DESCRIPTION: A study from Illinois involved five weeks of training in cognitive reappraisal and attention steering. Only attention changed; what this means for emotion regulation in everyday life, and what the press release failed to mention.

Ten sessions spent practising reframing. Attention shifted.

A research team at the University of Illinois taught students two strategies for emotion regulation over five weeks: diverting attention away from the emotional stimulus or reframing the situation in a positive light. By the end, the participants were using attention redirection more frequently in their everyday lives. There was little progress with reframing. The authors state this openly, though the press release does not mention it.

When you know full well that you could see it differently

You’ll recognise the situation. Someone says something in a meeting, in the kitchen, in the car, and you feel it hit you. At the same time, a second thought process is running in your head, which knows that the other person is under pressure, that the remark says nothing about you. That could reframe the scene into a milder interpretation within a second.

And yet it doesn’t happen. You say the thing you didn’t want to say, or you fall silent in a way that speaks louder than any words. Later, in the car, the kinder interpretation comes to you of its own accord. It was there all along. It just came too late.

Anyone who recognises this in themselves has a practical question: can this ability be trained so that it kicks in automatically in the moment? The group led by Sanda and Florin Dolcos has investigated this.

What was studied in Illinois

The study builds on an earlier investigation involving veterans. This time, the participants were students aged between 18 and 22 – in other words, people under constant, moderate stress, rather than a clinical group. Anyone currently undergoing treatment or taking medication was excluded.

Two strategies were practised. The first is called ‘Focused Attention’: consciously directing one’s gaze to the neutral background of an image. In the case of a photograph of a road traffic accident, this means looking at the trees, the sky, or the undamaged cars at the side of the road. The second is called ‘cognitive reappraisal’: reinterpreting the same scene in a less distressing way.

Both were practised using images and the participants’ own material, including distressing memories and worries about the future—ten sessions, two per week, over five weeks. Participants were also asked to apply both strategies in their everyday lives and to report after each session how often they had done so and with what success.

Measurements were taken before and after: mood questionnaires, image evaluations, eye movements via eye-tracking, and, for part of the group, a resting-state MRI scan.

Attention shifted, but reinterpretation did not.

The most interesting finding of the study is a null result, and the authors make no secret of it.

The daily use of attention redirection increased significantly (t(21) = 2.672; p = 0.014; d = 0.57). The daily use of reappraisal remained unchanged (t(21) = 0.400; p = 0.693; d = 0.085). The questionnaire revealed the same pattern: the scale for positive attention shifting increased (p = 0.020), whilst the scale for positive reappraisal remained unchanged (p = 0.14). The ability to generate alternative interpretations of a situation also remained unchanged.

The authors draw the obvious conclusion from this. The training had ‘primarily strengthened the early attention-related regulatory processes, whilst higher-order cognitive restructuring processes such as reappraisal and the generation of alternatives probably require more targeted or prolonged training to bring about behavioural changes’. And in the concluding section: the changes had come about ‘without increased recourse to effortful reappraisal strategies’.

Ten guided sessions, with exercise materials, homework and weekly feedback. That is more structured practice than most readers of a self-help guide will ever invest in. And yet the reinterpretation did not take hold.

Why this sequence is already established in James Gross’s work

The result is less surprising when you compare it with James Gross’s process model. Gross ranks the regulation strategies according to when they intervene in the course of an emotion. At the very top are situation selection and situation modification. Next comes attention redirection. Only then does cognitive change – that is, re-evaluation – follow. Finally, there is reaction modulation: the suppression of expression.

The earlier a strategy is applied, the less effort it requires. Redirecting attention involves shifting one’s gaze. Reinterpreting a situation requires holding the old interpretation in working memory, generating a new one, weighing both, and asserting the new one against the current emotion. This is work carried out by the prefrontal cortex.

So the Illinois group trained in both the low-cost and high-cost strategies in the same programme. It was the low-cost strategy that showed improvement.

