Chapter 4 External Focus of Attention

Wulf’s experiments

In the late 1990s the motor-learning researcher Gabriele Wulf and her colleagues taught people to balance on a ski simulator and a stabilometer, and compared two kinds of instruction. One directed attention to the body (“put pressure on your outer foot”); the other to the effect of the movement on the equipment (“put pressure on the outer wheels”). The external instruction led to better learning (Wulf et al. 1998). Over the following two decades, the same comparison was made in golf, basketball, jumping, swimming, music and many other skills, usually with the same direction of result.

Evidence

Wulf proposed the constrained-action hypothesis: attending to one’s own body movements makes people control them consciously, which interferes with the automatic processes that normally coordinate a skilled movement, whereas attending to the movement’s effect lets those processes work (Wulf et al. 2001). In support, studies found less muscle activity (EMG) for the same result under an external focus, for example in biceps curls (Vance et al. 2004) and in basketball free throws, together with greater accuracy (Zachry et al. 2005).

In music and singing

Evidence

Duke and colleagues found that pianists performed a passage more accurately when focusing on the sound or on the movement of the piano’s hammers than when focusing on their fingers (Duke et al. 2011). In singing, Atkins and Duke found that untrained singers’ tone changed with their focus of attention (Atkins and Duke 2013); in later studies with trained singers, tone was rated highest when attention was directed to targets outside and far from the body, such as a point on the far wall, rather than to the soft palate or to producing vibrato (Atkins 2013).

How large is the effect?

Evidence

A meta-analysis by Chua and colleagues pooled 73 studies of performance and 40 of learning, and reported benefits of an external focus for performance (Hedges’ $g = 0.26$ ), retention ( $g = 0.58$ ), transfer ( $g = 0.58$ ) and muscle activity ( $g = 0.83$ ) (Chua et al. 2021). A reanalysis by McKay and colleagues found moderate to strong evidence of publication bias in every one of these analyses; after correcting for it, the estimates fell to 0.01 for performance, 0.15 for retention, 0.09 for transfer and 0.06 for EMG, with the data favouring no average effect, and with clear differences between studies that are not yet explained (McKay et al. 2024).

Calculation

What these effect sizes mean in practice: how often a person with an external focus would outperform one with an internal focus, and how much the two distributions overlap.

import math
from statistics import NormalDist

N = NormalDist()

def describe(g):
    sup = N.cdf(g / math.sqrt(2))                 # probability of superiority
    overlap = 2 * N.cdf(-abs(g) / 2)              # overlapping coefficient
    return f"g = {g:5.2f}: external focus better {sup:5.1%} of the time, distributions overlap {overlap:5.1%}"

print("Chua et al. 2021, pooled estimates")
for name, g in (("performance", 0.264), ("retention", 0.583), ("EMG (less activity)", 0.833)):
    print(f"  {name:20s} {describe(g)}")
print("McKay et al., the same literature corrected for publication bias")
for name, g in (("performance", 0.01), ("retention", 0.15), ("EMG (less activity)", 0.06)):
    print(f"  {name:20s} {describe(g)}")
# Chua et al. 2021, pooled estimates
#   performance          g =  0.26: external focus better 57.4% of the time, distributions overlap 89.5%
#   retention            g =  0.58: external focus better 66.0% of the time, distributions overlap 77.1%
#   EMG (less activity)  g =  0.83: external focus better 72.2% of the time, distributions overlap 67.7%
# McKay et al., the same literature corrected for publication bias
#   performance          g =  0.01: external focus better 50.3% of the time, distributions overlap 99.6%
#   retention            g =  0.15: external focus better 54.2% of the time, distributions overlap 94.0%
#   EMG (less activity)  g =  0.06: external focus better 51.7% of the time, distributions overlap 97.6%

Calculation

Why small studies and publication bias go together: the number of participants needed to detect each effect, and the smallest effect a small study can declare “significant”.

import math
from statistics import NormalDist

z = NormalDist().inv_cdf

def per_group(d, alpha=0.05, power=0.80):
    """Two independent groups, two-sided test."""
    return math.ceil(2 * ((z(1 - alpha / 2) + z(power)) / d) ** 2 + z(1 - alpha / 2) ** 2 / 4)

for d in (0.83, 0.58, 0.26, 0.15):
    print(f"true effect d = {d:.2f}: {per_group(d):5d} participants per group for 80 % power")

