on sharing the task with another person. If simply having an arrow task running nearby were enough, the computer condition should also show the effect, but it did not.
Does this prove numbers are inherently spatial?
No. Earlier work suggests the number-space link needs both magnitude and direction to be actively processed; this study shows those two pieces can be spread across two minds, which supports a flexible, task-dependent association rather than a fixed one.
More on Numerical cognition