The Attention Training Technique (ATT) is delivered as a spoken script. A voice — a therapist in the clinic, a recording elsewhere — names sounds around the listener and tells them where to put their attention next. Wells’ 2009 treatment manual prints that script in full, and its wording is doing deliberate work: almost every line exists to prevent a specific way of getting the exercise wrong[1].

This page describes what the script asks for. It is a summary, not a reproduction — the verbatim text belongs to the manual.

Before the instructions begin

The script opens by setting up the room and the stance. The listener fixes their gaze on a spot — the manual has them focus on a visual fixation point, such as a dot, and hold that visual focus throughout[1]. Several sounds are introduced at different locations: a minimum of three actual competing sounds in the room, two more nearby, and two further off in the distance[1].

Some of those locations are what the manual calls potential sounds — a direction you are asked to attend to where nothing may actually be sounding at the time[1]. Attending to a quiet direction is part of the exercise, not a fault in the setup.

The introduction also tells the listener what not to do with their thoughts. Intruding thoughts, images and feelings are to be treated as additional noise and not resisted; the script is explicit that you must not try to stop them, and asks you to think of them as passing inner noises[1]. The aim, as the manual puts it, is not to remove awareness of inner events but to practise controlling attention in a particular way[1].

Phase 1 — selective attention (about 5 minutes)

The script names one sound and asks for full attention on it, to the exclusion of the others. Then it names another, and another. Each instruction includes the recovery clause: if your attention begins to stray, or is captured by another sound, refocus on the one being named[1].

That clause is the phase in miniature. The script assumes your attention will be captured. Being pulled away and hauling yourself back is the repetition being trained.

Phase 2 — attention switching (about 5 minutes)

The same naming, faster. The script moves the listener between sounds at roughly ten-second intervals, then increases the pace to one sound every five seconds[1].

This is the phase people report as the hardest. In a component study that took per-block effort ratings, the switching condition drew the highest effort ratings of the three phases[2].

Phase 3 — divided attention (about 2 minutes)

The instruction changes shape. Instead of one sound at a time, the script asks the listener to expand the breadth and depth of their attention and take in all of the sounds at once, including a covert count of how many can be heard simultaneously[1]. It is the shortest phase and the one that gives the session its ending.

Why the wording matters

Read as a whole, the script is defensive. It repeatedly rules out the misreadings that would turn the exercise into something else:

  • Not thought suppression. The manual states the technique counteracts the use of active thought suppression, and that it is not intended to produce a blank mind[1].
  • Not relaxation or symptom management. It should not be employed as a distraction, avoidance, or symptom-management strategy, and is not intended as an emotion-management strategy[1].
  • Not meant to get easy. The exercise is designed to be attentionally demanding and not to become substantially less demanding with practice — otherwise, Wells writes, it would not systematically strengthen control processes[1].

The manual also lists the strategies that quietly defeat the exercise while it appears to be going fine: thought suppression, daydreaming, and diverting attention to worry[1].

Who says the words

In the original clinical setting the therapist reads the script live, using the sounds available in the consulting room. Wells notes that recorded versions of sounds have also been used to deliver ATT[1] — which is what makes an audio or app version of the exercise a recognisable delivery of the same protocol rather than a different practice.

What the script does not fix is the sound bed. The manual specifies varying the sounds and their arrangement between sessions, because that variety offsets the effects of practice on task difficulty[1]. Same words, different soundscape.

For the exercise in full — where it came from, how it is supposed to work, and what published research reports — see the Attention Training Technique explained.

Questions and answers

Is there an official ATT script?
Yes. Wells' 2009 treatment manual prints a therapist script for the exercise, covering the introduction and all three phases. It is the reference version practitioners work from.
Do I have to say the words out loud?
No. The script is spoken by a therapist or a recording; the person practicing only listens and follows. Recorded versions of the sounds have also been used to deliver the exercise.
Does the script change between sessions?
The instructions stay the same; the sounds and their arrangement are deliberately varied, which Wells describes as offsetting the effects of practice on task difficulty.

References

  1. Wells, A. (2009). Metacognitive Therapy for Anxiety and Depression. Guilford Press. Chapter 4: Attention Training Techniques.
  2. Rosenbaum, D., Maier, M. J., Hudak, J., Metzger, F. G., Wells, A., Fallgatter, A. J., & Ehlis, A.-C. (2018). Neurophysiological correlates of the attention training technique: A component study. NeuroImage: Clinical, 19, 1018–1024. https://pmc.ncbi.nlm.nih.gov/articles/PMC6039840/

Published by the makers of Heed, an app for practicing ATT-style sessions. iPhone · Android. Heed is a self-practice tool inspired by ATT research — not therapy, and not a medical device.