Reinterpretation requires the prefrontal cortex, and this is in short supply under pressure.

Amy Arnsten has shown how the prefrontal cortex behaves under stress. Moderate stress improves its performance. High stress shuts it down, shifting control inwards to the limbic system. At the very moment when you need your analytical abilities most urgently, they are least available.

This results in an unpleasant asymmetry. Re-evaluation works well where you need it least: in a calm conversation, in retrospect, in the car on the way home. It fails where it really matters. Directing attention, on the other hand, remains available because it requires hardly any effort. A glance out of the window, naming three objects in the room, standing up and walking to the coffee machine.

Anyone who has been wondering for years why they have mastered these techniques yet still cannot draw on them at the crucial moment will find part of the answer here. The problem lies in the demands of the technique.

What the eye movements revealed

Eye-tracking supports this. After the training, the participants spent less time focusing on the emotional foreground of negative images and more time on the neutral background. The interaction was statistically significant (F(1,22) = 4.66; p = 0.042). They did this without being prompted. Their gaze had learnt something that no longer required a conscious decision.

This is the form of learning that emotion regulation is all about. Knowledge of what one ought to do is located in the prefrontal cortex and fails under pressure. Practised routines are located elsewhere and remain accessible. A gaze that wanders to the window of its own accord belongs to the second category.

Limitation: At this stage, the control group comprised 6 participants.

Brain scans and what they reveal

In 23 participants, measurements were taken whilst at rest to determine the extent to which different brain regions synchronise with one another. The coupling between the amygdala and several cortical areas decreased. The coupling between a gaze-control area in the frontal lobe and the visual cortex increased. This fits with the picture of a system that reacts less strongly to emotional stimuli and directs the gaze more effectively.

However, this measurement carries less weight than a headline might suggest. There were 23 participants, eight of whom were in the control group. The statistical threshold at the level of individual pixels was set without correction. One area did not survive the stricter correction. The measurements were taken at rest, not under stress, and resting-state connectivity, as pointed out, says nothing about the direction of an influence. Above all, no correlation is reported between the changes in the brain and the changes in behaviour. The link between the two is the authors’ interpretation, not a measurement.

The authors themselves state that the imaging results should be “treated with caution”.

What the press release said

The report circulating on social media was headlined as stating that emotion regulation training had reduced stress and altered neural connectivity in the brain. Both parts are roughly correct. However, the study’s text provides a more nuanced account.

The claim that participants found negative images less distressing after the training is based on a comparison within the training group. The overarching statistical test, which examines the group, image type and time point together, was not statistically significant (F(1,22) = 1.359; p = 0.256). It is only at a lower level, in the individual comparisons, that the value which made it into the press release appears.

The situation is similar with respect to the perceived sense of control. Here, too, the interaction between the groups was not significant (p = 0.090). It was the training group itself that showed a change.

What remains clear is the scale for negative mood. It fell in the training group from 30.2 to 24.1; the interaction was clear (F(1,29) = 9.953; p = 0.004). This is a genuine result. It is also the result most at risk due to the nature of the group allocation.

The allocation that throws everything into disarray

The groups were not allocated at random. Those who exceeded the stress thresholds during the initial assessment were given priority for the training group so that they would not have to wait. Humanly speaking, this was the right thing to do; methodologically, it was costly. Those who start high inevitably drop in the second measurement due to statistical gravity. This is precisely what the authors acknowledge: the results could ‘in part reflect regression to the mean’. They responded by recalculating the figures reported in the text using a narrower sample that excluded the eight participants with the highest stress levels. The pattern remained the same. This does not dispel the objection, nor do the authors claim that it does.

There are several other issues: the control group waited and did nothing else; there was therefore no sham programme to balance out the attention and expectations. The study was not registered in advance. 34 of the 39 participants analysed were female. And halfway through the study, the pandemic meant that face-to-face sessions were replaced with interactive PDF files sent via email.

There were no follow-up assessments. Nobody knows whether any of the effects remain after six months. An acute stress test was never carried out. It remains to be seen whether what was learnt will stand up to the test of real-life challenges.

