Neurofeedback and archers' scores: Paul 2012 study

Twenty-four Indian university archers were split into two groups of 12 (8 men and 4 women each). The first did twelve neurofeedback sessions over four weeks, learning to raise the sensorimotor rhythm of the electroencephalogram, a band between 12 and 15 Hz; the second did nothing. Before and after, everyone shot a 72-arrow competition at 70 metres.

The score did not change to any real extent: the neurofeedback group went from 527.33 to 529.92 points and the control group from 526.58 to 526.67, and the difference between the groups after the intervention is not significant (p = 0.561). The trained group's 'precision' also stayed exactly the same, 5.17 before and after.

What did change were some measures of how the archers felt. In the neurofeedback group self-rated arousal before competition fell from 6.83 to 5.75 and after competition from 7.29 to 6.25, on a grid from 1 to 9; mean heart rate while shooting fell from 116.08 to 113.75 beats per minute, and the ratio of sensorimotor rhythm to theta waves at rest rose from 0.3314 to 0.3950. Compared with the control group after the intervention, pleasure and arousal before competition, arousal after competition and the sensorimotor-to-theta ratio come out significant; heart rate does not.

The study was carried out at Guru Nanak Dev University in Amritsar, as the first author's master's thesis. The archers were all right-handed, with a mean age of 21.96 and 4.31 years of experience, and none had done neurofeedback before. The trained group had three 20-minute sessions a week with a ProComp5 Infiniti unit: one electrode at the top of the head (Cz), references on the earlobes, and the task of 'keeping the animation moving' on the screen, which ran on, with a reward tone, when the sensorimotor rhythm rose while theta (4-7 Hz) and high beta (22-26 Hz) stayed low.

The competition followed the FITA Olympic round, 72 arrows in 12 ends of six at 70 metres. During shooting a Polar S410 heart rate monitor recorded the average heart rate every five seconds. Before and after competition each archer marked on a 9-by-9 grid how good they felt (pleasure) and how activated they felt (arousal). The EEG was recorded only seated, at the start and end of the sessions, never while shooting. The comparisons use paired t-tests within each group and independent t-tests on the final measures.

The control group received no treatment at all, not even sham neurofeedback, so the changes in self-rated feelings may come from the attention received and from expectation rather than from the trained brain rhythm. The participants were chosen 'at random', but the paper does not say how they were allocated to the groups. The comparison between groups uses only the final values, ignoring the starting ones: the control group began with lower pre-competition pleasure (5.83 against 6.67) and higher post-competition arousal (8.00 against 7.29), so part of the final difference was there from the outset.

The tables have several problems. The rise in the sensorimotor-to-theta ratio in the trained group has t = 4.980 but is marked not significant, while the summary calls it significant; in the second table the standard errors for precision and score appear to be swapped; two correlations of -0.034 and -0.041 are marked significant, which is impossible with 24 people. 'Precision' is not clearly defined, and the discussion states that the trained group shot more Xs, while the table gives the same value before and after. The conclusion that neurofeedback makes scoring more consistent finds no support in the data presented.

For anyone who shoots, the result that matters is the one on score, and there nothing happened: twelve neurofeedback sessions changed neither points nor precision in a 72-arrow competition at 70 metres. The trained archers reported feeling less activated before and after competition, but without a group receiving a sham treatment there is no way to separate the effect of brain training from that of having been looked after for a month.

The study therefore does not support neurofeedback as a way to shoot better. It leaves open only the question of whether this kind of training changes the archer's state in competition, a question that needs a sham control and an EEG recorded during the shot.