See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/360081453

Breaking boundaries: How to exploit congruency cues conveyed by the

previous trial

Poster · April 2022

CITATIONS READS

0 44

3 authors:

Luis Jiménez David Gallego

University of Santiago de Compostela University of Santiago de Compostela
60 PUBLICATIONS 2,125 CITATIONS 5 PUBLICATIONS 2 CITATIONS

SEE PROFILE SEE PROFILE

Castor Méndez Paz

University of Santiago de Compostela
57 PUBLICATIONS 641 CITATIONS

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Conflict-triggered learning View project

Deconstructing cognitive control: An analysis of its learning sources View project

All content following this page was uploaded by Luis Jiménez on 29 April 2022.

The user has requested enhancement of the downloaded file.

 How to exploit congruency cues conveyed by the previous trial
Luis Jiménez, David Gallego and Cástor Méndez
University of Santiago de Compostela (Spain)

• The purpose of cognitive control is to keep behavior in line with goals, even in the presence of more compelling and automatic tendencies.
• The efficiency of cognitive control is assessed through interference tasks, in which responses activated by irrelevant features interfere with the imperative ones.
• The amount of interference changes dynamically with practice, showing that cognitive control is continuously adapted to experience.
• However, it is hard to use predictive cues conveyed by the previous trial to prepare for the congruency of the upcoming trial (see Jiménez, et al., 2021).

Shenhav et al.’s (2013) Expected Value of Control

The specification of cognitive control involves a strategical decision
that depends on its expected value:
Jiménez et al. (2021) if Costs outweigh the expected Outcomes, regulation will not be
It Is Harder Than You Think: exerted.
On the Boundary Conditions of Exploiting Congruency Cues
Cueing was only useful with:
Cues presented between trials, but not conveyed by the previous trial. The use of control cues also depends on a balance between costs and perceived outcomes:
Blocked but not trial by trial cueing (i.e. comparing cued vs. control blocks).
Vocal but not Manual Stroop task. • If the task is more automatic (vocal rather than manual) there are more resources available to engage in cueing tasks.
Long (2250 ms) interstimulus Intervals. Experiment 7 • If cue value is constant (e.g. blocked cueing) participants are continuously motivated to use the cue.
• If there is time to release from past efforts (long ISIs) it could be easier to get involved in a new cueing effort.

vs. Why people fail to use the previous trial as a cue, but use a cue interspersed between successive trials?
Hypothesis:
Using the previous trial as a cue blends several control requirements:
Experiment 8
• Dealing with the potential conflict over the current trial.
• Processing the cue AND preparing for the conflict expected on the upcoming trial.

If this blend is in the origin of the observed boundary conditions, then removing the conflict from
the predictive trials (and perhaps facilitating the processing of the cue by adding a semantic
prompt) should restore people’s ability to use the cue provided on the previous trial.

Remove conflict from the predictive trials AND replace it with a semantic cue
Based on Experiment 8 from Jiménez et al. (2021), we removed the Stroop words from the predictive (odd) trials and replaced them by the words “difficult” or “easy” to indicate the congruency of the next trial.
Previous colors were still predictive, but participants were only informed about the semantic cues.
RT ms
Procedure:
800 Results
Three cueing blocks compared with three 700 • Participants took advantage of the cue when it predicted
control blocks: an easy (congruent) successor but not when it predicted a
600
difficult (incongruent) successor.
Three explicit cueing blocks: 500
The words “easy” and “difficult” predicted
CONTROL INCONGRUENT
CUED INCONGRUENT
CONTROL CONGRUENT
CUED CONGRUENT
• Responding was slower on trials containing cues.
the congruency of the successor 400
1 2 3
Hypothesis:
Three control blocks: CUEING COST
Two neutral words (“car”, “house”) replaced 700 If preparing for conflict is costly, perhaps presenting a large
the cueing words. 650
proportion of incongruent trials could make the cost too high.
600
550
Cueing incongruence could be better observed when
500 incongruent trials are scarce.
difficult easy control
words

Remove semantic cues BUT maintain the congruency

Reduce costs by decreasing the rate of conflict
predictive colors AND the use of non-conflict cueing trials.
Procedure: RT ms Results Procedure:
Use a high proportion of congruent trials 800
Replication of Experiment 2 with non-conflict (XXXXX) cues and congruency predictive colors.
700 • Cueing was found Results
600 on both congruent RT ms
• Cueing was found
800
and incongruent on both congruent
500
80% CONTROL INCONGRUENT
CUED INCONGRUENT
CONTROL CONGRUENT
CUED CONGRUENT successors. 700
and incongruent
Cueing

400
1 2 3 80% 600
successors.
Cueing

20% CUEING COST

• Responding was 500

slower on cueing
CONTROL INCONGRUENT CONTROL CONGRUENT
700
• Responding was
CUED INCONGRUENT CUED CONGRUENT
400
650
trials, especially 20% 1 2 3
600 faster on cueing
Control

550 on those CUEING COST (non word) trials,

500
difficult easy control predicting conflict. 700
but less so on trials
Control

650
words
600
predicting a difficult
550
Hypothesis: Adding semantic cues could reduce the costs of processing cues, but it 500 successor.
difficult easy control
could also add specific costs of processing semantic information. words
Removing semantic cues would affect the cueing effect.

The use of congruency cues conveyed by the previous trial is hard to obtain, but it is possible under certain conditions:
• AS PREDICTED, cues need to be presented in non-conflict trials (i.e., processing the cue involves the same resources required to resolve potential conflicts).
• Processing conflict cues is costly, even when they are conveyed by SEMANTICALLY TRANSPARENT CUES. They are only used when conflict is infrequent.
• Cues predicting a CONGRUENT SUCCESSOR are easier to follow, but this may be due to cognitive shortcuts, such as responding to the distractor.
• Arbitrary CUES BASED ON THE PREVIOUS RESPONSE can be used with no cost if they occur in non-conflict trials and with infrequent incongruent trials.
• In all these cases, cueing was explicit. It is an open question whether people can learn to use these cues implicitly (but see Jiménez et al., 2020)

PID2020-116942GB-I00

send me your comments to:

You can download this
View publication stats
[email protected] poster from Research Gate