How you can apply this in everyday life

No specific instructions can be derived from a study of this scale. It provides a clue that points in the same direction as the process model and the research on the prefrontal cortex under stress.

The suggestion is this: if you want to have a single strategy at your disposal in the heat of the moment, choose the cheapest one. Shift your gaze. Stand up. Change rooms. Name three things you can see. It costs almost nothing and therefore remains available when the more complex tools fail.

Reinterpretation still has its place. It just comes later. After the conversation, on the way home, the next morning, when the prefrontal cortex is functioning again. Anyone who wants to carry out their reinterpretation in the very moment when emotions are running high is demanding a performance from the brain that it cannot deliver in that state. Ten guided sessions have done nothing to change that.

And the relief in this: the fact that the insightful interpretation only occurs to you in the car is not a sign of sluggishness. It is simply the sequence in which this system operates.

This is where the protocol I set out in THE 90-SECOND REGULATION SYSTEM begins. It opens with the easy steps rather than with thinking: noticing what the body is doing (SENSE), dropping the fight against it (YIELD), putting the pressure into words (SPEAK). Appraisal comes afterwards (THINK/TRANSFORM), then action (ENGAGE), then rehearsal (MERGE). The sequence follows availability under stress rather than difficulty.

The Illinois study proves none of this; it is too small and too shaky. It does offer one inconvenient piece of evidence in the same direction. Ten guided sessions were not enough to carry the expensive strategy into daily life. Anyone who makes reappraisal the first step is building a system on the part that fails first under pressure.

Key points in brief

• A study by the University of Illinois trained 20 students in attention steering and cognitive reappraisal over five weeks, with 19 students serving as a control group.

• The daily use of attention steering increased (p = 0.014; d = 0.57). The use of reappraisal remained unchanged (p = 0.693).

• The authors attribute this to the fact that reappraisal, as a more resource-intensive strategy, requires longer or more targeted training.

• This is consistent with James Gross’s process model: attention redirection takes effect sooner and requires less effort than cognitive re-evaluation.

• Under high stress, the performance of the prefrontal cortex declines. Consequently, the very strategy that relies most heavily on it fails.

• Eye-tracking showed that, following training, participants’ gaze was directed more frequently towards neutral parts of the image even without prompting (p = 0.042).

• The brain scans, based on 23 participants, were collected whilst at rest and cannot be correlated with the behavioural changes.

• Two of the three findings reported in the press release stem from comparisons within the training group, after the main test failed to yield significant results.

• The groups were not allocated at random; participants under greater stress were given priority for the training. Regression towards the mean cannot be ruled out.

• As for everyday life, the fact remains: in the moment itself, what we see is more reliable than what we think. Reinterpretation comes later – and that is fine.

Sources

· Hohl K, Dolcos S, Mull DD, et al. Enhancing resilience, flexibility and well-being through cognitive-emotional training: behavioural and neural evidence. Scientific Reports 2026;16(1):24914. https://doi.org/10.1038/s41598-026-63059-0

· Technology Networks: Emotion Regulation Training Reduced Distress and Altered Brain Connectivity, 19 August 2026. https://www.technologynetworks.com/neuroscience/news/emotion-regulation-training-reduced-distress-and-altered-brain-connectivity-415767

· Gross JJ. Emotion regulation: current status and prospects. Psychological Inquiry 2015;26(1):1–26. https://doi.org/10.1080/1047840X.2014.940781

· Arnsten AFT. Stress signalling pathways that impair prefrontal cortex structure and function. Nature Reviews Neuroscience 2009;10(6):410–422. https://doi.org/10.1038/nrn2648

· Dolcos S, Hu Y, Williams C, et al. Cultivating affective resilience: proof-of-principle evidence of translational benefits from a novel cognitive-emotional training intervention. Frontiers in Psychology 2021;12:585536. https://doi.org/10.3389/fpsyg.2021.585536