# Many studies in this literature are small. With 15 per group,
# only an observed effect larger than this reaches p < .05:
n = 15
d_min = (z(0.975)) * math.sqrt(2 / n)
print(f"with {n} per group the smallest 'significant' observed effect is d ≈ {d_min:.2f}")
# true effect d = 0.83:    24 participants per group for 80 % power
# true effect d = 0.58:    48 participants per group for 80 % power
# true effect d = 0.26:   234 participants per group for 80 % power
# true effect d = 0.15:   699 participants per group for 80 % power
# with 15 per group the smallest 'significant' observed effect is d ≈ 0.72

The honest summary is that external-focus instructions are cheap, safe and supported by many experiments, but that the average size of their benefit is uncertain and may be small once unpublished null results are taken into account. Where they work, they may work for some tasks, learners and wordings and not others. That is a reason to test specific instructions properly, not to drop them.

Touch as an anchor

Many singing teachers, the author among them, combine an external focus with touch: for example, the student lightly holds the teacher’s hand or rests a hand on a stable surface while singing towards a point in the room. The idea is that a steady, non-threatening touch gives the body a sense of safety and a reference that is not the larynx, so that attention can go outward.

Hypothesis · not established

Not established. A light, steady hand contact combined with an external auditory-spatial focus reduces unnecessary laryngeal-area muscle activity within minutes, and so improves pitch accuracy in beginners. The hypothesis predicts lower surface-EMG activity over the anterior neck and a smaller mean absolute pitch error (in cents) after five minutes with the combined cue than after five minutes with an internal-focus instruction. It would be refuted by the trial in Chapter 8 if the combined cue did not reduce pitch error by the pre-specified margin. Because touch and focus are combined, a positive result would not show which of the two is responsible.

Note

The touch described here is a light, consented contact, such as holding hands or resting a hand on a table. Teachers should always ask first, offer a non-contact alternative (such as holding a smooth stone or the back of a chair), and never touch the student’s neck or throat.

References

Atkins, Rebecca L., and Robert A. Duke. 2013. “Changes in Tone Production as a Function of Focus of Attention in Untrained Singers.” International Journal of Research in Choral Singing 4 (2): 28–36.

Atkins, Rebecca Lynn. 2013. “Focus of Attention Affects Singer’s Tone Production.” PhD thesis, University of Texas at Austin.

Chua, Lee-Kuen, Judith Jimenez-Diaz, Rebecca Lewthwaite, Taewon Kim, and Gabriele Wulf. 2021. “Superiority of External Attentional Focus for Motor Performance and Learning: Systematic Reviews and Meta-Analyses.” Psychological Bulletin 147 (6): 618–45.

Duke, Robert A., Carla Davis Cash, and Sarah E. Allen. 2011. “Focus of Attention Affects Performance of Motor Skills in Music.” Journal of Research in Music Education 59 (1): 44–55.

McKay, Brad, Abbey E. Corson, Jeswende Seedu, et al. 2024. Reporting Bias, Not External Focus: A Robust Bayesian Meta-Analysis and Systematic Review of the Attentional Focus Literature. SportRxiv preprint, version 3.

Vance, J., G. Wulf, T. Töllner, N. H. McNevin, and J. Mercer. 2004. “EMG Activity as a Function of the Performer’s Focus of Attention.” Journal of Motor Behavior 36: 450–59.

Wulf, Gabriele, Markus Höß, and Wolfgang Prinz. 1998. “Instructions for Motor Learning: Differential Effects of Internal Versus External Focus of Attention.” Journal of Motor Behavior 30 (2): 169–79.

Wulf, Gabriele, Nancy H. McNevin, and Charles H. Shea. 2001. “The Automaticity of Complex Motor Skill Learning as a Function of Attentional Focus.” Quarterly Journal of Experimental Psychology 54A: 1143–54.

Zachry, T., G. Wulf, J. Mercer, and N. Bezodis. 2005. “Increased Movement Accuracy and Reduced EMG Activity as the Result of Adopting an External Focus of Attention.” Brain Research Bulletin 67: 304–9.

Fan, Y. (2026). The Biomechanics of Phonation and Attentional Shift: Myoelastic-Aerodynamic Theory, the Singer's Formant, Auditory-Motor Feedback and External Focus of Attention [发声的生物力学与注意力转移] (Version 1.0.0). 5D Library. https://doi.org/10.5281/zenodo.23719625 · Licensed under Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International (CC BY-NC-ND 4.0).